clones/php-src
1
0
Fork 0
mirror of https://github.com/php/php-src.git synced 2024-09-21 18:07:23 +00:00
Code Issues Packages Projects Releases Wiki Activity
553a0c52b1
php-src/ext/standard/string.c

5992 lines
142 KiB
C
Raw Normal View History

First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
/*
+----------------------------------------------------------------------+
Remove year range from copyright notice
2019-01-30 09:00:23 +00:00
| Copyright (c) The PHP Group |
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
+----------------------------------------------------------------------+
bump year and license version
2006-01-01 12:51:34 +00:00
| This source file is subject to version 3.01 of the PHP license, |
License update
1999-07-16 13:13:16 +00:00
| that is bundled with this package in the file LICENSE, and is |
updating license information in the headers.
2003-06-10 20:04:29 +00:00
| available through the world-wide-web at the following url: |
bump year and license version
2006-01-01 12:51:34 +00:00
| http://www.php.net/license/3_01.txt |
License update
1999-07-16 13:13:16 +00:00
| If you did not receive a copy of the PHP license and are unable to |
| obtain it through the world-wide-web, please send a note to |
| license@php.net so we can mail you a copy immediately. |
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
+----------------------------------------------------------------------+
Need a PHPAPI version of basename for some stuff I am working on. Also fixed a bug along the way in the basename function. If it was fed something like "filename.ext/////" it would return the string with all the slashes whereas if you fed it "/path/filename.ext////" it would get it right. @ Fixed basename() bug where "file.ext///" would not return the same @ as "/path/file.ext///" (Rasmus)
2000-05-23 20:16:14 +00:00
| Authors: Rasmus Lerdorf <rasmus@php.net> |
Author Name Character Fix
2018-04-03 16:52:49 +00:00
| Stig Sæther Bakken <ssb@php.net> |
Future-proof email addresses...
2018-11-01 16:30:28 +00:00
| Zeev Suraski <zeev@php.net> |
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
+----------------------------------------------------------------------+
*/
#include <stdio.h>
#include "php.h"
Fix warning.
2002-09-26 19:18:35 +00:00
#include "php_rand.h"
Clean up php3.*\.h files. The files itself are renamed, and references in all .*\.[ch] files were changed. There is a slight chance that my script missed a few changes, please correct them manually.
1999-12-04 19:19:57 +00:00
#include "php_string.h"
Tried to centralize global variable registration as much as possible: - Added $HTTP_ENV_VARS[] and $HTTP_SERVER_VARS[] support, which similarly to $HTTP_GET_VARS[], contain environment and server variables. Setting register_globals to Off will now also prevent registration of the environment and server variables into the global scope (Zeev) - Renamed gpc_globals to register_globals (Zeev) - Introduced variables_order that deprecates gpc_order, and allows control over the server and environment variables, in addition to GET/POST/Cookies (Zeev)
2000-01-28 17:24:53 +00:00
#include "php_variables.h"
Remove checks for locale.h, setlocale, localeconv The `<loccale.h>` header file, setlocale, and localeconv are part of the standard C89 [1] and on current systems can be used unconditionally. Since PHP 7.4 requires at least C89 or greater, the `HAVE_LOCALE_H`, `HAVE_SETLOCALE`, and `HAVE_LOCALECONV` symbols defined by Autoconf in configure.ac [2] can be ommitted and simplifed. The bundled libmagic (file) has also been patched already in version 5.35 and up in upstream location so when it will be patched also in php-src the check for locale.h header is still left in the configure.ac and in windows headers definition file. [1] https://port70.net/~nsz/c/c89/c89-draft.html#4.4 [2] https://git.savannah.gnu.org/cgit/autoconf.git/tree/lib/autoconf/headers.m4 Omit the bundled libmagic files
2019-02-24 09:21:25 +00:00
#include <locale.h>
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
#ifdef HAVE_LANGINFO_H
# include <langinfo.h>
#endif
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
MFB: LC_MESSAGES support
2003-05-27 00:42:39 +00:00
#ifdef HAVE_LIBINTL
# include <libintl.h> /* For LC_MESSAGES */
#endif
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
Added sscanf() function.
2000-06-06 18:58:15 +00:00
#include "scanf.h"
#include "zend_API.h"
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
#include "zend_execute.h"
#include "php_globals.h"
Move include into source file
1999-12-28 18:47:59 +00:00
#include "basic_functions.h"
Move smart_str implementation into Zend/ So we can use it there as well... For now I've retained the zend_smart_str_public.h header, though it would probably be better to just move that one struct into zend_types.h.
2014-09-20 20:42:02 +00:00
#include "zend_smart_str.h"
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
#include <Zend/zend_exceptions.h>
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
#ifdef ZTS
#include "TSRM.h"
#endif
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
/* For str_getcsv() support */
#include "ext/standard/file.h"
Move utf8_encode and utf8_decode to ext/standard
2016-10-13 22:33:33 +00:00
/* For php_next_utf8_char() */
#include "ext/standard/html.h"
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
(PHP pathinfo) Fixed leaks, saved some memory allocations, added more error checking.
2000-09-11 13:52:22 +00:00
#define STR_PAD_LEFT 0
#define STR_PAD_RIGHT 1
#define STR_PAD_BOTH 2
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
#define PHP_PATHINFO_DIRNAME 1
#define PHP_PATHINFO_BASENAME 2
#define PHP_PATHINFO_EXTENSION 4
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
#define PHP_PATHINFO_FILENAME 8
#define PHP_PATHINFO_ALL (PHP_PATHINFO_DIRNAME | PHP_PATHINFO_BASENAME | PHP_PATHINFO_EXTENSION | PHP_PATHINFO_FILENAME)
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
#define STR_STRSPN 0
#define STR_STRCSPN 1
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ register_string_constants */
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
void register_string_constants(INIT_FUNC_ARGS)
{
master renames phase 1
2014-08-25 17:24:55 +00:00
REGISTER_LONG_CONSTANT("STR_PAD_LEFT", STR_PAD_LEFT, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("STR_PAD_RIGHT", STR_PAD_RIGHT, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("STR_PAD_BOTH", STR_PAD_BOTH, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("PATHINFO_DIRNAME", PHP_PATHINFO_DIRNAME, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("PATHINFO_BASENAME", PHP_PATHINFO_BASENAME, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("PATHINFO_EXTENSION", PHP_PATHINFO_EXTENSION, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("PATHINFO_FILENAME", PHP_PATHINFO_FILENAME, CONST_CS | CONST_PERSISTENT);
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
REGISTER_LONG_CONSTANT("PATHINFO_ALL", PHP_PATHINFO_ALL, CONST_CS | CONST_PERSISTENT);
Added constants for locale categories.
2000-12-10 15:55:34 +00:00
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* If last members of struct lconv equal CHAR_MAX, no grouping is done */
master renames phase 1
2014-08-25 17:24:55 +00:00
REGISTER_LONG_CONSTANT("CHAR_MAX", CHAR_MAX, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("LC_CTYPE", LC_CTYPE, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("LC_NUMERIC", LC_NUMERIC, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("LC_TIME", LC_TIME, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("LC_COLLATE", LC_COLLATE, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("LC_MONETARY", LC_MONETARY, CONST_CS | CONST_PERSISTENT);
REGISTER_LONG_CONSTANT("LC_ALL", LC_ALL, CONST_CS | CONST_PERSISTENT);
Don't define LC_MESSAGES, if LC_MESSAGES is not defined. PR: #8259
2000-12-14 15:58:14 +00:00
# ifdef LC_MESSAGES
master renames phase 1
2014-08-25 17:24:55 +00:00
REGISTER_LONG_CONSTANT("LC_MESSAGES", LC_MESSAGES, CONST_CS | CONST_PERSISTENT);
Fix Win32 build
2000-12-11 16:27:48 +00:00
# endif
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
}
Fix folding and clean up some extensions
2001-06-06 13:06:12 +00:00
/* }}} */
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
Cleanup (avoid reallocatios and side effects in php_strip_tags)
2015-07-01 16:04:18 +00:00
int php_tag_find(char *tag, size_t len, const char *set);
More maintainer-mode cleanups
1999-12-05 14:16:37 +00:00
Make basic/string functions thread-safe
1999-11-28 00:31:02 +00:00
/* this is read-only, so it's ok */
Add "const". Move constant strings to read-only memory.
2020-09-07 18:35:48 +00:00
ZEND_SET_ALIGNED(16, static const char hexconvtab[]) = "0123456789abcdef";
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
/* localeconv mutex */
#ifdef ZTS
static MUTEX_T locale_mutex = NULL;
#endif
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_bin2hex */
Use better data structures (incomplete)
2014-02-18 21:12:05 +00:00
static zend_string *php_bin2hex(const unsigned char *old, const size_t oldlen)
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
{
Use better data structures (incomplete)
2014-02-18 21:12:05 +00:00
zend_string *result;
signed/unsigned warning removed
1999-08-12 10:53:29 +00:00
size_t i, j;
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
master renames phase 1
2014-08-25 17:24:55 +00:00
result = zend_string_safe_alloc(oldlen, 2 * sizeof(char), 0, 0);
Remove _GNU_SOURCE, add local heap sort
2013-01-10 12:09:14 +00:00
ws fixes.
2002-10-02 17:56:04 +00:00
for (i = j = 0; i < oldlen; i++) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[j++] = hexconvtab[old[i] >> 4];
ZSTR_VAL(result)[j++] = hexconvtab[old[i] & 15];
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[j] = '\0';
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
Use better data structures (incomplete)
2014-02-18 21:12:05 +00:00
return result;
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
}
Fix folding and clean up some extensions
2001-06-06 13:06:12 +00:00
/* }}} */
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_hex2bin */
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
static zend_string *php_hex2bin(const unsigned char *old, const size_t oldlen)
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
{
size_t target_length = oldlen >> 1;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_string *str = zend_string_alloc(target_length, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
unsigned char *ret = (unsigned char *)ZSTR_VAL(str);
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
size_t i, j;
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
for (i = j = 0; i < target_length; i++) {
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
unsigned char c = old[j++];
Make bin2hex() and hex2bin() timing safe
2015-08-04 20:07:13 +00:00
unsigned char l = c & ~0x20;
int is_letter = ((unsigned int) ((l - 'A') ^ (l - 'F' - 1))) >> (8 * sizeof(unsigned int) - 1);
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
unsigned char d;
Fix bug #73648 - integer overflow in substr
2017-01-16 20:11:50 +00:00
/* basically (c >= '0' && c <= '9') || (l >= 'A' && l <= 'F') */
Make bin2hex() and hex2bin() timing safe
2015-08-04 20:07:13 +00:00
if (EXPECTED((((c ^ '0') - 10) >> (8 * sizeof(unsigned int) - 1)) | is_letter)) {
d = (l - 0x10 - 0x27 * is_letter) << 4;
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
} else {
Avoid useless checks, using zend_string_efree(), in cases where the string is known to be a temporary allocated zend_string.
2018-05-08 14:30:15 +00:00
zend_string_efree(str);
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
return NULL;
}
c = old[j++];
Make bin2hex() and hex2bin() timing safe
2015-08-04 20:07:13 +00:00
l = c & ~0x20;
is_letter = ((unsigned int) ((l - 'A') ^ (l - 'F' - 1))) >> (8 * sizeof(unsigned int) - 1);
if (EXPECTED((((c ^ '0') - 10) >> (8 * sizeof(unsigned int) - 1)) | is_letter)) {
d |= l - 0x10 - 0x27 * is_letter;
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
} else {
Avoid useless checks, using zend_string_efree(), in cases where the string is known to be a temporary allocated zend_string.
2018-05-08 14:30:15 +00:00
zend_string_efree(str);
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
return NULL;
}
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
ret[i] = d;
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
}
cleanup
2014-08-13 15:43:34 +00:00
ret[i] = '\0';
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
return str;
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
}
/* }}} */
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* {{{ localeconv_r
* glibc's localeconv is not reentrant, so lets make it so ... sorta */
- Fixed bug #30630: Added a BSD based strtod function that is locale-independent. (PHP part)
2004-11-03 23:36:51 +00:00
PHPAPI struct lconv *localeconv_r(struct lconv *out)
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
{
Remove checks for locale.h, setlocale, localeconv The `<loccale.h>` header file, setlocale, and localeconv are part of the standard C89 [1] and on current systems can be used unconditionally. Since PHP 7.4 requires at least C89 or greater, the `HAVE_LOCALE_H`, `HAVE_SETLOCALE`, and `HAVE_LOCALECONV` symbols defined by Autoconf in configure.ac [2] can be ommitted and simplifed. The bundled libmagic (file) has also been patched already in version 5.35 and up in upstream location so when it will be patched also in php-src the check for locale.h header is still left in the configure.ac and in windows headers definition file. [1] https://port70.net/~nsz/c/c89/c89-draft.html#4.4 [2] https://git.savannah.gnu.org/cgit/autoconf.git/tree/lib/autoconf/headers.m4 Omit the bundled libmagic files
2019-02-24 09:21:25 +00:00
#ifdef ZTS
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
tsrm_mutex_lock( locale_mutex );
Remove checks for locale.h, setlocale, localeconv The `<loccale.h>` header file, setlocale, and localeconv are part of the standard C89 [1] and on current systems can be used unconditionally. Since PHP 7.4 requires at least C89 or greater, the `HAVE_LOCALE_H`, `HAVE_SETLOCALE`, and `HAVE_LOCALECONV` symbols defined by Autoconf in configure.ac [2] can be ommitted and simplifed. The bundled libmagic (file) has also been patched already in version 5.35 and up in upstream location so when it will be patched also in php-src the check for locale.h header is still left in the configure.ac and in windows headers definition file. [1] https://port70.net/~nsz/c/c89/c89-draft.html#4.4 [2] https://git.savannah.gnu.org/cgit/autoconf.git/tree/lib/autoconf/headers.m4 Omit the bundled libmagic files
2019-02-24 09:21:25 +00:00
#endif
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
fix compilation with vc14
2015-04-05 18:59:26 +00:00
/* cur->locinfo is struct __crt_locale_info which implementation is
hidden in vc14. TODO revisit this and check if a workaround available
and needed. */
#if defined(PHP_WIN32) && _MSC_VER < 1900 && defined(ZTS)
Fixed bug #65769 localeconv() broken in TS builds
2014-12-05 10:06:06 +00:00
{
/* Even with the enabled per thread locale, localeconv
won't check any locale change in the master thread. */
_locale_t cur = _get_current_locale();
fix bug #72024 (microtime() leaks memory)
2016-07-07 12:24:05 +00:00
*out = *cur->locinfo->lconv;
_free_locale(cur);
Fixed bug #65769 localeconv() broken in TS builds
2014-12-05 10:06:06 +00:00
}
#else
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
/* localeconv doesn't return an error condition */
fix bug #72024 (microtime() leaks memory)
2016-07-07 12:24:05 +00:00
*out = *localeconv();
Fixed bug #65769 localeconv() broken in TS builds
2014-12-05 10:06:06 +00:00
#endif
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
Remove checks for locale.h, setlocale, localeconv The `<loccale.h>` header file, setlocale, and localeconv are part of the standard C89 [1] and on current systems can be used unconditionally. Since PHP 7.4 requires at least C89 or greater, the `HAVE_LOCALE_H`, `HAVE_SETLOCALE`, and `HAVE_LOCALECONV` symbols defined by Autoconf in configure.ac [2] can be ommitted and simplifed. The bundled libmagic (file) has also been patched already in version 5.35 and up in upstream location so when it will be patched also in php-src the check for locale.h header is still left in the configure.ac and in windows headers definition file. [1] https://port70.net/~nsz/c/c89/c89-draft.html#4.4 [2] https://git.savannah.gnu.org/cgit/autoconf.git/tree/lib/autoconf/headers.m4 Omit the bundled libmagic files
2019-02-24 09:21:25 +00:00
#ifdef ZTS
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
tsrm_mutex_unlock( locale_mutex );
Remove checks for locale.h, setlocale, localeconv The `<loccale.h>` header file, setlocale, and localeconv are part of the standard C89 [1] and on current systems can be used unconditionally. Since PHP 7.4 requires at least C89 or greater, the `HAVE_LOCALE_H`, `HAVE_SETLOCALE`, and `HAVE_LOCALECONV` symbols defined by Autoconf in configure.ac [2] can be ommitted and simplifed. The bundled libmagic (file) has also been patched already in version 5.35 and up in upstream location so when it will be patched also in php-src the check for locale.h header is still left in the configure.ac and in windows headers definition file. [1] https://port70.net/~nsz/c/c89/c89-draft.html#4.4 [2] https://git.savannah.gnu.org/cgit/autoconf.git/tree/lib/autoconf/headers.m4 Omit the bundled libmagic files
2019-02-24 09:21:25 +00:00
#endif
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
return out;
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
Remove checks for locale.h, setlocale, localeconv The `<loccale.h>` header file, setlocale, and localeconv are part of the standard C89 [1] and on current systems can be used unconditionally. Since PHP 7.4 requires at least C89 or greater, the `HAVE_LOCALE_H`, `HAVE_SETLOCALE`, and `HAVE_LOCALECONV` symbols defined by Autoconf in configure.ac [2] can be ommitted and simplifed. The bundled libmagic (file) has also been patched already in version 5.35 and up in upstream location so when it will be patched also in php-src the check for locale.h header is still left in the configure.ac and in windows headers definition file. [1] https://port70.net/~nsz/c/c89/c89-draft.html#4.4 [2] https://git.savannah.gnu.org/cgit/autoconf.git/tree/lib/autoconf/headers.m4 Omit the bundled libmagic files
2019-02-24 09:21:25 +00:00
#ifdef ZTS
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ PHP_MINIT_FUNCTION */
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
PHP_MINIT_FUNCTION(localeconv)
{
locale_mutex = tsrm_mutex_alloc();
return SUCCESS;
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ PHP_MSHUTDOWN_FUNCTION */
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
PHP_MSHUTDOWN_FUNCTION(localeconv)
{
tsrm_mutex_free( locale_mutex );
locale_mutex = NULL;
return SUCCESS;
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
#endif
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Converts the binary representation of data to hex */
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
PHP_FUNCTION(bin2hex)
{
Use better data structures (incomplete)
2014-02-18 21:12:05 +00:00
zend_string *result;
fixed strspn
2014-08-17 20:56:26 +00:00
zend_string *data;
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(data)
ZEND_PARSE_PARAMETERS_END();
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result = php_bin2hex((unsigned char *)ZSTR_VAL(data), ZSTR_LEN(data));
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use better data structures (incomplete)
2014-02-18 21:12:05 +00:00
RETURN_STR(result);
add fast bin2hex string function
1999-05-16 11:12:23 +00:00
}
Will watch the diffs as much as I can.
2000-02-24 10:11:42 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Converts the hex representation of data to binary */
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
PHP_FUNCTION(hex2bin)
{
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
zend_string *result, *data;
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(data)
ZEND_PARSE_PARAMETERS_END();
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(data) % 2 != 0) {
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Hexadecimal input string must have an even length");
Fix bug #61660: bin2hex(hex2bin($data)) != $data If the input data has an odd length a warning is thrown and false is returned.
2012-04-08 20:36:50 +00:00
RETURN_FALSE;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result = php_hex2bin((unsigned char *)ZSTR_VAL(data), ZSTR_LEN(data));
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
if (!result) {
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Input string must be hexadecimal string");
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
RETURN_FALSE;
}
cleanup (use zend_string* instead of char*)
2014-08-13 10:51:48 +00:00
RETVAL_STR(result);
Added fabled hex2bin() function
2011-06-03 00:12:22 +00:00
}
/* }}} */
fix folding
2007-06-05 13:35:26 +00:00
static void php_spn_common_handler(INTERNAL_FUNCTION_PARAMETERS, int behavior) /* {{{ */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixed strspn
2014-08-17 20:56:26 +00:00
zend_string *s11, *s22;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long start = 0, len = 0;
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
zend_bool len_is_null = 1;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 4)
Z_PARAM_STR(s11)
Z_PARAM_STR(s22)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(start)
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
Z_PARAM_LONG_OR_NULL(len, len_is_null)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Don't throw for out of bounds offsets in strspn() Make strspn($str1, $str2, $offset, $length) behaviorally equivalent to strspn(substr($str1, $offset, $length), $str2) by not throwing for out of bounds offset. There have been two reports that this change cause issues, including bug #80285.
2020-10-27 10:23:49 +00:00
size_t remain_len = ZSTR_LEN(s11);
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
if (start < 0) {
Don't throw for out of bounds offsets in strspn() Make strspn($str1, $str2, $offset, $length) behaviorally equivalent to strspn(substr($str1, $offset, $length), $str2) by not throwing for out of bounds offset. There have been two reports that this change cause issues, including bug #80285.
2020-10-27 10:23:49 +00:00
start += remain_len;
if (start < 0) {
start = 0;
}
} else if ((size_t) start > remain_len) {
start = remain_len;
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Don't throw for out of bounds offsets in strspn() Make strspn($str1, $str2, $offset, $length) behaviorally equivalent to strspn(substr($str1, $offset, $length), $str2) by not throwing for out of bounds offset. There have been two reports that this change cause issues, including bug #80285.
2020-10-27 10:23:49 +00:00
remain_len -= start;
Check string bounds in strspn/strcspn strspn/strcspn are string search functions, and as such should throw ValueError on out-of-bounds offsets, just like strpos etc do.
2020-09-22 08:42:24 +00:00
if (!len_is_null) {
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
if (len < 0) {
Check string bounds in strspn/strcspn strspn/strcspn are string search functions, and as such should throw ValueError on out-of-bounds offsets, just like strpos etc do.
2020-09-22 08:42:24 +00:00
len += remain_len;
Don't throw for out of bounds offsets in strspn() Make strspn($str1, $str2, $offset, $length) behaviorally equivalent to strspn(substr($str1, $offset, $length), $str2) by not throwing for out of bounds offset. There have been two reports that this change cause issues, including bug #80285.
2020-10-27 10:23:49 +00:00
if (len < 0) {
len = 0;
}
} else if ((size_t) len > remain_len) {
len = remain_len;
Check string bounds in strspn/strcspn strspn/strcspn are string search functions, and as such should throw ValueError on out-of-bounds offsets, just like strpos etc do.
2020-09-22 08:42:24 +00:00
}
} else {
len = remain_len;
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Check string bounds in strspn/strcspn strspn/strcspn are string search functions, and as such should throw ValueError on out-of-bounds offsets, just like strpos etc do.
2020-09-22 08:42:24 +00:00
if (len == 0) {
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(0);
Fix chunk_split fix - avoid using floats Fix money_format - don't give strfmon more arguments then supplied Fix str[c]spn integer overflow
2007-06-06 17:59:07 +00:00
}
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
if (behavior == STR_STRSPN) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(php_strspn(ZSTR_VAL(s11) + start /*str1_start*/,
ZSTR_VAL(s22) /*str2_start*/,
ZSTR_VAL(s11) + start + len /*str1_end*/,
ZSTR_VAL(s22) + ZSTR_LEN(s22) /*str2_end*/));
Check string bounds in strspn/strcspn strspn/strcspn are string search functions, and as such should throw ValueError on out-of-bounds offsets, just like strpos etc do.
2020-09-22 08:42:24 +00:00
} else {
ZEND_ASSERT(behavior == STR_STRCSPN);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(php_strcspn(ZSTR_VAL(s11) + start /*str1_start*/,
ZSTR_VAL(s22) /*str2_start*/,
ZSTR_VAL(s11) + start + len /*str1_end*/,
ZSTR_VAL(s22) + ZSTR_LEN(s22) /*str2_end*/));
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
}
}
fix folding
2007-06-05 13:35:26 +00:00
/* }}} */
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds length of initial segment consisting entirely of characters found in mask. If start or/and length is provided works like strspn(substr($s,$start,$len),$good_chars) */
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
PHP_FUNCTION(strspn)
{
php_spn_common_handler(INTERNAL_FUNCTION_PARAM_PASSTHRU, STR_STRSPN);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds length of initial segment consisting entirely of characters not found in mask. If start or/and length is provide works like strcspn(substr($s,$start,$len),$bad_chars) */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strcspn)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Added common handler for strspn() and strcspn(). Almost of the code is identical. Modified to accept up to 4 params. Last 2 optional start & len. They are analogic to start & len of substr(). So the behavior when start & len are used is like strspn(substr($s,$start,$len),$good_chars) and strcspn(substr($s,$start,$len), $bad_chars)
2002-09-25 19:06:29 +00:00
php_spn_common_handler(INTERNAL_FUNCTION_PARAM_PASSTHRU, STR_STRCSPN);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
/* {{{ PHP_MINIT_FUNCTION(nl_langinfo) */
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
#if HAVE_NL_LANGINFO
PHP_MINIT_FUNCTION(nl_langinfo)
{
master renames phase 1
2014-08-25 17:24:55 +00:00
#define REGISTER_NL_LANGINFO_CONSTANT(x) REGISTER_LONG_CONSTANT(#x, x, CONST_CS | CONST_PERSISTENT)
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#ifdef ABDAY_1
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_1);
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_2);
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_3);
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_4);
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_5);
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_6);
REGISTER_NL_LANGINFO_CONSTANT(ABDAY_7);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef DAY_1
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(DAY_1);
REGISTER_NL_LANGINFO_CONSTANT(DAY_2);
REGISTER_NL_LANGINFO_CONSTANT(DAY_3);
REGISTER_NL_LANGINFO_CONSTANT(DAY_4);
REGISTER_NL_LANGINFO_CONSTANT(DAY_5);
REGISTER_NL_LANGINFO_CONSTANT(DAY_6);
REGISTER_NL_LANGINFO_CONSTANT(DAY_7);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef ABMON_1
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ABMON_1);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_2);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_3);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_4);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_5);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_6);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_7);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_8);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_9);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_10);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_11);
REGISTER_NL_LANGINFO_CONSTANT(ABMON_12);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef MON_1
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(MON_1);
REGISTER_NL_LANGINFO_CONSTANT(MON_2);
REGISTER_NL_LANGINFO_CONSTANT(MON_3);
REGISTER_NL_LANGINFO_CONSTANT(MON_4);
REGISTER_NL_LANGINFO_CONSTANT(MON_5);
REGISTER_NL_LANGINFO_CONSTANT(MON_6);
REGISTER_NL_LANGINFO_CONSTANT(MON_7);
REGISTER_NL_LANGINFO_CONSTANT(MON_8);
REGISTER_NL_LANGINFO_CONSTANT(MON_9);
REGISTER_NL_LANGINFO_CONSTANT(MON_10);
REGISTER_NL_LANGINFO_CONSTANT(MON_11);
REGISTER_NL_LANGINFO_CONSTANT(MON_12);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef AM_STR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(AM_STR);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef PM_STR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(PM_STR);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef D_T_FMT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(D_T_FMT);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef D_FMT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(D_FMT);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef T_FMT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(T_FMT);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef T_FMT_AMPM
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(T_FMT_AMPM);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef ERA
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ERA);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#ifdef ERA_YEAR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ERA_YEAR);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#ifdef ERA_D_T_FMT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ERA_D_T_FMT);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef ERA_D_FMT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ERA_D_FMT);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef ERA_T_FMT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ERA_T_FMT);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
#ifdef ALT_DIGITS
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(ALT_DIGITS);
Fix some build probs on *bsd
2001-08-10 18:02:59 +00:00
#endif
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#ifdef INT_CURR_SYMBOL
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(INT_CURR_SYMBOL);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
#ifdef CURRENCY_SYMBOL
REGISTER_NL_LANGINFO_CONSTANT(CURRENCY_SYMBOL);
#endif
#ifdef CRNCYSTR
REGISTER_NL_LANGINFO_CONSTANT(CRNCYSTR);
#endif
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#ifdef MON_DECIMAL_POINT
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(MON_DECIMAL_POINT);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef MON_THOUSANDS_SEP
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(MON_THOUSANDS_SEP);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef MON_GROUPING
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(MON_GROUPING);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef POSITIVE_SIGN
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(POSITIVE_SIGN);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef NEGATIVE_SIGN
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(NEGATIVE_SIGN);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef INT_FRAC_DIGITS
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(INT_FRAC_DIGITS);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef FRAC_DIGITS
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(FRAC_DIGITS);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef P_CS_PRECEDES
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(P_CS_PRECEDES);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef P_SEP_BY_SPACE
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(P_SEP_BY_SPACE);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef N_CS_PRECEDES
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(N_CS_PRECEDES);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef N_SEP_BY_SPACE
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(N_SEP_BY_SPACE);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef P_SIGN_POSN
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(P_SIGN_POSN);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
#ifdef N_SIGN_POSN
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(N_SIGN_POSN);
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
#endif
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
#ifdef DECIMAL_POINT
REGISTER_NL_LANGINFO_CONSTANT(DECIMAL_POINT);
- HEAD and PHP_5_2 have it like this, PHP_5_3 should not be different!
2009-07-27 09:23:06 +00:00
#endif
#ifdef RADIXCHAR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(RADIXCHAR);
#endif
#ifdef THOUSANDS_SEP
REGISTER_NL_LANGINFO_CONSTANT(THOUSANDS_SEP);
- HEAD and PHP_5_2 have it like this, PHP_5_3 should not be different!
2009-07-27 09:23:06 +00:00
#endif
#ifdef THOUSEP
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(THOUSEP);
#endif
More ifdefs, just in case.
2001-07-05 10:53:15 +00:00
#ifdef GROUPING
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(GROUPING);
More ifdefs, just in case.
2001-07-05 10:53:15 +00:00
#endif
#ifdef YESEXPR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(YESEXPR);
More ifdefs, just in case.
2001-07-05 10:53:15 +00:00
#endif
#ifdef NOEXPR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(NOEXPR);
More ifdefs, just in case.
2001-07-05 10:53:15 +00:00
#endif
#ifdef YESSTR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(YESSTR);
More ifdefs, just in case.
2001-07-05 10:53:15 +00:00
#endif
#ifdef NOSTR
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
REGISTER_NL_LANGINFO_CONSTANT(NOSTR);
More ifdefs, just in case.
2001-07-05 10:53:15 +00:00
#endif
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
#ifdef CODESET
REGISTER_NL_LANGINFO_CONSTANT(CODESET);
#endif
#undef REGISTER_NL_LANGINFO_CONSTANT
return SUCCESS;
}
Added loads of #ifdefs; should fix Bug #11889.
2001-07-05 10:49:58 +00:00
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Query language and locale information */
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
PHP_FUNCTION(nl_langinfo)
{
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long item;
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
char *value;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_LONG(item)
ZEND_PARSE_PARAMETERS_END();
MFH: add workaround for glibc issue
2007-12-13 22:29:48 +00:00
switch(item) { /* {{{ */
#ifdef ABDAY_1
case ABDAY_1:
case ABDAY_2:
case ABDAY_3:
case ABDAY_4:
case ABDAY_5:
case ABDAY_6:
case ABDAY_7:
#endif
#ifdef DAY_1
case DAY_1:
case DAY_2:
case DAY_3:
case DAY_4:
case DAY_5:
case DAY_6:
case DAY_7:
#endif
#ifdef ABMON_1
case ABMON_1:
case ABMON_2:
case ABMON_3:
case ABMON_4:
case ABMON_5:
case ABMON_6:
case ABMON_7:
case ABMON_8:
case ABMON_9:
case ABMON_10:
case ABMON_11:
case ABMON_12:
#endif
#ifdef MON_1
case MON_1:
case MON_2:
case MON_3:
case MON_4:
case MON_5:
case MON_6:
case MON_7:
case MON_8:
case MON_9:
case MON_10:
case MON_11:
case MON_12:
#endif
#ifdef AM_STR
case AM_STR:
#endif
#ifdef PM_STR
case PM_STR:
#endif
#ifdef D_T_FMT
case D_T_FMT:
#endif
#ifdef D_FMT
case D_FMT:
#endif
#ifdef T_FMT
case T_FMT:
#endif
#ifdef T_FMT_AMPM
case T_FMT_AMPM:
#endif
#ifdef ERA
case ERA:
#endif
#ifdef ERA_YEAR
case ERA_YEAR:
#endif
#ifdef ERA_D_T_FMT
case ERA_D_T_FMT:
#endif
#ifdef ERA_D_FMT
case ERA_D_FMT:
#endif
#ifdef ERA_T_FMT
case ERA_T_FMT:
#endif
#ifdef ALT_DIGITS
case ALT_DIGITS:
#endif
#ifdef INT_CURR_SYMBOL
case INT_CURR_SYMBOL:
#endif
#ifdef CURRENCY_SYMBOL
case CURRENCY_SYMBOL:
#endif
#ifdef CRNCYSTR
case CRNCYSTR:
#endif
#ifdef MON_DECIMAL_POINT
case MON_DECIMAL_POINT:
#endif
#ifdef MON_THOUSANDS_SEP
case MON_THOUSANDS_SEP:
#endif
#ifdef MON_GROUPING
case MON_GROUPING:
#endif
#ifdef POSITIVE_SIGN
case POSITIVE_SIGN:
#endif
#ifdef NEGATIVE_SIGN
case NEGATIVE_SIGN:
#endif
#ifdef INT_FRAC_DIGITS
case INT_FRAC_DIGITS:
#endif
#ifdef FRAC_DIGITS
case FRAC_DIGITS:
#endif
#ifdef P_CS_PRECEDES
case P_CS_PRECEDES:
#endif
#ifdef P_SEP_BY_SPACE
case P_SEP_BY_SPACE:
#endif
#ifdef N_CS_PRECEDES
case N_CS_PRECEDES:
#endif
#ifdef N_SEP_BY_SPACE
case N_SEP_BY_SPACE:
#endif
#ifdef P_SIGN_POSN
case P_SIGN_POSN:
#endif
#ifdef N_SIGN_POSN
case N_SIGN_POSN:
#endif
#ifdef DECIMAL_POINT
case DECIMAL_POINT:
Fixed bug #48872 (string.c: errors: duplicate case values) (Only in PHP_5_3)
2009-07-27 07:14:40 +00:00
#elif defined(RADIXCHAR)
MFH: add workaround for glibc issue
2007-12-13 22:29:48 +00:00
case RADIXCHAR:
#endif
#ifdef THOUSANDS_SEP
case THOUSANDS_SEP:
Fixed bug #48872 (string.c: errors: duplicate case values) (Only in PHP_5_3)
2009-07-27 07:14:40 +00:00
#elif defined(THOUSEP)
MFH: add workaround for glibc issue
2007-12-13 22:29:48 +00:00
case THOUSEP:
#endif
#ifdef GROUPING
case GROUPING:
#endif
#ifdef YESEXPR
case YESEXPR:
#endif
#ifdef NOEXPR
case NOEXPR:
#endif
#ifdef YESSTR
case YESSTR:
#endif
#ifdef NOSTR
case NOSTR:
#endif
#ifdef CODESET
case CODESET:
#endif
break;
default:
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Item '" ZEND_LONG_FMT "' is not valid", item);
MFH: add workaround for glibc issue
2007-12-13 22:29:48 +00:00
RETURN_FALSE;
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
}
MFH: add workaround for glibc issue
2007-12-13 22:29:48 +00:00
/* }}} */
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
MFH: add workaround for glibc issue
2007-12-13 22:29:48 +00:00
value = nl_langinfo(item);
Ws fix. Forgot to add this in the previous commit.
2002-10-02 18:13:56 +00:00
if (value == NULL) {
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
RETURN_FALSE;
Ws fix. Forgot to add this in the previous commit.
2002-10-02 18:13:56 +00:00
} else {
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
RETURN_STRING(value);
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
}
}
#endif
Fix for #17271 @ Fix crash bug in stripslashes() when working in sybase mode (Rasmus)
2002-05-16 14:41:15 +00:00
/* }}} */
(PHP nl_langinfo) Added function when provided by OS (PHP htmlentities, htmlspecialchars) Uses nl_langinfo to determine charset @- Added nl_langinfo() (when OS provides it) that returns locale information. (Wez Furlong) # There are a lot of constants used by nl_langinfo; should we do something # along the lines of what we do for syslog?
2001-07-04 10:10:30 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Compares two strings using the current locale */
Added string comparison function strcoll(). It uses the current locale to do the comparisons. @- Added localeconv() and strcoll() functions for localization. (Sean)
2001-01-17 01:10:50 +00:00
PHP_FUNCTION(strcoll)
{
rename zval_get_long and fixes to some string functions
2014-08-18 06:39:39 +00:00
zend_string *s1, *s2;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(s1)
Z_PARAM_STR(s2)
ZEND_PARSE_PARAMETERS_END();
Added string comparison function strcoll(). It uses the current locale to do the comparisons. @- Added localeconv() and strcoll() functions for localization. (Sean)
2001-01-17 01:10:50 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(strcoll((const char *) ZSTR_VAL(s1),
(const char *) ZSTR_VAL(s2)));
Added string comparison function strcoll(). It uses the current locale to do the comparisons. @- Added localeconv() and strcoll() functions for localization. (Sean)
2001-01-17 01:10:50 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Added string comparison function strcoll(). It uses the current locale to do the comparisons. @- Added localeconv() and strcoll() functions for localization. (Sean)
2001-01-17 01:10:50 +00:00
Un-revert patch
2001-08-03 23:09:05 +00:00
/* {{{ php_charmask
* Fills a 256-byte bytemask with input. You can specify a range like 'a..z',
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
* it needs to be incrementing.
MFH: fix typo
2006-01-05 20:49:37 +00:00
* Returns: FAILURE/SUCCESS whether the input was correct (i.e. no range errors)
Un-revert patch
2001-08-03 23:09:05 +00:00
*/
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
static inline int php_charmask(const unsigned char *input, size_t len, char *mask)
Un-revert patch
2001-08-03 23:09:05 +00:00
{
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
const unsigned char *end;
Un-revert patch
2001-08-03 23:09:05 +00:00
unsigned char c;
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
int result = SUCCESS;
Un-revert patch
2001-08-03 23:09:05 +00:00
memset(mask, 0, 256);
ws fixes.
2002-10-02 17:56:04 +00:00
for (end = input+len; input < end; input++) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
c=*input;
if ((input+3 < end) && input[1] == '.' && input[2] == '.'
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
&& input[3] >= c) {
memset(mask+c, 1, input[3] - c + 1);
Un-revert patch
2001-08-03 23:09:05 +00:00
input+=3;
ws fixes.
2002-10-02 17:56:04 +00:00
} else if ((input+1 < end) && input[0] == '.' && input[1] == '.') {
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
/* Error, try to be as helpful as possible:
(a range ending/starting with '.' won't be captured here) */
ws fixes.
2002-10-02 17:56:04 +00:00
if (end-len >= input) { /* there was no 'left' char */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Invalid '..'-range, no character to the left of '..'");
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
result = FAILURE;
continue;
}
ws fixes.
2002-10-02 17:56:04 +00:00
if (input+2 >= end) { /* there is no 'right' char */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Invalid '..'-range, no character to the right of '..'");
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
result = FAILURE;
continue;
}
if (input[-1] > input[2]) { /* wrong order */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Invalid '..'-range, '..'-range needs to be incrementing");
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
result = FAILURE;
continue;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
/* FIXME: better error (a..b..c is the only left possibility?) */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_error_docref(NULL, E_WARNING, "Invalid '..'-range");
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
result = FAILURE;
continue;
} else {
Un-revert patch
2001-08-03 23:09:05 +00:00
mask[c]=1;
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
}
Un-revert patch
2001-08-03 23:09:05 +00:00
}
Error-handling for the second parameter of [l|r]trim
2001-08-05 19:38:49 +00:00
return result;
Un-revert patch
2001-08-03 23:09:05 +00:00
}
/* }}} */
trim() micro-optimization
2017-05-30 14:14:08 +00:00
/* {{{ php_trim_int()
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
* mode 1 : trim left
* mode 2 : trim right
* mode 3 : trim left and right
* what indicates which chars are to be trimmed. NULL->default (' \t\n\r\v\0')
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
*/
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
static zend_always_inline zend_string *php_trim_int(zend_string *str, const char *what, size_t what_len, int mode)
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
{
trim() micro-optimization
2017-05-30 14:14:08 +00:00
const char *start = ZSTR_VAL(str);
const char *end = start + ZSTR_LEN(str);
Un-revert patch
2001-08-03 23:09:05 +00:00
char mask[256];
if (what) {
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
if (what_len == 1) {
Micro optimization
2015-06-24 13:35:13 +00:00
char p = *what;
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
if (mode & 1) {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
while (start != end) {
if (*start == p) {
start++;
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
} else {
break;
}
Improved trim/ltrim/rtrim functions. Now php_trim() operates on zend_string and returns zend_string (it may return the same zend_string).
2014-12-23 01:29:41 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
if (mode & 2) {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
while (start != end) {
if (*(end-1) == p) {
end--;
} else {
break;
}
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
}
}
} else {
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
php_charmask((const unsigned char *) what, what_len, mask);
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
if (mode & 1) {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
while (start != end) {
if (mask[(unsigned char)*start]) {
start++;
Improved trim/ltrim/rtrim functions. Now php_trim() operates on zend_string and returns zend_string (it may return the same zend_string).
2014-12-23 01:29:41 +00:00
} else {
break;
}
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
}
}
if (mode & 2) {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
while (start != end) {
if (mask[(unsigned char)*(end-1)]) {
end--;
} else {
break;
}
Optimized trim for what_len == 1
2015-01-19 12:45:30 +00:00
}
Improved trim/ltrim/rtrim functions. Now php_trim() operates on zend_string and returns zend_string (it may return the same zend_string).
2014-12-23 01:29:41 +00:00
}
}
} else {
if (mode & 1) {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
while (start != end) {
unsigned char c = (unsigned char)*start;
if (c <= ' ' &&
(c == ' ' || c == '\n' || c == '\r' || c == '\t' || c == '\v' || c == '\0')) {
start++;
fix signed/unsigned mismatch
2014-09-16 20:15:27 +00:00
} else {
break;
}
Improved trim/ltrim/rtrim functions. Now php_trim() operates on zend_string and returns zend_string (it may return the same zend_string).
2014-12-23 01:29:41 +00:00
}
}
if (mode & 2) {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
while (start != end) {
unsigned char c = (unsigned char)*(end-1);
if (c <= ' ' &&
(c == ' ' || c == '\n' || c == '\r' || c == '\t' || c == '\v' || c == '\0')) {
end--;
} else {
break;
}
Improved trim/ltrim/rtrim functions. Now php_trim() operates on zend_string and returns zend_string (it may return the same zend_string).
2014-12-23 01:29:41 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
}
trim() micro-optimization
2017-05-30 14:14:08 +00:00
if (ZSTR_LEN(str) == end - start) {
Improved trim/ltrim/rtrim functions. Now php_trim() operates on zend_string and returns zend_string (it may return the same zend_string).
2014-12-23 01:29:41 +00:00
return zend_string_copy(str);
trim() micro-optimization
2017-05-30 14:14:08 +00:00
} else if (end - start == 0) {
return ZSTR_EMPTY_ALLOC();
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
} else {
trim() micro-optimization
2017-05-30 14:14:08 +00:00
return zend_string_init(start, end - start, 0);
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
}
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
/* }}} */
trim() micro-optimization
2017-05-30 14:14:08 +00:00
/* {{{ php_trim_int()
* mode 1 : trim left
* mode 2 : trim right
* mode 3 : trim left and right
* what indicates which chars are to be trimmed. NULL->default (' \t\n\r\v\0')
*/
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
PHPAPI zend_string *php_trim(zend_string *str, const char *what, size_t what_len, int mode)
trim() micro-optimization
2017-05-30 14:14:08 +00:00
{
return php_trim_int(str, what, what_len, mode);
}
/* }}} */
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
/* {{{ php_do_trim
* Base for trim(), rtrim() and ltrim() functions.
*/
trim() micro-optimization
2017-05-30 14:14:08 +00:00
static zend_always_inline void php_do_trim(INTERNAL_FUNCTION_PARAMETERS, int mode)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str;
zend_string *what = NULL;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
Z_PARAM_STR(str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
Z_PARAM_STR(what)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
trim() micro-optimization
2017-05-30 14:14:08 +00:00
ZVAL_STR(return_value, php_trim_int(str, (what ? ZSTR_VAL(what) : NULL), (what ? ZSTR_LEN(what) : 0), mode));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Strips whitespace from the beginning and end of a string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(trim)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
php_do_trim(INTERNAL_FUNCTION_PARAM_PASSTHRU, 3);
}
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Removes trailing whitespace */
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
PHP_FUNCTION(rtrim)
{
php_do_trim(INTERNAL_FUNCTION_PARAM_PASSTHRU, 2);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Strips whitespace from the beginning of a string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(ltrim)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
- Made the trim functions use same base function. (less code :) - Renamed php_trim2 -> php_trim (not used anyplace else, yet) - Made php_trim abit more usable for external use. (not necessary to use zvals with it anymore)
2002-07-13 04:13:43 +00:00
php_do_trim(INTERNAL_FUNCTION_PARAM_PASSTHRU, 1);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Wraps buffer to selected number of characters using string break char */
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
PHP_FUNCTION(wordwrap)
{
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
zend_string *text;
char *breakchar = "\n";
size_t newtextlen, chk, breakchar_len = 1;
Fixed overflow on 64bit systems in str_repeat() and wordwrap(). Disabled CURLOPT_FOLLOWLOCATION in curl when open_basedir or safe_mode are enabled. # Patches by Stefan E.
2006-08-10 14:40:13 +00:00
size_t alloced;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long current = 0, laststart = 0, lastspace = 0;
zend_long linelength = 75;
Fixed segfault in wordwrap() when wrapping to zero width and using multi-character break or trying to force cut (bug #12768, now fails and issues a warning because forcing a zero-width cut doesn't make sense). Also converted to new paramater-passing API and avoid making an extra copy of the return values. # also added tests. @- Fixed segfault in wordwrap() when wrapping to zero width and using @ multi-character break or trying to force cut (bug #12768). (Jim)
2002-01-04 19:48:08 +00:00
zend_bool docut = 0;
Fixed memory leak in ext/standard/tests/strings/bug20927.phpt
2014-02-23 10:17:39 +00:00
zend_string *newtext;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 4)
Z_PARAM_STR(text)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(linelength)
Z_PARAM_STRING(breakchar, breakchar_len)
Z_PARAM_BOOL(docut)
ZEND_PARSE_PARAMETERS_END();
fixes to string functions
2014-08-18 13:39:38 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(text) == 0) {
Fixed bug #29034 (wordwrap() returns a boolean when passed empty string).
2004-07-06 22:16:42 +00:00
RETURN_EMPTY_STRING();
}
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
if (breakchar_len == 0) {
Improve a last couple of argument error messages Closes GH-5404
2020-04-16 22:09:15 +00:00
zend_argument_value_error(3, "cannot be empty");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Fixed "Floating point exception" inside wordwrap()
2007-07-22 15:55:16 +00:00
}
Fixed segfault in wordwrap() when wrapping to zero width and using multi-character break or trying to force cut (bug #12768, now fails and issues a warning because forcing a zero-width cut doesn't make sense). Also converted to new paramater-passing API and avoid making an extra copy of the return values. # also added tests. @- Fixed segfault in wordwrap() when wrapping to zero width and using @ multi-character break or trying to force cut (bug #12768). (Jim)
2002-01-04 19:48:08 +00:00
if (linelength == 0 && docut) {
Improve a last couple of argument error messages Closes GH-5404
2020-04-16 22:09:15 +00:00
zend_argument_value_error(4, "cannot be true when argument #2 ($width) is 0");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
- Fix for bug #6673 - Added a "cut" option to wordwrap (as per feature request #6429) @ Added an optional parameter to wordwrap that cuts a string if the length of a word is longer than the maximum allowed with (Derick)
2000-09-11 20:06:24 +00:00
}
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
/* Special case for a single-character break as it needs no
additional storage space */
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
if (breakchar_len == 1 && !docut) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
newtext = zend_string_init(ZSTR_VAL(text), ZSTR_LEN(text), 0);
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
laststart = lastspace = 0;
Fixed compilation warnings
2016-06-23 09:47:06 +00:00
for (current = 0; current < (zend_long)ZSTR_LEN(text); current++) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_VAL(text)[current] == breakchar[0]) {
Fixed bug #49361 (wordwrap() wraps incorrectly on end of line boundaries).
2009-08-31 12:28:46 +00:00
laststart = lastspace = current + 1;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
} else if (ZSTR_VAL(text)[current] == ' ') {
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
if (current - laststart >= linelength) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(newtext)[current] = breakchar[0];
Fixed bug #28386 (wordwrap() wraps lines 1 character too soon).
2004-05-13 17:44:23 +00:00
laststart = current + 1;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
lastspace = current;
ws fixes
2002-10-10 18:41:24 +00:00
} else if (current - laststart >= linelength && laststart != lastspace) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(newtext)[lastspace] = breakchar[0];
MFH: patch for #37846 (wordwrap() wraps incorrectly) by Dmitry Kononov <ddk at krasn dot ru>
2006-07-28 12:23:18 +00:00
laststart = lastspace + 1;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(newtext);
ws fixes
2002-10-10 18:41:24 +00:00
} else {
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
/* Multiple character line break or forced cut */
if (linelength > 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
chk = (size_t)(ZSTR_LEN(text)/linelength + 1);
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
newtext = zend_string_safe_alloc(chk, breakchar_len, ZSTR_LEN(text), 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
alloced = ZSTR_LEN(text) + chk * breakchar_len + 1;
ws fixes
2002-10-10 18:41:24 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
chk = ZSTR_LEN(text);
alloced = ZSTR_LEN(text) * (breakchar_len + 1) + 1;
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
newtext = zend_string_safe_alloc(ZSTR_LEN(text), breakchar_len + 1, 0, 0);
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
}
Fixed bug #20927 and updated the test case for the bug.
2002-12-12 13:38:19 +00:00
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
/* now keep track of the actual new text length */
newtextlen = 0;
laststart = lastspace = 0;
Fixed compilation warnings
2016-06-23 09:47:06 +00:00
for (current = 0; current < (zend_long)ZSTR_LEN(text); current++) {
Fix unsigned comparisons and remove dead code Fix unsigned comparison Cleanup never executed block Fix unsigned comparison Fix unsigned comparison, diff can't be < 0 Fix unsigned comparison Fix unsigned comparison Remove dead code
2017-11-17 20:35:22 +00:00
if (chk == 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
alloced += (size_t) (((ZSTR_LEN(text) - current + 1)/linelength + 1) * breakchar_len) + 1;
Optimize zend_string_realloc() add more specialized versions zend_string_extend() and zend_string_truncate()
2015-03-19 23:02:42 +00:00
newtext = zend_string_extend(newtext, alloced, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
chk = (size_t) ((ZSTR_LEN(text) - current)/linelength) + 1;
Fixed bug #20927 and updated the test case for the bug.
2002-12-12 13:38:19 +00:00
}
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
/* when we hit an existing break, copy to new buffer, and
* fix up laststart and lastspace */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_VAL(text)[current] == breakchar[0]
&& current + breakchar_len < ZSTR_LEN(text)
&& !strncmp(ZSTR_VAL(text) + current, breakchar, breakchar_len)) {
memcpy(ZSTR_VAL(newtext) + newtextlen, ZSTR_VAL(text) + laststart, current - laststart + breakchar_len);
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
newtextlen += current - laststart + breakchar_len;
current += breakchar_len - 1;
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
laststart = lastspace = current + 1;
Fixed bug #20927 and updated the test case for the bug.
2002-12-12 13:38:19 +00:00
chk--;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
/* if it is a space, check if it is at the line boundary,
* copy and insert a break, or just keep track of it */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
else if (ZSTR_VAL(text)[current] == ' ') {
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
if (current - laststart >= linelength) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, ZSTR_VAL(text) + laststart, current - laststart);
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
newtextlen += current - laststart;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, breakchar, breakchar_len);
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
newtextlen += breakchar_len;
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
laststart = current + 1;
Fixed bug #20927 and updated the test case for the bug.
2002-12-12 13:38:19 +00:00
chk--;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
lastspace = current;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
/* if we are cutting, and we've accumulated enough
More tweaking of wordwrap() with the cut parameter set. It was being a little too aggressive and cutting words without breaking at spaces first. (A couple of tests were incorrect.)
2002-01-05 23:49:58 +00:00
* characters, and we haven't see a space for this line,
* copy and insert a break. */
else if (current - laststart >= linelength
&& docut && laststart >= lastspace) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, ZSTR_VAL(text) + laststart, current - laststart);
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
newtextlen += current - laststart;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, breakchar, breakchar_len);
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
newtextlen += breakchar_len;
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
laststart = lastspace = current;
Fixed bug #20927 and updated the test case for the bug.
2002-12-12 13:38:19 +00:00
chk--;
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
}
/* if the current word puts us over the linelength, copy
* back up until the last space, insert a break, and move
* up the laststart */
else if (current - laststart >= linelength
&& laststart < lastspace) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, ZSTR_VAL(text) + laststart, lastspace - laststart);
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
newtextlen += lastspace - laststart;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, breakchar, breakchar_len);
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
newtextlen += breakchar_len;
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
laststart = lastspace = lastspace + 1;
Fixed bug #20927 and updated the test case for the bug.
2002-12-12 13:38:19 +00:00
chk--;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
}
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
/* copy over any stragglers */
if (laststart != current) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(newtext) + newtextlen, ZSTR_VAL(text) + laststart, current - laststart);
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
newtextlen += current - laststart;
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(newtext)[newtextlen] = '\0';
Fixed a buffer overflow that occurs when wordwrap is unable to calculate the correct number of times the multi-byte break needs to be inserted into the string.
2002-09-05 14:00:28 +00:00
/* free unused memory */
Optimize zend_string_realloc() add more specialized versions zend_string_extend() and zend_string_truncate()
2015-03-19 23:02:42 +00:00
newtext = zend_string_truncate(newtext, newtextlen, 0);
New memcpy()-based wordwrap() implementation. The simple case (single-character break, no forced break) appears to be about 60% faster, and there's simply no comparison for non-simple cases with non-trivial amounts of text. The old algorithm was O(n^2) (with an unfortunately large constant factor) because of the use of strncat(), the new one is O(n). Added some more tests, too. @ - Made wordwrap() significantly faster. (Jim) # test case: $t = join('',file('ChangeLog')); $w = wordwrap($t,10,"\n",1); # new code completes in less than a second. i'm still waiting for the # old code to finish.
2002-01-05 20:46:43 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(newtext);
wordwrap function from Chris Russel <russel@yorku.ca> differences from his patch: - wordwrap width and wrap-string now optional parameters (default to 75 and "\n" respectively) - wordwrap_byte is now just an automatic special case of wordwrap - Zend API compliant @- Added new function "wordwrap" to wordwrap long strings from Chris @ Russel <russel@yorku.ca> (David Croft)
2000-07-22 01:12:24 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_explode */
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
PHPAPI void php_explode(const zend_string *delim, zend_string *str, zval *return_value, zend_long limit)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
const char *p1 = ZSTR_VAL(str);
const char *endp = ZSTR_VAL(str) + ZSTR_LEN(str);
const char *p2 = php_memnstr(ZSTR_VAL(str), ZSTR_VAL(delim), ZSTR_LEN(delim), endp);
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
zval tmp;
@- explode() is now binary-safe. (Thies) we save one buffer-copy as well (hehe)
1999-12-06 18:29:22 +00:00
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (p2 == NULL) {
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
ZVAL_STR_COPY(&tmp, str);
zend_hash_next_index_insert_new(Z_ARRVAL_P(return_value), &tmp);
(explode) fixed newly introduced bug (them shall check twice *before* committing!)
1999-12-07 08:14:00 +00:00
} else {
do {
Add helper APIs for maybe-interned string creation Add ZVAL_CHAR/RETVAL_CHAR/RETURN_CHAR as a shortcut for using ZVAL_INTERNED_STRING and ZSTR_CHAR. Add zend_string_init_fast() as a helper for the empty string / one char interned string / zend_string_init() pattern. Also add corresponding ZVAL_STRINGL_FAST etc macros. Closes GH-5684.
2020-06-08 10:45:01 +00:00
ZVAL_STRINGL_FAST(&tmp, p1, p2 - p1);
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
zend_hash_next_index_insert_new(Z_ARRVAL_P(return_value), &tmp);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p1 = p2 + ZSTR_LEN(delim);
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
p2 = php_memnstr(p1, ZSTR_VAL(delim), ZSTR_LEN(delim), endp);
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
} while (p2 != NULL && --limit > 1);
(explode) fixed newly introduced bug (them shall check twice *before* committing!)
1999-12-07 08:14:00 +00:00
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
if (p1 <= endp) {
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
ZVAL_STRINGL(&tmp, p1, endp - p1);
zend_hash_next_index_insert_new(Z_ARRVAL_P(return_value), &tmp);
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_explode_negative_limit */
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
PHPAPI void php_explode_negative_limit(const zend_string *delim, zend_string *str, zval *return_value, zend_long limit)
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
{
MFH: explode() stuff: - Fixed bug #47560 (explode()'s limit parameter odd behaviour) by reverting change for bug #47546 - Changed int to long where needed (should fix memory errors from overflow seen in bug #47854) - Simplified logic a bit with limit and its default value - php_explode_negative_limit(): removed safe_emalloc (not needed; plain erealloc is used later) - Moved declarations/allocation to optimize if the delimiter isn't found - Changed ALLOC_STEP size for less realloc's (and maybe better memory block alignment?)
2009-04-01 17:05:36 +00:00
#define EXPLODE_ALLOC_STEP 64
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
const char *p1 = ZSTR_VAL(str);
const char *endp = ZSTR_VAL(str) + ZSTR_LEN(str);
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *p2 = php_memnstr(ZSTR_VAL(str), ZSTR_VAL(delim), ZSTR_LEN(delim), endp);
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
zval tmp;
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
if (p2 == NULL) {
/*
do nothing since limit <= -1, thus if only one chunk - 1 + (limit) <= 0
by doing nothing we return empty array
*/
} else {
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t allocated = EXPLODE_ALLOC_STEP, found = 0;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long i, to_return;
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
const char **positions = emalloc(allocated * sizeof(char *));
MFH: explode() stuff: - Fixed bug #47560 (explode()'s limit parameter odd behaviour) by reverting change for bug #47546 - Changed int to long where needed (should fix memory errors from overflow seen in bug #47854) - Simplified logic a bit with limit and its default value - php_explode_negative_limit(): removed safe_emalloc (not needed; plain erealloc is used later) - Moved declarations/allocation to optimize if the delimiter isn't found - Changed ALLOC_STEP size for less realloc's (and maybe better memory block alignment?)
2009-04-01 17:05:36 +00:00
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
positions[found++] = p1;
do {
if (found >= allocated) {
allocated = found + EXPLODE_ALLOC_STEP;/* make sure we have enough memory */
Fix MSVC level 1 (severe) warnings We fix (hopefully) all instances of: * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4005> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4024> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4028> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4047> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4087> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4090> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4273> * <https://docs.microsoft.com/en-us/cpp/error-messages/compiler-warnings/compiler-warning-level-1-c4312> `zend_llist_add_element()` and `zend_llist_prepend_element()` now explicitly expect a *const* pointer. We use the macro `ZEND_VOIDP()` instead of a `(void*)` cast to suppress C4090; this should prevent accidential removal of the cast by clarifying the intention, and makes it easier to remove the casts if the issue[1] will be resolved sometime. [1] <https://developercommunity.visualstudio.com/content/problem/390711/c-compiler-incorrect-propagation-of-const-qualifie.html>
2020-05-27 07:58:10 +00:00
positions = erealloc(ZEND_VOIDP(positions), allocated*sizeof(char *));
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
positions[found++] = p1 = p2 + ZSTR_LEN(delim);
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
p2 = php_memnstr(p1, ZSTR_VAL(delim), ZSTR_LEN(delim), endp);
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
} while (p2 != NULL);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
to_return = limit + found;
/* limit is at least -1 therefore no need of bounds checking : i will be always less than found */
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
for (i = 0; i < to_return; i++) { /* this checks also for to_return > 0 */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZVAL_STRINGL(&tmp, positions[i], (positions[i+1] - ZSTR_LEN(delim)) - positions[i]);
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
zend_hash_next_index_insert_new(Z_ARRVAL_P(return_value), &tmp);
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
}
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
efree((void *)positions);
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
}
#undef EXPLODE_ALLOC_STEP
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Splits a string on string separator and return array of components. If limit is positive only limit number of components is returned. If limit is negative all components except the last abs(limit) are returned. */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
PHP_FUNCTION(explode)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Fixed several memory leaks
2014-02-24 07:03:35 +00:00
zend_string *str, *delim;
master renames phase 3
2014-08-25 18:22:49 +00:00
zend_long limit = ZEND_LONG_MAX; /* No limit */
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
zval tmp;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Z_PARAM_STR(delim)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
master renames phase 1
2014-08-25 17:24:55 +00:00
Z_PARAM_LONG(limit)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(delim) == 0) {
Improve a last couple of argument error messages Closes GH-5404
2020-04-16 22:09:15 +00:00
zend_argument_value_error(1, "cannot be empty");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
@- explode() is now binary-safe. (Thies) we save one buffer-copy as well (hehe)
1999-12-06 18:29:22 +00:00
Remove pointless checks for array_init() return values.
2002-12-05 22:28:02 +00:00
array_init(return_value);
@- explode() is now binary-safe. (Thies) we save one buffer-copy as well (hehe)
1999-12-06 18:29:22 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(str) == 0) {
Promote warnings to errors in explode()
2019-08-22 10:09:55 +00:00
if (limit >= 0) {
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
ZVAL_EMPTY_STRING(&tmp);
zend_hash_index_add_new(Z_ARRVAL_P(return_value), 0, &tmp);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
Add support for negative values of limit of explode(). If limit is negative then all components are returned except the last abs(limit) ones.
2004-07-18 11:34:28 +00:00
return;
}
MFH: explode() stuff: - Fixed bug #47560 (explode()'s limit parameter odd behaviour) by reverting change for bug #47546 - Changed int to long where needed (should fix memory errors from overflow seen in bug #47854) - Simplified logic a bit with limit and its default value - php_explode_negative_limit(): removed safe_emalloc (not needed; plain erealloc is used later) - Moved declarations/allocation to optimize if the delimiter isn't found - Changed ALLOC_STEP size for less realloc's (and maybe better memory block alignment?)
2009-04-01 17:05:36 +00:00
if (limit > 1) {
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
php_explode(delim, str, return_value, limit);
MFH: explode() stuff: - Fixed bug #47560 (explode()'s limit parameter odd behaviour) by reverting change for bug #47546 - Changed int to long where needed (should fix memory errors from overflow seen in bug #47854) - Simplified logic a bit with limit and its default value - php_explode_negative_limit(): removed safe_emalloc (not needed; plain erealloc is used later) - Moved declarations/allocation to optimize if the delimiter isn't found - Changed ALLOC_STEP size for less realloc's (and maybe better memory block alignment?)
2009-04-01 17:05:36 +00:00
} else if (limit < 0) {
Directly use zend_string for php_implode/php_explode Also drop some unnecessary zval separations and copies
2014-10-07 14:43:50 +00:00
php_explode_negative_limit(delim, str, return_value, limit);
again (due to feature request in bug id #3251) added optional 3rd parameter 'limit' to explode() as in split()
2000-05-23 05:04:28 +00:00
} else {
e more efficient zend_hash_*() functions
2015-05-05 13:20:29 +00:00
ZVAL_STR_COPY(&tmp, str);
zend_hash_index_add_new(Z_ARRVAL_P(return_value), 0, &tmp);
again (due to feature request in bug id #3251) added optional 3rd parameter 'limit' to explode() as in split()
2000-05-23 05:04:28 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ An alias for implode */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
/* }}} */
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_implode */
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
PHPAPI void php_implode(const zend_string *glue, HashTable *pieces, zval *return_value)
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
{
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
zval *tmp;
Also unreserve T_CLASS
2015-05-25 20:58:30 +00:00
int numelems;
zend_string *str;
char *cptr;
size_t len = 0;
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
struct {
zend_string *str;
zend_long lval;
} *strings, *ptr;
ALLOCA_FLAG(use_heap)
have implode use the smart_str_*() functions, this should speed things up quite a bit...
2002-10-04 20:44:19 +00:00
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
numelems = zend_hash_num_elements(pieces);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
style & ws fixes
2003-02-08 15:26:17 +00:00
if (numelems == 0) {
Fixed segfault and memory leak in the new implementation of implode()
2002-10-06 11:28:11 +00:00
RETURN_EMPTY_STRING();
Also unreserve T_CLASS
2015-05-25 20:58:30 +00:00
} else if (numelems == 1) {
/* loop to search the first not undefined element... */
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
ZEND_HASH_FOREACH_VAL_IND(pieces, tmp) {
Also unreserve T_CLASS
2015-05-25 20:58:30 +00:00
RETURN_STR(zval_get_string(tmp));
} ZEND_HASH_FOREACH_END();
Fixed segfault and memory leak in the new implementation of implode()
2002-10-06 11:28:11 +00:00
}
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
ptr = strings = do_alloca((sizeof(*strings)) * numelems, use_heap);
Use zval_get_string in implode Also get rid of separate case for IS_OBJECT - this is already handled by the default branch.
2014-05-28 16:45:00 +00:00
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
ZEND_HASH_FOREACH_VAL_IND(pieces, tmp) {
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
if (EXPECTED(Z_TYPE_P(tmp) == IS_STRING)) {
ptr->str = Z_STR_P(tmp);
len += ZSTR_LEN(ptr->str);
ptr->lval = 0;
ptr++;
} else if (UNEXPECTED(Z_TYPE_P(tmp) == IS_LONG)) {
Fixed bug #72100 (implode() inserts garbage into resulting string when joins very big integer). (Mikhail Galanin)
2016-04-26 10:04:06 +00:00
zend_long val = Z_LVAL_P(tmp);
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
ptr->str = NULL;
ptr->lval = val;
ptr++;
Fixed bug #72100 (implode() inserts garbage into resulting string when joins very big integer). (Mikhail Galanin)
2016-04-26 10:04:06 +00:00
if (val <= 0) {
len++;
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
}
Fixed bug #72100 (implode() inserts garbage into resulting string when joins very big integer). (Mikhail Galanin)
2016-04-26 10:04:06 +00:00
while (val) {
val /= 10;
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
len++;
}
} else {
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
ptr->str = zval_get_string_func(tmp);
len += ZSTR_LEN(ptr->str);
ptr->lval = 1;
ptr++;
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
}
Also unreserve T_CLASS
2015-05-25 20:58:30 +00:00
} ZEND_HASH_FOREACH_END();
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
/* numelems can not be 0, we checked above */
Refactor: rename variables to match documentation Rename 'arr' to 'pieces' Rename 'delim' to 'glue'
2016-11-20 21:50:56 +00:00
str = zend_string_safe_alloc(numelems - 1, ZSTR_LEN(glue), len, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
cptr = ZSTR_VAL(str) + ZSTR_LEN(str);
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
*cptr = 0;
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
while (1) {
ptr--;
if (EXPECTED(ptr->str)) {
cptr -= ZSTR_LEN(ptr->str);
memcpy(cptr, ZSTR_VAL(ptr->str), ZSTR_LEN(ptr->str));
if (ptr->lval) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(ptr->str, 0);
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
}
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
} else {
char *oldPtr = cptr;
char oldVal = *cptr;
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
cptr = zend_print_long_to_buf(cptr, ptr->lval);
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
*oldPtr = oldVal;
}
Speed up implod()'ing between 5% and 30% (by time) Significantly speed up implode()'ing on integers now... Previously I accidentally committed the first part of the patch in 69b54ba926b714dff0f8b54bea385e9a278c5849
2015-05-25 23:12:23 +00:00
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
if (ptr == strings) {
break;
}
Refactor: rename variables to match documentation Rename 'arr' to 'pieces' Rename 'delim' to 'glue'
2016-11-20 21:50:56 +00:00
cptr -= ZSTR_LEN(glue);
memcpy(cptr, ZSTR_VAL(glue), ZSTR_LEN(glue));
Revert "Simplify php_implode(). Fill the resulting string in direct order." The patch slightly improved performance of string imploding, but reduced speed of integers imploding. This reverts commit 36d39b70443dcefeeb62d189ce67b414caf5486a.
2015-06-23 13:46:41 +00:00
}
Simplify php_implode(). Fill the resulting string in direct order.
2015-06-23 09:57:36 +00:00
Avoid overhead of memory allocation and useless referene-counting in implode()
2017-12-26 10:29:18 +00:00
free_alloca(strings, use_heap);
Also unreserve T_CLASS
2015-05-25 20:58:30 +00:00
RETURN_NEW_STR(str);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Joins array elements placing glue string between items and return one string */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
PHP_FUNCTION(implode)
{
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
zend_string *arg1_str = NULL;
HashTable *arg1_array = NULL;
zend_array *pieces = NULL;
@ Make the glue argument to implode() optional, if it is not specified @ default to using "". (Sterling)
2002-10-04 16:53:14 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Consolidate new union type ZPP macro names They will now follow the canonical order of types. Older macros are left intact due to maintaining BC. Closes GH-6112
2020-09-10 22:23:54 +00:00
Z_PARAM_ARRAY_HT_OR_STR(arg1_array, arg1_str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
Z_PARAM_ARRAY_HT_OR_NULL(pieces)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
if (pieces == NULL) {
if (arg1_array == NULL) {
Change argument error message format Closes GH-5211
2020-02-25 19:58:33 +00:00
zend_type_error("%s(): Argument #1 ($pieces) must be of type array, string given", get_active_function_name());
Use RETURN_THROWS() after zend_type_error()
2020-01-01 13:23:21 +00:00
RETURN_THROWS();
@ Make the glue argument to implode() optional, if it is not specified @ default to using "". (Sterling)
2002-10-04 16:53:14 +00:00
}
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
arg1_str = ZSTR_EMPTY_ALLOC();
pieces = arg1_array;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
} else {
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
if (arg1_str == NULL) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_type_error(1, "must be of type string, array given");
Use RETURN_THROWS() after zend_type_error()
2020-01-01 13:23:21 +00:00
RETURN_THROWS();
@ Make the glue argument to implode() optional, if it is not specified @ default to using "". (Sterling)
2002-10-04 16:53:14 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Remove support for mixing parameter order in implode()
2019-12-12 09:04:31 +00:00
php_implode(arg1_str, pieces, return_value);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
/* }}} */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
#define STRTOK_TABLE(p) BG(strtok_table)[(unsigned char) *p]
Un-revert patch
2001-08-03 23:09:05 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Tokenize a string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strtok)
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str, *tok = NULL;
Un-revert patch
2001-08-03 23:09:05 +00:00
char *token;
char *token_end;
char *p;
char *pe;
fixed trim() and strtok() to work with big strings
2014-08-28 19:59:00 +00:00
size_t skipped = 0;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
FAST ZPP for strtok() and nl2br() Looking at some template-driven code that calls these hundreds of times per request
2015-01-17 16:30:40 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
Z_PARAM_STR_OR_NULL(tok)
FAST ZPP for strtok() and nl2br() Looking at some template-driven code that calls these hundreds of times per request
2015-01-17 16:30:40 +00:00
ZEND_PARSE_PARAMETERS_END();
style & ws fixes
2003-02-08 15:26:17 +00:00
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
if (!tok) {
- New parameter parsing API (zend_get_parameters_array_ex--) - Fixed tests
2008-08-16 21:32:41 +00:00
tok = str;
} else {
Clean up strtok implementation Store the zend_string instead of performing a copy and storing in a zval. Also make sure the string is released immediately if it's no longer needed. Finally, avoid null pointer offset UB if no string has been set -- though I'm wondering if this case shouldn't be generating a warning?
2020-08-27 10:24:58 +00:00
if (BG(strtok_string)) {
zend_string_release(BG(strtok_string));
}
BG(strtok_string) = zend_string_copy(str);
BG(strtok_last) = ZSTR_VAL(str);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
BG(strtok_len) = ZSTR_LEN(str);
Replace our strtok implementation with a binary-safe one which code does not cause headaches when reading it. Also happens to be a bit quicker.
2001-08-03 08:49:19 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Clean up strtok implementation Store the zend_string instead of performing a copy and storing in a zval. Also make sure the string is released immediately if it's no longer needed. Finally, avoid null pointer offset UB if no string has been set -- though I'm wondering if this case shouldn't be generating a warning?
2020-08-27 10:24:58 +00:00
if (!BG(strtok_string)) {
/* String to tokenize not set. */
// TODO: Should this warn?
RETURN_FALSE;
}
Replace our strtok implementation with a binary-safe one which code does not cause headaches when reading it. Also happens to be a bit quicker.
2001-08-03 08:49:19 +00:00
Clean up strtok implementation Store the zend_string instead of performing a copy and storing in a zval. Also make sure the string is released immediately if it's no longer needed. Finally, avoid null pointer offset UB if no string has been set -- though I'm wondering if this case shouldn't be generating a warning?
2020-08-27 10:24:58 +00:00
p = BG(strtok_last); /* Where we start to search */
pe = ZSTR_VAL(BG(strtok_string)) + BG(strtok_len);
if (p >= pe) {
/* Reached the end of the string. */
Un-revert patch
2001-08-03 23:09:05 +00:00
RETURN_FALSE;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
token = ZSTR_VAL(tok);
token_end = token + ZSTR_LEN(tok);
Un-revert patch
2001-08-03 23:09:05 +00:00
style & ws fixes
2003-02-08 15:26:17 +00:00
while (token < token_end) {
Un-revert patch
2001-08-03 23:09:05 +00:00
STRTOK_TABLE(token++) = 1;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Un-revert patch
2001-08-03 23:09:05 +00:00
/* Skip leading delimiters */
ws fixes
2002-10-10 18:41:24 +00:00
while (STRTOK_TABLE(p)) {
Un-revert patch
2001-08-03 23:09:05 +00:00
if (++p >= pe) {
/* no other chars left */
Clean up strtok implementation Store the zend_string instead of performing a copy and storing in a zval. Also make sure the string is released immediately if it's no longer needed. Finally, avoid null pointer offset UB if no string has been set -- though I'm wondering if this case shouldn't be generating a warning?
2020-08-27 10:24:58 +00:00
goto return_false;
Reverting last patch, segfaults.
2001-08-03 22:42:02 +00:00
}
@- Fix behaviour of strtok. Bug 13866 (jmoore) # I have brought the behaviour of strtok into line with how the # libc strtok's behave. currently given # <string1><token><string2><token><token>string> # three recursive calls to strtok returns <string1>. <string2>, <token><string3> # it now returns <string1>, <string2>, <string3>. (there was some # debate in #php.bugs if it should return <string1>, <string2>, false, <string3> # but php's strtok now behaves the same way as the libc version.
2001-11-02 19:19:24 +00:00
skipped++;
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* We know at this place that *p is no delimiter, so skip it */
style & ws fixes
2003-02-08 15:26:17 +00:00
while (++p < pe) {
if (STRTOK_TABLE(p)) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
goto return_token;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Un-revert patch
2001-08-03 23:09:05 +00:00
if (p - BG(strtok_last)) {
return_token:
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
RETVAL_STRINGL(BG(strtok_last) + skipped, (p - BG(strtok_last)) - skipped);
Un-revert patch
2001-08-03 23:09:05 +00:00
BG(strtok_last) = p + 1;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
} else {
Clean up strtok implementation Store the zend_string instead of performing a copy and storing in a zval. Also make sure the string is released immediately if it's no longer needed. Finally, avoid null pointer offset UB if no string has been set -- though I'm wondering if this case shouldn't be generating a warning?
2020-08-27 10:24:58 +00:00
return_false:
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
RETVAL_FALSE;
Clean up strtok implementation Store the zend_string instead of performing a copy and storing in a zval. Also make sure the string is released immediately if it's no longer needed. Finally, avoid null pointer offset UB if no string has been set -- though I'm wondering if this case shouldn't be generating a warning?
2020-08-27 10:24:58 +00:00
zend_string_release(BG(strtok_string));
BG(strtok_string) = NULL;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Un-revert patch
2001-08-03 23:09:05 +00:00
style & ws fixes
2003-02-08 15:26:17 +00:00
/* Restore table -- usually faster then memset'ing the table on every invocation */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
token = ZSTR_VAL(tok);
style & ws fixes
2003-02-08 15:26:17 +00:00
while (token < token_end) {
Un-revert patch
2001-08-03 23:09:05 +00:00
STRTOK_TABLE(token++) = 0;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strtoupper */
(PHP strtoupper, strtolower, substr_replace) Made binary-safe. @Made strtoupper(), strtolower(), substr_replace() binary-safe. (Andrei)
1999-12-09 23:45:30 +00:00
PHPAPI char *php_strtoupper(char *s, size_t len)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
unsigned char *c;
const unsigned char *e;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
c = (unsigned char *)s;
e = (unsigned char *)c+len;
Slight optimization of php_strtoupper & php_strtoupper functions.
2002-08-25 19:08:07 +00:00
ws fixes.
2002-10-02 17:56:04 +00:00
while (c < e) {
Slight optimization of php_strtoupper & php_strtoupper functions.
2002-08-25 19:08:07 +00:00
*c = toupper(*c);
c++;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
return s;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_string_toupper */
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
PHPAPI zend_string *php_string_toupper(zend_string *s)
{
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
unsigned char *c;
const unsigned char *e;
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
c = (unsigned char *)ZSTR_VAL(s);
e = c + ZSTR_LEN(s);
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
while (c < e) {
Optimize strtolower()/strtoupper()
2015-12-13 05:22:50 +00:00
if (islower(*c)) {
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
register unsigned char *r;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
zend_string *res = zend_string_alloc(ZSTR_LEN(s), 0);
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (c != (unsigned char*)ZSTR_VAL(s)) {
memcpy(ZSTR_VAL(res), ZSTR_VAL(s), c - (unsigned char*)ZSTR_VAL(s));
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
r = c + (ZSTR_VAL(res) - ZSTR_VAL(s));
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
while (c < e) {
*r = toupper(*c);
r++;
c++;
}
*r = '\0';
return res;
}
c++;
}
return zend_string_copy(s);
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Makes a string uppercase */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strtoupper)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *arg;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use FAST_ZPP in few more frequently used functions
2015-01-12 16:26:12 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(arg)
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
RETURN_STR(php_string_toupper(arg));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strtolower */
(PHP strtoupper, strtolower, substr_replace) Made binary-safe. @Made strtoupper(), strtolower(), substr_replace() binary-safe. (Andrei)
1999-12-09 23:45:30 +00:00
PHPAPI char *php_strtolower(char *s, size_t len)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
unsigned char *c;
const unsigned char *e;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
c = (unsigned char *)s;
Slight optimization of php_strtoupper & php_strtoupper functions.
2002-08-25 19:08:07 +00:00
e = c+len;
ws fixes.
2002-10-02 17:56:04 +00:00
while (c < e) {
Slight optimization of php_strtoupper & php_strtoupper functions.
2002-08-25 19:08:07 +00:00
*c = tolower(*c);
c++;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
return s;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_string_tolower */
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
PHPAPI zend_string *php_string_tolower(zend_string *s)
{
Use const pointers at some appropriate places
2018-03-16 14:42:41 +00:00
unsigned char *c;
const unsigned char *e;
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
Ensure ctype_string is NULL for C locale We already document that this is the case, but currently it's only true if setlocale() has not been called. Make sure ctype_string is always NULL, even with an explicit "C" locale call, so we can more efficiently check whether we are in the "C" locale. Closes GH-5542.
2020-05-07 16:50:42 +00:00
if (EXPECTED(!BG(ctype_string))) {
Use ascii lowercase if locale is default C
2020-04-28 03:33:02 +00:00
return zend_string_tolower(s);
} else {
c = (unsigned char *)ZSTR_VAL(s);
e = c + ZSTR_LEN(s);
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
Use ascii lowercase if locale is default C
2020-04-28 03:33:02 +00:00
while (c < e) {
if (isupper(*c)) {
register unsigned char *r;
zend_string *res = zend_string_alloc(ZSTR_LEN(s), 0);
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
Use ascii lowercase if locale is default C
2020-04-28 03:33:02 +00:00
if (c != (unsigned char*)ZSTR_VAL(s)) {
memcpy(ZSTR_VAL(res), ZSTR_VAL(s), c - (unsigned char*)ZSTR_VAL(s));
}
r = c + (ZSTR_VAL(res) - ZSTR_VAL(s));
while (c < e) {
*r = tolower(*c);
r++;
c++;
}
*r = '\0';
return res;
Revert "Merge branch 'sse2_strtolower' of https://github.com/laruence/php-src" This reverts commit 5a6373f904c6497551cd53baf323ddb854a553e1, reversing changes made to 45cb42166d4a53fe0154be08097f112d6ec72a27. This breaks strtolower-win32.phpt.
2020-04-29 07:49:01 +00:00
}
Use ascii lowercase if locale is default C
2020-04-28 03:33:02 +00:00
c++;
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
}
Use ascii lowercase if locale is default C
2020-04-28 03:33:02 +00:00
return zend_string_copy(s);
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Makes a string lowercase */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strtolower)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
Z_PARAM_STR(str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Add php_string_tolower/toupper which is locale depened
2014-12-25 03:28:54 +00:00
RETURN_STR(php_string_tolower(str));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_basename */
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
PHPAPI zend_string *php_basename(const char *s, size_t len, const char *suffix, size_t suffix_len)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
/* State 0 is directly after a directory separator (or at the start of the string).
* State 1 is everything else. */
int state = 0;
const char *basename_start = s;
const char *basename_end = s;
while (len > 0) {
int inc_len = (*s == '\0' ? 1 : php_mblen(s, len));
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
switch (inc_len) {
case 0:
goto quit_loop;
case 1:
Remove Netware support If this does not break the Unix system somehow, I'll be amazed. This should get most of it out, apologies for any errors this may cause on non-Windows ends which I cannot test atm.
2016-11-12 10:20:01 +00:00
#if defined(PHP_WIN32)
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
if (*s == '/' || *s == '\\') {
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
#else
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
if (*s == '/') {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
#endif
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
if (state == 1) {
state = 0;
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
basename_end = s;
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
}
Remove Netware support If this does not break the Unix system somehow, I'll be amazed. This should get most of it out, apologies for any errors this may cause on non-Windows ends which I cannot test atm.
2016-11-12 10:20:01 +00:00
#if defined(PHP_WIN32)
improved the fix for bug #66395 - fixed the traverse vulnerability in case like c:d:file.txt - ensure paths containing NTFS streams are still worky
2014-01-05 18:05:04 +00:00
/* Catch relative paths in c:file.txt style. They're not to confuse
trailing whitespace removal
2015-01-03 09:22:58 +00:00
with the NTFS streams. This part ensures also, that no drive
improved the fix for bug #66395 - fixed the traverse vulnerability in case like c:d:file.txt - ensure paths containing NTFS streams are still worky
2014-01-05 18:05:04 +00:00
letter traversing happens. */
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
} else if ((*s == ':' && (s - basename_start == 1))) {
improved the fix for bug #66395 - fixed the traverse vulnerability in case like c:d:file.txt - ensure paths containing NTFS streams are still worky
2014-01-05 18:05:04 +00:00
if (state == 0) {
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
basename_start = s;
improved the fix for bug #66395 - fixed the traverse vulnerability in case like c:d:file.txt - ensure paths containing NTFS streams are still worky
2014-01-05 18:05:04 +00:00
state = 1;
} else {
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
basename_end = s;
improved the fix for bug #66395 - fixed the traverse vulnerability in case like c:d:file.txt - ensure paths containing NTFS streams are still worky
2014-01-05 18:05:04 +00:00
state = 0;
}
#endif
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
} else {
if (state == 0) {
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
basename_start = s;
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
state = 1;
}
}
- fix bug #37945 pathinfo() cannot handle argument with special characters like german "Umlaute"
2006-07-17 20:36:47 +00:00
break;
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
default:
Treat invalid characters in basename() consistently Always simply ignore (pass through) them. Previously the behavior depended on where the invalid character occurred, as it messed up the state management.
2020-04-29 16:42:25 +00:00
if (inc_len < 0) {
/* If character is invalid, treat it like other non-significant characters. */
inc_len = 1;
php_mb_reset();
}
- fix bug #37945 pathinfo() cannot handle argument with special characters like german "Umlaute"
2006-07-17 20:36:47 +00:00
if (state == 0) {
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
basename_start = s;
- fix bug #37945 pathinfo() cannot handle argument with special characters like german "Umlaute"
2006-07-17 20:36:47 +00:00
state = 1;
}
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
break;
}
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
s += inc_len;
len -= inc_len;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
Fix bug #26574 (basename() doesn't work properly with multibyte characters)
2003-12-10 07:15:28 +00:00
quit_loop:
if (state == 1) {
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
basename_end = s;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
basename() now binary safe, testcases added
2003-05-05 13:46:14 +00:00
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
if (suffix != NULL && suffix_len < (size_t)(basename_end - basename_start) &&
memcmp(basename_end - suffix_len, suffix, suffix_len) == 0) {
basename_end -= suffix_len;
}
basename() now binary safe, testcases added
2003-05-05 13:46:14 +00:00
Clean up naming in basename() implementation
2020-04-29 16:37:51 +00:00
return zend_string_init(basename_start, basename_end - basename_start, 0);
Need a PHPAPI version of basename for some stuff I am working on. Also fixed a bug along the way in the basename function. If it was fed something like "filename.ext/////" it would return the string with all the slashes whereas if you fed it "/path/filename.ext////" it would get it right. @ Fixed basename() bug where "file.ext///" would not return the same @ as "/path/file.ext///" (Rasmus)
2000-05-23 20:16:14 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Need a PHPAPI version of basename for some stuff I am working on. Also fixed a bug along the way in the basename function. If it was fed something like "filename.ext/////" it would return the string with all the slashes whereas if you fed it "/path/filename.ext////" it would get it right. @ Fixed basename() bug where "file.ext///" would not return the same @ as "/path/file.ext///" (Rasmus)
2000-05-23 20:16:14 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns the filename component of the path */
Need a PHPAPI version of basename for some stuff I am working on. Also fixed a bug along the way in the basename function. If it was fed something like "filename.ext/////" it would return the string with all the slashes whereas if you fed it "/path/filename.ext////" it would get it right. @ Fixed basename() bug where "file.ext///" would not return the same @ as "/path/file.ext///" (Rasmus)
2000-05-23 20:16:14 +00:00
PHP_FUNCTION(basename)
{
Refactor basename using zend_string
2014-02-23 10:08:35 +00:00
char *string, *suffix = NULL;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t string_len, suffix_len = 0;
Need a PHPAPI version of basename for some stuff I am working on. Also fixed a bug along the way in the basename function. If it was fed something like "filename.ext/////" it would return the string with all the slashes whereas if you fed it "/path/filename.ext////" it would get it right. @ Fixed basename() bug where "file.ext///" would not return the same @ as "/path/file.ext///" (Rasmus)
2000-05-23 20:16:14 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STRING(string, string_len)
Z_PARAM_OPTIONAL
Z_PARAM_STRING(suffix, suffix_len)
ZEND_PARSE_PARAMETERS_END();
@added optional suffix removal parameter to basename() (Hartmut) added optional suffix removal parameter to basename() this makes it more simmilar to the basename command
2001-06-23 00:51:03 +00:00
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
RETURN_STR(php_basename(string, string_len, suffix, suffix_len));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* {{{ php_dirname
Adjust dirname() on Win32 to match CWD per drive semantics.
2002-05-16 16:04:45 +00:00
Returns directory name component of path */
Made dirname() binary-safe.
2003-04-12 21:53:46 +00:00
PHPAPI size_t php_dirname(char *path, size_t len)
- Fix VC++ warning
2000-10-03 14:43:04 +00:00
{
[DOC] Add compile-time __DIR__ constant which implements dirname(__FILE__)
2008-02-12 00:21:15 +00:00
return zend_dirname(path, len);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns the directory name component of the path */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(dirname)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
New parameter parsing API for string, part I
2008-06-22 19:22:41 +00:00
char *str;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t str_len;
Cleanup and also include the error path into test
2015-07-27 14:22:13 +00:00
zend_string *ret;
Fix #70112 RFE Allow dirname to go up various times
2015-07-25 14:58:36 +00:00
zend_long levels = 1;
Made dirname() binary-safe.
2003-04-12 21:53:46 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STRING(str, str_len)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(levels)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
master renames phase 1
2014-08-25 17:24:55 +00:00
ret = zend_string_init(str, str_len, 0);
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
Fix #70112 RFE Allow dirname to go up various times
2015-07-25 14:58:36 +00:00
if (levels == 1) {
Fix some misspellings
2018-02-05 00:16:50 +00:00
/* Default case */
Fixed the UTF-8 and long path support in the streams on Windows. Since long the default PHP charset is UTF-8, however the Windows part is out of step with this important point. The current implementation in PHP doesn't technically permit to handle UTF-8 filepath and several other things. Till now, only the ANSI compatible APIs are being used. Here is more about it https://msdn.microsoft.com/en-us/library/windows/desktop/dd317752%28v=vs.85%29.aspx The patch fixes not only issues with multibyte filenames under incompatible codepages, but indirectly also issues with some other multibyte encodings like BIG5, Shift-JIS, etc. by providing a clean way to access filenames in UTF-8. Below is a small list of issues from the bug tracker, that are getting fixed: https://bugs.php.net/63401 https://bugs.php.net/41199 https://bugs.php.net/50203 https://bugs.php.net/71509 https://bugs.php.net/64699 https://bugs.php.net/64506 https://bugs.php.net/30195 https://bugs.php.net/65358 https://bugs.php.net/61315 https://bugs.php.net/70943 https://bugs.php.net/70903 https://bugs.php.net/63593 https://bugs.php.net/54977 https://bugs.php.net/54028 https://bugs.php.net/43148 https://bugs.php.net/30730 https://bugs.php.net/33350 https://bugs.php.net/35300 https://bugs.php.net/46990 https://bugs.php.net/61309 https://bugs.php.net/69333 https://bugs.php.net/45517 https://bugs.php.net/70551 https://bugs.php.net/50197 https://bugs.php.net/72200 https://bugs.php.net/37672 Yet more related tickets can for sure be found - on bugs.php.net, Stackoverflow and Github. Some of the bugs are pretty recent, some descend to early 2000th, but the user comments in there last even till today. Just for example, bug #30195 was opened in 2004, the latest comment in there was made in 2014. It is certain, that these bugs descend not only to pure PHP use cases, but get also redirected from the popular PHP based projects. Given the modern systems (and those supported by PHP) are always based on NTFS, there is no excuse to keep these issues unresolved. The internalization approach on Windows is in many ways different from UNIX and Linux, while it supports and is based on Unicode. It depends on the current system code page, APIs used and exact kind how the binary was compiled The locale doesn't affect the way Unicode or ANSI API work. PHP in particular is being compiled without _UNICODE defined and this is conditioned by the way we handle strings. Here is more about it https://msdn.microsoft.com/en-us/library/tsbaswba.aspx However, with any system code page ANSI functions automatically convert paths to UTF-16. Paths in some encodings incompatible with the current system code page, won't work correctly with ANSI APIs. PHP till now only uses the ANSI Windows APIs. For example, on a system with the current code page 1252, the paths in cp1252 are supported and transparently converted to UTF-16 by the ANSI functions. Once one wants to handle a filepath encoded with cp932 on that particular system, an ANSI or a POSIX compatible function used in PHP will produce an erroneous result. When trying to convert that cp932 path to UTF-8 and passing to the ANSI functions, an ANSI function would likely interpret the UTF-8 string as some string in the current code page and create a filepath that represents every single byte of the UTF-8 string. These behaviors are not only broken but also disregard the documented INI settings. This patch solves the issies with the multibyte paths on Windows by intelligently enforcing the usage of the Unicode aware APIs. For functions expect Unicode (fe CreateFileW, FindFirstFileW, etc.), arguments will be converted to UTF-16 wide chars. For functions returning Unicode aware data (fe GetCurrentDirectoryW, etc.), resulting wide string is converted back to char's depending on the current PHP charset settings, either to the current ANSI codepage (this is the behavior prior to this patch) or to UTF-8 (the default behavior). In a particular case, users might have to explicitly set internal_encoding or default_charset, if filenames in ANSI codepage are necessary. Current tests show no regressions and witness that this will be an exotic case, the current default UTF-8 encoding is compatible with any supported system. The dependency libraries are long switching to Unicode APIs, so some tests were also added for extensions not directly related to streams. At large, the patch brings over 150 related tests into the core. Those target and was run on various environments with European, Asian, etc. codepages. General PHP frameworks was tested and showed no regressions. The impact on the current C code base is low, the most places affected are the Windows only places in the three files tsrm_win32.c, zend_virtual_cwd.c and plain_wrapper.c. The actual implementation of the most of the wide char supporting functionality is in win32/ioutil.* and win32/codepage.*, several low level functionsare extended in place to avoid reimplementation for now. No performance impact was sighted. As previously mentioned, the ANSI APIs used prior the patch perform Unicode conversions internally. Using the Unicode APIs directly while doing custom conversions just retains the status quo. The ways to optimize it are open (fe. by implementing caching for the strings converted to wide variants). The long path implementation is user transparent. If a path exceeds the length of _MAX_PATH, it'll be automatically prefixed with \\?\. The MAXPATHLEN is set to 2048 bytes. Appreciation to Pierre Joye, Matt Ficken, @algo13 and others for tips, ideas and testing. Thanks.
2016-06-20 07:32:19 +00:00
#ifdef PHP_WIN32
ZSTR_LEN(ret) = php_win32_ioutil_dirname(ZSTR_VAL(ret), str_len);
#else
Fix #70112 RFE Allow dirname to go up various times
2015-07-25 14:58:36 +00:00
ZSTR_LEN(ret) = zend_dirname(ZSTR_VAL(ret), str_len);
Fixed the UTF-8 and long path support in the streams on Windows. Since long the default PHP charset is UTF-8, however the Windows part is out of step with this important point. The current implementation in PHP doesn't technically permit to handle UTF-8 filepath and several other things. Till now, only the ANSI compatible APIs are being used. Here is more about it https://msdn.microsoft.com/en-us/library/windows/desktop/dd317752%28v=vs.85%29.aspx The patch fixes not only issues with multibyte filenames under incompatible codepages, but indirectly also issues with some other multibyte encodings like BIG5, Shift-JIS, etc. by providing a clean way to access filenames in UTF-8. Below is a small list of issues from the bug tracker, that are getting fixed: https://bugs.php.net/63401 https://bugs.php.net/41199 https://bugs.php.net/50203 https://bugs.php.net/71509 https://bugs.php.net/64699 https://bugs.php.net/64506 https://bugs.php.net/30195 https://bugs.php.net/65358 https://bugs.php.net/61315 https://bugs.php.net/70943 https://bugs.php.net/70903 https://bugs.php.net/63593 https://bugs.php.net/54977 https://bugs.php.net/54028 https://bugs.php.net/43148 https://bugs.php.net/30730 https://bugs.php.net/33350 https://bugs.php.net/35300 https://bugs.php.net/46990 https://bugs.php.net/61309 https://bugs.php.net/69333 https://bugs.php.net/45517 https://bugs.php.net/70551 https://bugs.php.net/50197 https://bugs.php.net/72200 https://bugs.php.net/37672 Yet more related tickets can for sure be found - on bugs.php.net, Stackoverflow and Github. Some of the bugs are pretty recent, some descend to early 2000th, but the user comments in there last even till today. Just for example, bug #30195 was opened in 2004, the latest comment in there was made in 2014. It is certain, that these bugs descend not only to pure PHP use cases, but get also redirected from the popular PHP based projects. Given the modern systems (and those supported by PHP) are always based on NTFS, there is no excuse to keep these issues unresolved. The internalization approach on Windows is in many ways different from UNIX and Linux, while it supports and is based on Unicode. It depends on the current system code page, APIs used and exact kind how the binary was compiled The locale doesn't affect the way Unicode or ANSI API work. PHP in particular is being compiled without _UNICODE defined and this is conditioned by the way we handle strings. Here is more about it https://msdn.microsoft.com/en-us/library/tsbaswba.aspx However, with any system code page ANSI functions automatically convert paths to UTF-16. Paths in some encodings incompatible with the current system code page, won't work correctly with ANSI APIs. PHP till now only uses the ANSI Windows APIs. For example, on a system with the current code page 1252, the paths in cp1252 are supported and transparently converted to UTF-16 by the ANSI functions. Once one wants to handle a filepath encoded with cp932 on that particular system, an ANSI or a POSIX compatible function used in PHP will produce an erroneous result. When trying to convert that cp932 path to UTF-8 and passing to the ANSI functions, an ANSI function would likely interpret the UTF-8 string as some string in the current code page and create a filepath that represents every single byte of the UTF-8 string. These behaviors are not only broken but also disregard the documented INI settings. This patch solves the issies with the multibyte paths on Windows by intelligently enforcing the usage of the Unicode aware APIs. For functions expect Unicode (fe CreateFileW, FindFirstFileW, etc.), arguments will be converted to UTF-16 wide chars. For functions returning Unicode aware data (fe GetCurrentDirectoryW, etc.), resulting wide string is converted back to char's depending on the current PHP charset settings, either to the current ANSI codepage (this is the behavior prior to this patch) or to UTF-8 (the default behavior). In a particular case, users might have to explicitly set internal_encoding or default_charset, if filenames in ANSI codepage are necessary. Current tests show no regressions and witness that this will be an exotic case, the current default UTF-8 encoding is compatible with any supported system. The dependency libraries are long switching to Unicode APIs, so some tests were also added for extensions not directly related to streams. At large, the patch brings over 150 related tests into the core. Those target and was run on various environments with European, Asian, etc. codepages. General PHP frameworks was tested and showed no regressions. The impact on the current C code base is low, the most places affected are the Windows only places in the three files tsrm_win32.c, zend_virtual_cwd.c and plain_wrapper.c. The actual implementation of the most of the wide char supporting functionality is in win32/ioutil.* and win32/codepage.*, several low level functionsare extended in place to avoid reimplementation for now. No performance impact was sighted. As previously mentioned, the ANSI APIs used prior the patch perform Unicode conversions internally. Using the Unicode APIs directly while doing custom conversions just retains the status quo. The ways to optimize it are open (fe. by implementing caching for the strings converted to wide variants). The long path implementation is user transparent. If a path exceeds the length of _MAX_PATH, it'll be automatically prefixed with \\?\. The MAXPATHLEN is set to 2048 bytes. Appreciation to Pierre Joye, Matt Ficken, @algo13 and others for tips, ideas and testing. Thanks.
2016-06-20 07:32:19 +00:00
#endif
Fix #70112 RFE Allow dirname to go up various times
2015-07-25 14:58:36 +00:00
} else if (levels < 1) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(2, "must be greater than or equal to 1");
Avoid useless checks, using zend_string_efree(), in cases where the string is known to be a temporary allocated zend_string.
2018-05-08 14:30:15 +00:00
zend_string_efree(ret);
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Fix #70112 RFE Allow dirname to go up various times
2015-07-25 14:58:36 +00:00
} else {
/* Some levels up */
do {
Fixed the UTF-8 and long path support in the streams on Windows. Since long the default PHP charset is UTF-8, however the Windows part is out of step with this important point. The current implementation in PHP doesn't technically permit to handle UTF-8 filepath and several other things. Till now, only the ANSI compatible APIs are being used. Here is more about it https://msdn.microsoft.com/en-us/library/windows/desktop/dd317752%28v=vs.85%29.aspx The patch fixes not only issues with multibyte filenames under incompatible codepages, but indirectly also issues with some other multibyte encodings like BIG5, Shift-JIS, etc. by providing a clean way to access filenames in UTF-8. Below is a small list of issues from the bug tracker, that are getting fixed: https://bugs.php.net/63401 https://bugs.php.net/41199 https://bugs.php.net/50203 https://bugs.php.net/71509 https://bugs.php.net/64699 https://bugs.php.net/64506 https://bugs.php.net/30195 https://bugs.php.net/65358 https://bugs.php.net/61315 https://bugs.php.net/70943 https://bugs.php.net/70903 https://bugs.php.net/63593 https://bugs.php.net/54977 https://bugs.php.net/54028 https://bugs.php.net/43148 https://bugs.php.net/30730 https://bugs.php.net/33350 https://bugs.php.net/35300 https://bugs.php.net/46990 https://bugs.php.net/61309 https://bugs.php.net/69333 https://bugs.php.net/45517 https://bugs.php.net/70551 https://bugs.php.net/50197 https://bugs.php.net/72200 https://bugs.php.net/37672 Yet more related tickets can for sure be found - on bugs.php.net, Stackoverflow and Github. Some of the bugs are pretty recent, some descend to early 2000th, but the user comments in there last even till today. Just for example, bug #30195 was opened in 2004, the latest comment in there was made in 2014. It is certain, that these bugs descend not only to pure PHP use cases, but get also redirected from the popular PHP based projects. Given the modern systems (and those supported by PHP) are always based on NTFS, there is no excuse to keep these issues unresolved. The internalization approach on Windows is in many ways different from UNIX and Linux, while it supports and is based on Unicode. It depends on the current system code page, APIs used and exact kind how the binary was compiled The locale doesn't affect the way Unicode or ANSI API work. PHP in particular is being compiled without _UNICODE defined and this is conditioned by the way we handle strings. Here is more about it https://msdn.microsoft.com/en-us/library/tsbaswba.aspx However, with any system code page ANSI functions automatically convert paths to UTF-16. Paths in some encodings incompatible with the current system code page, won't work correctly with ANSI APIs. PHP till now only uses the ANSI Windows APIs. For example, on a system with the current code page 1252, the paths in cp1252 are supported and transparently converted to UTF-16 by the ANSI functions. Once one wants to handle a filepath encoded with cp932 on that particular system, an ANSI or a POSIX compatible function used in PHP will produce an erroneous result. When trying to convert that cp932 path to UTF-8 and passing to the ANSI functions, an ANSI function would likely interpret the UTF-8 string as some string in the current code page and create a filepath that represents every single byte of the UTF-8 string. These behaviors are not only broken but also disregard the documented INI settings. This patch solves the issies with the multibyte paths on Windows by intelligently enforcing the usage of the Unicode aware APIs. For functions expect Unicode (fe CreateFileW, FindFirstFileW, etc.), arguments will be converted to UTF-16 wide chars. For functions returning Unicode aware data (fe GetCurrentDirectoryW, etc.), resulting wide string is converted back to char's depending on the current PHP charset settings, either to the current ANSI codepage (this is the behavior prior to this patch) or to UTF-8 (the default behavior). In a particular case, users might have to explicitly set internal_encoding or default_charset, if filenames in ANSI codepage are necessary. Current tests show no regressions and witness that this will be an exotic case, the current default UTF-8 encoding is compatible with any supported system. The dependency libraries are long switching to Unicode APIs, so some tests were also added for extensions not directly related to streams. At large, the patch brings over 150 related tests into the core. Those target and was run on various environments with European, Asian, etc. codepages. General PHP frameworks was tested and showed no regressions. The impact on the current C code base is low, the most places affected are the Windows only places in the three files tsrm_win32.c, zend_virtual_cwd.c and plain_wrapper.c. The actual implementation of the most of the wide char supporting functionality is in win32/ioutil.* and win32/codepage.*, several low level functionsare extended in place to avoid reimplementation for now. No performance impact was sighted. As previously mentioned, the ANSI APIs used prior the patch perform Unicode conversions internally. Using the Unicode APIs directly while doing custom conversions just retains the status quo. The ways to optimize it are open (fe. by implementing caching for the strings converted to wide variants). The long path implementation is user transparent. If a path exceeds the length of _MAX_PATH, it'll be automatically prefixed with \\?\. The MAXPATHLEN is set to 2048 bytes. Appreciation to Pierre Joye, Matt Ficken, @algo13 and others for tips, ideas and testing. Thanks.
2016-06-20 07:32:19 +00:00
#ifdef PHP_WIN32
ZSTR_LEN(ret) = php_win32_ioutil_dirname(ZSTR_VAL(ret), str_len = ZSTR_LEN(ret));
#else
Cleanup and also include the error path into test
2015-07-27 14:22:13 +00:00
ZSTR_LEN(ret) = zend_dirname(ZSTR_VAL(ret), str_len = ZSTR_LEN(ret));
Fixed the UTF-8 and long path support in the streams on Windows. Since long the default PHP charset is UTF-8, however the Windows part is out of step with this important point. The current implementation in PHP doesn't technically permit to handle UTF-8 filepath and several other things. Till now, only the ANSI compatible APIs are being used. Here is more about it https://msdn.microsoft.com/en-us/library/windows/desktop/dd317752%28v=vs.85%29.aspx The patch fixes not only issues with multibyte filenames under incompatible codepages, but indirectly also issues with some other multibyte encodings like BIG5, Shift-JIS, etc. by providing a clean way to access filenames in UTF-8. Below is a small list of issues from the bug tracker, that are getting fixed: https://bugs.php.net/63401 https://bugs.php.net/41199 https://bugs.php.net/50203 https://bugs.php.net/71509 https://bugs.php.net/64699 https://bugs.php.net/64506 https://bugs.php.net/30195 https://bugs.php.net/65358 https://bugs.php.net/61315 https://bugs.php.net/70943 https://bugs.php.net/70903 https://bugs.php.net/63593 https://bugs.php.net/54977 https://bugs.php.net/54028 https://bugs.php.net/43148 https://bugs.php.net/30730 https://bugs.php.net/33350 https://bugs.php.net/35300 https://bugs.php.net/46990 https://bugs.php.net/61309 https://bugs.php.net/69333 https://bugs.php.net/45517 https://bugs.php.net/70551 https://bugs.php.net/50197 https://bugs.php.net/72200 https://bugs.php.net/37672 Yet more related tickets can for sure be found - on bugs.php.net, Stackoverflow and Github. Some of the bugs are pretty recent, some descend to early 2000th, but the user comments in there last even till today. Just for example, bug #30195 was opened in 2004, the latest comment in there was made in 2014. It is certain, that these bugs descend not only to pure PHP use cases, but get also redirected from the popular PHP based projects. Given the modern systems (and those supported by PHP) are always based on NTFS, there is no excuse to keep these issues unresolved. The internalization approach on Windows is in many ways different from UNIX and Linux, while it supports and is based on Unicode. It depends on the current system code page, APIs used and exact kind how the binary was compiled The locale doesn't affect the way Unicode or ANSI API work. PHP in particular is being compiled without _UNICODE defined and this is conditioned by the way we handle strings. Here is more about it https://msdn.microsoft.com/en-us/library/tsbaswba.aspx However, with any system code page ANSI functions automatically convert paths to UTF-16. Paths in some encodings incompatible with the current system code page, won't work correctly with ANSI APIs. PHP till now only uses the ANSI Windows APIs. For example, on a system with the current code page 1252, the paths in cp1252 are supported and transparently converted to UTF-16 by the ANSI functions. Once one wants to handle a filepath encoded with cp932 on that particular system, an ANSI or a POSIX compatible function used in PHP will produce an erroneous result. When trying to convert that cp932 path to UTF-8 and passing to the ANSI functions, an ANSI function would likely interpret the UTF-8 string as some string in the current code page and create a filepath that represents every single byte of the UTF-8 string. These behaviors are not only broken but also disregard the documented INI settings. This patch solves the issies with the multibyte paths on Windows by intelligently enforcing the usage of the Unicode aware APIs. For functions expect Unicode (fe CreateFileW, FindFirstFileW, etc.), arguments will be converted to UTF-16 wide chars. For functions returning Unicode aware data (fe GetCurrentDirectoryW, etc.), resulting wide string is converted back to char's depending on the current PHP charset settings, either to the current ANSI codepage (this is the behavior prior to this patch) or to UTF-8 (the default behavior). In a particular case, users might have to explicitly set internal_encoding or default_charset, if filenames in ANSI codepage are necessary. Current tests show no regressions and witness that this will be an exotic case, the current default UTF-8 encoding is compatible with any supported system. The dependency libraries are long switching to Unicode APIs, so some tests were also added for extensions not directly related to streams. At large, the patch brings over 150 related tests into the core. Those target and was run on various environments with European, Asian, etc. codepages. General PHP frameworks was tested and showed no regressions. The impact on the current C code base is low, the most places affected are the Windows only places in the three files tsrm_win32.c, zend_virtual_cwd.c and plain_wrapper.c. The actual implementation of the most of the wide char supporting functionality is in win32/ioutil.* and win32/codepage.*, several low level functionsare extended in place to avoid reimplementation for now. No performance impact was sighted. As previously mentioned, the ANSI APIs used prior the patch perform Unicode conversions internally. Using the Unicode APIs directly while doing custom conversions just retains the status quo. The ways to optimize it are open (fe. by implementing caching for the strings converted to wide variants). The long path implementation is user transparent. If a path exceeds the length of _MAX_PATH, it'll be automatically prefixed with \\?\. The MAXPATHLEN is set to 2048 bytes. Appreciation to Pierre Joye, Matt Ficken, @algo13 and others for tips, ideas and testing. Thanks.
2016-06-20 07:32:19 +00:00
#endif
Cleanup and also include the error path into test
2015-07-27 14:22:13 +00:00
} while (ZSTR_LEN(ret) < str_len && --levels);
Fix #70112 RFE Allow dirname to go up various times
2015-07-25 14:58:36 +00:00
}
Cleanup and also include the error path into test
2015-07-27 14:22:13 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(ret);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns information about a certain string */
@- Added the pathinfo() function. (Sterling)
2000-09-10 05:59:16 +00:00
PHP_FUNCTION(pathinfo)
{
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zval tmp;
Refactor basename using zend_string
2014-02-23 10:08:35 +00:00
char *path, *dirname;
's' works with size_t round 2
2014-08-27 15:31:24 +00:00
size_t path_len;
int have_basename;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long opt = PHP_PATHINFO_ALL;
Fixed memory leak in ext/standard/tests/strings/bug24098.phpt
2014-02-24 08:14:46 +00:00
zend_string *ret = NULL;
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STRING(path, path_len)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(opt)
ZEND_PARSE_PARAMETERS_END();
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
MFH
2006-07-17 06:45:08 +00:00
have_basename = ((opt & PHP_PATHINFO_BASENAME) == PHP_PATHINFO_BASENAME);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
array_init(&tmp);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
if ((opt & PHP_PATHINFO_DIRNAME) == PHP_PATHINFO_DIRNAME) {
Refactor basename using zend_string
2014-02-23 10:08:35 +00:00
dirname = estrndup(path, path_len);
php_dirname(dirname, path_len);
if (*dirname) {
Cleanup (1-st round)
2014-04-15 11:40:40 +00:00
add_assoc_string(&tmp, "dirname", dirname);
style & ws fixes
2003-02-08 15:26:17 +00:00
}
Refactor basename using zend_string
2014-02-23 10:08:35 +00:00
efree(dirname);
@- Added the pathinfo() function. (Sterling)
2000-09-10 05:59:16 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
MFH
2006-07-17 06:45:08 +00:00
if (have_basename) {
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
ret = php_basename(path, path_len, NULL, 0);
master renames phase 1
2014-08-25 17:24:55 +00:00
add_assoc_str(&tmp, "basename", zend_string_copy(ret));
MFH
2006-07-17 06:45:08 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
if ((opt & PHP_PATHINFO_EXTENSION) == PHP_PATHINFO_EXTENSION) {
Kill some more constness warnings in com_dotnet/phar/standard
2010-11-19 16:03:00 +00:00
const char *p;
fixed trim() and strtok() to work with big strings
2014-08-28 19:59:00 +00:00
ptrdiff_t idx;
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
if (!have_basename) {
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
ret = php_basename(path, path_len, NULL, 0);
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = zend_memrchr(ZSTR_VAL(ret), '.', ZSTR_LEN(ret));
@- Added the pathinfo() function. (Sterling)
2000-09-10 05:59:16 +00:00
@- Fixed crash in pathinfo()
2001-04-13 18:22:13 +00:00
if (p) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
idx = p - ZSTR_VAL(ret);
add_assoc_stringl(&tmp, "extension", ZSTR_VAL(ret) + idx + 1, ZSTR_LEN(ret) - idx - 1);
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
}
@- Added the pathinfo() function. (Sterling)
2000-09-10 05:59:16 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
if ((opt & PHP_PATHINFO_FILENAME) == PHP_PATHINFO_FILENAME) {
Kill some more constness warnings in com_dotnet/phar/standard
2010-11-19 16:03:00 +00:00
const char *p;
fixed trim() and strtok() to work with big strings
2014-08-28 19:59:00 +00:00
ptrdiff_t idx;
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
typo fixes
2013-07-13 12:37:04 +00:00
/* Have we already looked up the basename? */
MFH
2006-07-17 06:45:08 +00:00
if (!have_basename && !ret) {
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
ret = php_basename(path, path_len, NULL, 0);
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = zend_memrchr(ZSTR_VAL(ret), '.', ZSTR_LEN(ret));
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
Fixed compilation warnings
2016-06-23 09:47:06 +00:00
idx = p ? (p - ZSTR_VAL(ret)) : (ptrdiff_t)ZSTR_LEN(ret);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
add_assoc_stringl(&tmp, "filename", ZSTR_VAL(ret), idx);
MFH
2006-07-17 06:45:08 +00:00
}
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
Fixed memory leak in ext/standard/tests/strings/bug24098.phpt
2014-02-24 08:14:46 +00:00
if (ret) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(ret, 0);
- MFH Pathinfo allows to get filename (Toby S, Christian S)
2006-07-16 10:37:24 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
if (opt == PHP_PATHINFO_ALL) {
Avoid zval duplication in ZVAL_ZVAL() macro (it was necessary only in few places). Switch from ZVAL_ZVAL() to simpler macros where possible (it makes sense to review remaining places)
2015-06-12 09:33:23 +00:00
ZVAL_COPY_VALUE(return_value, &tmp);
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
} else {
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zval *element;
if ((element = zend_hash_get_current_data(Z_ARRVAL(tmp))) != NULL) {
Uze ZVAL_COPY_DEREF() instead of ZVAL_DEREF() and ZVAL_COPY()
2018-07-09 09:46:46 +00:00
ZVAL_COPY_DEREF(return_value, element);
Fix for bug #15130. Way too much effort for this bug, but cleaned up code a bit, use zend_parse_parameters(), etc, etc. We only look for extensions in the basename, not the full path.
2002-01-22 03:35:23 +00:00
} else {
ZVAL_EMPTY_STRING(return_value);
}
Avoid zval duplication in ZVAL_ZVAL() macro (it was necessary only in few places). Switch from ZVAL_ZVAL() to simpler macros where possible (it makes sense to review remaining places)
2015-06-12 09:33:23 +00:00
zval_ptr_dtor(&tmp);
@- Added the pathinfo() function. (Sterling)
2000-09-10 05:59:16 +00:00
}
}
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* {{{ php_stristr
Fix typos in code comments [skip ci]
2019-02-18 16:14:22 +00:00
case insensitive strstr */
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
PHPAPI char *php_stristr(char *s, char *t, size_t s_len, size_t t_len)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Made strstr(), stristr(), and ucwords() binary-safe(). Also corrected stristr() behavior to match that of strstr() if needle is not a string. @ Made strstr(), stristr(), and ucwords() binary-safe() (Andrei)
1999-12-11 19:50:01 +00:00
php_strtolower(s, s_len);
php_strtolower(t, t_len);
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
return (char*)php_memnstr(s, t, t_len, s + s_len);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strspn */
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
PHPAPI size_t php_strspn(const char *s1, const char *s2, const char *s1_end, const char *s2_end)
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
{
register const char *p = s1, *spanp;
register char c = *p;
cont:
style & ws fixes
2003-02-08 15:26:17 +00:00
for (spanp = s2; p != s1_end && spanp != s2_end;) {
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
if (*spanp++ == c) {
c = *(++p);
goto cont;
}
style & ws fixes
2003-02-08 15:26:17 +00:00
}
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
return (p - s1);
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strcspn */
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
PHPAPI size_t php_strcspn(const char *s1, const char *s2, const char *s1_end, const char *s2_end)
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
{
register const char *p, *spanp;
register char c = *s1;
for (p = s1;;) {
spanp = s2;
do {
style & ws fixes
2003-02-08 15:26:17 +00:00
if (*spanp == c || p == s1_end) {
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
return p - s1;
style & ws fixes
2003-02-08 15:26:17 +00:00
}
MFH: fix #39032 (strcspn() stops on null character)
2006-10-04 11:14:32 +00:00
} while (spanp++ < (s2_end - 1));
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
c = *++p;
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
}
/* NOTREACHED */
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Made strspn() and strcspn() binary-safe. # Please test if you can, especially cases with embedded chr(0). @ Made strspn() and strcspn() binary-safe.
1999-12-13 23:40:36 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds first occurrence of a string within another, case insensitive */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(stristr)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_string *haystack, *needle;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *found = NULL;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t found_offset;
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
char *haystack_dup;
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
char *orig_needle;
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
zend_bool part = 0;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Z_PARAM_STR(haystack)
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
Z_PARAM_STR(needle)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_OPTIONAL
Z_PARAM_BOOL(part)
ZEND_PARSE_PARAMETERS_END();
avoid lowercasing the string
2001-09-06 08:54:54 +00:00
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
haystack_dup = estrndup(ZSTR_VAL(haystack), ZSTR_LEN(haystack));
orig_needle = estrndup(ZSTR_VAL(needle), ZSTR_LEN(needle));
found = php_stristr(haystack_dup, orig_needle, ZSTR_LEN(haystack), ZSTR_LEN(needle));
efree(orig_needle);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (found) {
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
found_offset = found - haystack_dup;
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
if (part) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETVAL_STRINGL(ZSTR_VAL(haystack), found_offset);
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETVAL_STRINGL(ZSTR_VAL(haystack) + found_offset, ZSTR_LEN(haystack) - found_offset);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
} else {
Another leak bites the dust?
2001-09-27 00:21:21 +00:00
RETVAL_FALSE;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
efree(haystack_dup);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds first occurrence of a string within another */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strstr)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_string *haystack, *needle;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *found = NULL;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long found_offset;
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
zend_bool part = 0;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Z_PARAM_STR(haystack)
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
Z_PARAM_STR(needle)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_OPTIONAL
Z_PARAM_BOOL(part)
ZEND_PARSE_PARAMETERS_END();
style & ws fixes
2003-02-08 15:26:17 +00:00
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
found = php_memnstr(ZSTR_VAL(haystack), ZSTR_VAL(needle), ZSTR_LEN(needle), ZSTR_VAL(haystack) + ZSTR_LEN(haystack));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (found) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
found_offset = found - ZSTR_VAL(haystack);
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
if (part) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_STRINGL(ZSTR_VAL(haystack), found_offset);
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_STRINGL(found, ZSTR_LEN(haystack) - found_offset);
MFH: New parameter 'before_needle'
2008-02-01 12:28:44 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Fixed reference problem (variable were converted)
2008-02-03 14:30:25 +00:00
RETURN_FALSE;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Checks if a string contains another */
Add str_contains() function RFC: https://wiki.php.net/rfc/str_contains Closes GH-5179.
2020-02-13 15:23:01 +00:00
PHP_FUNCTION(str_contains)
{
zend_string *haystack, *needle;
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(haystack)
Z_PARAM_STR(needle)
ZEND_PARSE_PARAMETERS_END();
RETURN_BOOL(php_memnstr(ZSTR_VAL(haystack), ZSTR_VAL(needle), ZSTR_LEN(needle), ZSTR_VAL(haystack) + ZSTR_LEN(haystack)));
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Checks if haystack starts with needle */
Add str_starts_with() and str_ends_with() RFC: https://wiki.php.net/rfc/add_str_starts_with_and_ends_with_functions Closes GH-5300.
2020-05-05 13:57:39 +00:00
PHP_FUNCTION(str_starts_with)
{
zend_string *haystack, *needle;
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(haystack)
Z_PARAM_STR(needle)
ZEND_PARSE_PARAMETERS_END();
if (ZSTR_LEN(needle) > ZSTR_LEN(haystack)) {
RETURN_FALSE;
}
RETURN_BOOL(memcmp(ZSTR_VAL(haystack), ZSTR_VAL(needle), ZSTR_LEN(needle)) == 0);
}
Folder marks
2020-05-07 15:03:25 +00:00
/* }}} */
Add str_starts_with() and str_ends_with() RFC: https://wiki.php.net/rfc/add_str_starts_with_and_ends_with_functions Closes GH-5300.
2020-05-05 13:57:39 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Checks if haystack ends with needle */
Add str_starts_with() and str_ends_with() RFC: https://wiki.php.net/rfc/add_str_starts_with_and_ends_with_functions Closes GH-5300.
2020-05-05 13:57:39 +00:00
PHP_FUNCTION(str_ends_with)
{
zend_string *haystack, *needle;
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(haystack)
Z_PARAM_STR(needle)
ZEND_PARSE_PARAMETERS_END();
if (ZSTR_LEN(needle) > ZSTR_LEN(haystack)) {
RETURN_FALSE;
}
RETURN_BOOL(memcmp(
ZSTR_VAL(haystack) + ZSTR_LEN(haystack) - ZSTR_LEN(needle),
ZSTR_VAL(needle), ZSTR_LEN(needle)) == 0);
}
Folder marks
2020-05-07 15:03:25 +00:00
/* }}} */
Add str_starts_with() and str_ends_with() RFC: https://wiki.php.net/rfc/add_str_starts_with_and_ends_with_functions Closes GH-5300.
2020-05-05 13:57:39 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ An alias for strstr */
More typos.
1999-08-09 20:52:58 +00:00
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds position of first occurrence of a string within another */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strpos)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_string *haystack, *needle;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *found = NULL;
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_long offset = 0;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
fix some cases with fast zpp
2014-08-16 12:00:02 +00:00
Z_PARAM_STR(haystack)
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
Z_PARAM_STR(needle)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
master renames phase 1
2014-08-25 17:24:55 +00:00
Z_PARAM_LONG(offset)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
strpos()/stripos(): Add support for negative offsets
2015-07-22 22:08:57 +00:00
if (offset < 0) {
offset += (zend_long)ZSTR_LEN(haystack);
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (offset < 0 || (size_t)offset > ZSTR_LEN(haystack)) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
}
@- strpos() is now binary-safe. (Thies)
1999-12-06 17:37:59 +00:00
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
found = (char*)php_memnstr(ZSTR_VAL(haystack) + offset,
ZSTR_VAL(needle), ZSTR_LEN(needle),
ZSTR_VAL(haystack) + ZSTR_LEN(haystack));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (found) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(found - ZSTR_VAL(haystack));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
} else {
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
RETURN_FALSE;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds position of first occurrence of a string within another, case insensitive */
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
PHP_FUNCTION(stripos)
{
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *found = NULL;
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_string *haystack, *needle;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long offset = 0;
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
zend_string *needle_dup = NULL, *haystack_dup;
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Z_PARAM_STR(haystack)
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
Z_PARAM_STR(needle)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_OPTIONAL
Z_PARAM_LONG(offset)
ZEND_PARSE_PARAMETERS_END();
style & ws fixes
2003-02-08 15:26:17 +00:00
strpos()/stripos(): Add support for negative offsets
2015-07-22 22:08:57 +00:00
if (offset < 0) {
offset += (zend_long)ZSTR_LEN(haystack);
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (offset < 0 || (size_t)offset > ZSTR_LEN(haystack)) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
}
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
if (ZSTR_LEN(needle) > ZSTR_LEN(haystack)) {
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
RETURN_FALSE;
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
}
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
haystack_dup = php_string_tolower(haystack);
needle_dup = php_string_tolower(needle);
found = (char*)php_memnstr(ZSTR_VAL(haystack_dup) + offset,
ZSTR_VAL(needle_dup), ZSTR_LEN(needle_dup), ZSTR_VAL(haystack_dup) + ZSTR_LEN(haystack));
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
if (found) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETVAL_LONG(found - ZSTR_VAL(haystack_dup));
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
} else {
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
RETVAL_FALSE;
}
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(haystack_dup, 0);
Remove unnecessary NULL check needle_dup cannot be NULL here.
2019-08-25 14:09:20 +00:00
zend_string_release_ex(needle_dup, 0);
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds position of last occurrence of a string within another string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strrpos)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
zend_string *needle;
fix some cases with fast zpp
2014-08-16 12:00:02 +00:00
zend_string *haystack;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long offset = 0;
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
const char *p, *e, *found;
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
fix some cases with fast zpp
2014-08-16 12:00:02 +00:00
Z_PARAM_STR(haystack)
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
Z_PARAM_STR(needle)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
master renames phase 1
2014-08-25 17:24:55 +00:00
Z_PARAM_LONG(offset)
Remove most uses of ZEND_PARSE_PARAMETERS_END_EX() As ZPP now throws, it makes no sense to specify an explicit return value.
2019-10-07 08:02:18 +00:00
ZEND_PARSE_PARAMETERS_END();
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
Rewrite strrpos and strripos to be binary safe and behave the way their prototype suggests (also like their counterparts strpos and stripos) i.e.: Allow string needles rather than just characters
2003-04-12 07:36:22 +00:00
if (offset >= 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if ((size_t)offset > ZSTR_LEN(haystack)) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Fix invalid memory access in strrpos().
2006-10-02 19:42:42 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(haystack) + (size_t)offset;
e = ZSTR_VAL(haystack) + ZSTR_LEN(haystack);
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
} else {
Promote warnings to exceptions in string search related functions GH-5004
2019-11-20 01:34:52 +00:00
if (offset < -ZEND_LONG_MAX || (size_t)(-offset) > ZSTR_LEN(haystack)) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Fix invalid memory access in strrpos().
2006-10-02 19:42:42 +00:00
}
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(haystack);
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
if ((size_t)-offset < ZSTR_LEN(needle)) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
e = ZSTR_VAL(haystack) + ZSTR_LEN(haystack);
Better fix
2015-01-19 02:38:25 +00:00
} else {
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
e = ZSTR_VAL(haystack) + ZSTR_LEN(haystack) + offset + ZSTR_LEN(needle);
Rewrite strrpos and strripos to be binary safe and behave the way their prototype suggests (also like their counterparts strpos and stripos) i.e.: Allow string needles rather than just characters
2003-04-12 07:36:22 +00:00
}
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
}
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
if ((found = zend_memnrstr(p, ZSTR_VAL(needle), ZSTR_LEN(needle), e))) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(found - ZSTR_VAL(haystack));
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
}
Rewrite strrpos and strripos to be binary safe and behave the way their prototype suggests (also like their counterparts strpos and stripos) i.e.: Allow string needles rather than just characters
2003-04-12 07:36:22 +00:00
RETURN_FALSE;
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds position of last occurrence of a string within another string */
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
PHP_FUNCTION(strripos)
{
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
zend_string *needle;
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *haystack;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long offset = 0;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *p, *e, *found;
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_string *needle_dup, *haystack_dup;
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Z_PARAM_STR(haystack)
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
Z_PARAM_STR(needle)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_OPTIONAL
Z_PARAM_LONG(offset)
Remove most uses of ZEND_PARSE_PARAMETERS_END_EX() As ZPP now throws, it makes no sense to specify an explicit return value.
2019-10-07 08:02:18 +00:00
ZEND_PARSE_PARAMETERS_END();
Added stripos() & strripos() functions. Added 3rd parameter (offset) to strrpos().
2003-01-11 17:52:38 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(needle) == 1) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* Single character search can shortcut memcmps
Optimize strrpos/strripos for single char strings and fix offset to report correctly
2003-12-03 01:31:56 +00:00
Can also avoid tolower emallocs */
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
char lowered;
Optimize strrpos/strripos for single char strings and fix offset to report correctly
2003-12-03 01:31:56 +00:00
if (offset >= 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if ((size_t)offset > ZSTR_LEN(haystack)) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
More offset fixes. Added E_NOTICE where appropriate.
2006-10-02 19:58:15 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(haystack) + (size_t)offset;
e = ZSTR_VAL(haystack) + ZSTR_LEN(haystack) - 1;
Optimize strrpos/strripos for single char strings and fix offset to report correctly
2003-12-03 01:31:56 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(haystack);
Promote warnings to exceptions in string search related functions GH-5004
2019-11-20 01:34:52 +00:00
if (offset < -ZEND_LONG_MAX || (size_t)(-offset) > ZSTR_LEN(haystack)) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Fixed a segfault. (It's possible for large offsets to make strrpos() read past the end of the haystack string...)
2004-05-06 16:11:50 +00:00
}
Avoid address calculation overflow
2019-06-20 11:06:56 +00:00
e = ZSTR_VAL(haystack) + (ZSTR_LEN(haystack) + (size_t)offset);
Optimize strrpos/strripos for single char strings and fix offset to report correctly
2003-12-03 01:31:56 +00:00
}
/* Borrow that ord_needle buffer to avoid repeatedly tolower()ing needle */
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
lowered = tolower(*ZSTR_VAL(needle));
Optimize strrpos/strripos for single char strings and fix offset to report correctly
2003-12-03 01:31:56 +00:00
while (e >= p) {
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
if (tolower(*e) == lowered) {
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(e - p + (offset > 0 ? offset : 0));
Optimize strrpos/strripos for single char strings and fix offset to report correctly
2003-12-03 01:31:56 +00:00
}
e--;
}
RETURN_FALSE;
}
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
haystack_dup = php_string_tolower(haystack);
Rewrite strrpos and strripos to be binary safe and behave the way their prototype suggests (also like their counterparts strpos and stripos) i.e.: Allow string needles rather than just characters
2003-04-12 07:36:22 +00:00
if (offset >= 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if ((size_t)offset > ZSTR_LEN(haystack)) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(haystack_dup, 0);
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
More offset fixes. Added E_NOTICE where appropriate.
2006-10-02 19:58:15 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(haystack_dup) + offset;
e = ZSTR_VAL(haystack_dup) + ZSTR_LEN(haystack);
ws fixes
2002-10-10 18:41:24 +00:00
} else {
Promote warnings to exceptions in string search related functions GH-5004
2019-11-20 01:34:52 +00:00
if (offset < -ZEND_LONG_MAX || (size_t)(-offset) > ZSTR_LEN(haystack)) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(haystack_dup, 0);
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
More offset fixes. Added E_NOTICE where appropriate.
2006-10-02 19:58:15 +00:00
}
Improve strpos and strstr function family implementation
2019-08-18 13:33:10 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(haystack_dup);
Fixed compilation warnings
2016-06-23 09:47:06 +00:00
if ((size_t)-offset < ZSTR_LEN(needle)) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
e = ZSTR_VAL(haystack_dup) + ZSTR_LEN(haystack);
Better fix
2015-01-19 02:38:25 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
e = ZSTR_VAL(haystack_dup) + ZSTR_LEN(haystack) + offset + ZSTR_LEN(needle);
Rewrite strrpos and strripos to be binary safe and behave the way their prototype suggests (also like their counterparts strpos and stripos) i.e.: Allow string needles rather than just characters
2003-04-12 07:36:22 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
needle_dup = php_string_tolower(needle);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if ((found = (char *)zend_memnrstr(p, ZSTR_VAL(needle_dup), ZSTR_LEN(needle_dup), e))) {
RETVAL_LONG(found - ZSTR_VAL(haystack_dup));
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(needle_dup, 0);
zend_string_release_ex(haystack_dup, 0);
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
} else {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(needle_dup, 0);
zend_string_release_ex(haystack_dup, 0);
Faster strrpos implementation
2015-01-12 09:24:37 +00:00
RETURN_FALSE;
Rewrite strrpos and strripos to be binary safe and behave the way their prototype suggests (also like their counterparts strpos and stripos) i.e.: Allow string needles rather than just characters
2003-04-12 07:36:22 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Finds the last occurrence of a character in a string within another */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strrchr)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
zend_string *haystack, *needle;
Kill some more constness warnings in com_dotnet/phar/standard
2010-11-19 16:03:00 +00:00
const char *found = NULL;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long found_offset;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(haystack)
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
Z_PARAM_STR(needle)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Always treat needles as strings This is part of https://wiki.php.net/rfc/deprecations_php_7_3.
2019-01-29 09:02:03 +00:00
found = zend_memrchr(ZSTR_VAL(haystack), *ZSTR_VAL(needle), ZSTR_LEN(haystack));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (found) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
found_offset = found - ZSTR_VAL(haystack);
RETURN_STRINGL(found, ZSTR_LEN(haystack) - found_offset);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
} else {
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
RETURN_FALSE;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_chunk_split */
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
static zend_string *php_chunk_split(const char *src, size_t srclen, const char *end, size_t endlen, size_t chunklen)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
char *q;
const char *p;
Make string size calculation in chunk_split more precise The +1 on the string length is unnecessary, as we need the string length without trailing NUL byte here. The +1 on the chunks is only necessary if there is a rest. If the string devides into chunks exactly, we don't need an extra chunk. This makes the allocations exactly as large as it needs to be.
2019-08-25 13:32:33 +00:00
size_t chunks;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t restlen;
Refactor chunk_split returning zend_string to avoid memleak
2014-02-24 09:35:47 +00:00
zend_string *dest;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
chunks = srclen / chunklen;
restlen = srclen - chunks * chunklen; /* srclen % chunklen */
Make string size calculation in chunk_split more precise The +1 on the string length is unnecessary, as we need the string length without trailing NUL byte here. The +1 on the chunks is only necessary if there is a rest. If the string devides into chunks exactly, we don't need an extra chunk. This makes the allocations exactly as large as it needs to be.
2019-08-25 13:32:33 +00:00
if (restlen) {
/* We want chunks to be rounded up rather than rounded down.
* Increment can't overflow because chunks <= SIZE_MAX/2 at this point. */
chunks++;
Fix chunk_split fix - avoid using floats Fix money_format - don't give strfmon more arguments then supplied Fix str[c]spn integer overflow
2007-06-06 17:59:07 +00:00
}
Make string size calculation in chunk_split more precise The +1 on the string length is unnecessary, as we need the string length without trailing NUL byte here. The +1 on the chunks is only necessary if there is a rest. If the string devides into chunks exactly, we don't need an extra chunk. This makes the allocations exactly as large as it needs to be.
2019-08-25 13:32:33 +00:00
Remove memory exhaustion checks in php_chunk_split()
2019-08-23 23:52:17 +00:00
dest = zend_string_safe_alloc(chunks, endlen, srclen, 0);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (p = src, q = ZSTR_VAL(dest); p < (src + srclen - chunklen + 1); ) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
memcpy(q, p, chunklen);
q += chunklen;
memcpy(q, end, endlen);
q += endlen;
p += chunklen;
}
ws fixes.
2002-10-02 17:56:04 +00:00
if (restlen) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
memcpy(q, p, restlen);
q += restlen;
memcpy(q, end, endlen);
q += endlen;
}
*q = '\0';
Make string size calculation in chunk_split more precise The +1 on the string length is unnecessary, as we need the string length without trailing NUL byte here. The +1 on the chunks is only necessary if there is a rest. If the string devides into chunks exactly, we don't need an extra chunk. This makes the allocations exactly as large as it needs to be.
2019-08-25 13:32:33 +00:00
ZEND_ASSERT(q - ZSTR_VAL(dest) == ZSTR_LEN(dest));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Refactor chunk_split returning zend_string to avoid memleak
2014-02-24 09:35:47 +00:00
return dest;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns split line */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
PHP_FUNCTION(chunk_split)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
char *end = "\r\n";
's' works with size_t round 2
2014-08-27 15:31:24 +00:00
size_t endlen = 2;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long chunklen = 76;
Refactor chunk_split returning zend_string to avoid memleak
2014-02-24 09:35:47 +00:00
zend_string *result;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 3)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(chunklen)
Z_PARAM_STRING(end, endlen)
ZEND_PARSE_PARAMETERS_END();
style & ws fixes
2003-02-08 15:26:17 +00:00
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
if (chunklen <= 0) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(2, "must be greater than 0");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Fix chunk_split (#6208)
2000-08-17 08:47:42 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if ((size_t)chunklen > ZSTR_LEN(str)) {
Fixed bug #32530 (chunk_split() does not append endstr if chunklen is longer then the original string).
2005-04-03 18:08:40 +00:00
/* to maintain BC, we must return original string + ending */
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
result = zend_string_safe_alloc(ZSTR_LEN(str), 1, endlen, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(result), ZSTR_VAL(str), ZSTR_LEN(str));
memcpy(ZSTR_VAL(result) + ZSTR_LEN(str), end, endlen);
ZSTR_VAL(result)[ZSTR_LEN(result)] = '\0';
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(result);
Fixed bug #31479 (Fixed crash in chunk_split(), when chunklen > strlen).
2005-01-18 15:47:57 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(str)) {
Fix chunk_split (#6208)
2000-08-17 08:47:42 +00:00
RETURN_EMPTY_STRING();
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result = php_chunk_split(ZSTR_VAL(str), ZSTR_LEN(str), end, endlen, (size_t)chunklen);
style & ws fixes
2003-02-08 15:26:17 +00:00
Remove memory exhaustion checks in php_chunk_split()
2019-08-23 23:52:17 +00:00
RETURN_STR(result);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns part of a string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(substr)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long l = 0, f;
Accept null lengths for substr functions() If a null $length is passed to any of these functions, behave as if no parameter was passed: - substr() - substr_count() - substr_compare() - iconv_substr()
2019-10-21 18:49:08 +00:00
zend_bool len_is_null = 1;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
Z_PARAM_STR(str)
master renames phase 1
2014-08-25 17:24:55 +00:00
Z_PARAM_LONG(f)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
Accept null lengths for substr functions() If a null $length is passed to any of these functions, behave as if no parameter was passed: - substr() - substr_count() - substr_compare() - iconv_substr()
2019-10-21 18:49:08 +00:00
Z_PARAM_LONG_OR_NULL(l, len_is_null)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
style & ws fixes
2003-02-08 15:26:17 +00:00
Normalize substr() behavior Make the behavior of substr(), mb_substr(), iconv_substr() and grapheme_substr() consistent when it comes to the handling of out of bounds offsets. substr() will now always clamp out of bounds offsets to the string boundary. Cases that previously returned false will now return an empty string. This means that substr() itself *always* returns a string now (like mb_substr() already did before.) Closes GH-6182.
2020-09-22 07:56:08 +00:00
if (f < 0) {
Optimize substr() edge-case conditions
2018-10-29 10:11:41 +00:00
/* if "from" position is negative, count start position from the end
* of the string
*/
if ((size_t)-f > ZSTR_LEN(str)) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
f = 0;
Optimize substr() edge-case conditions
2018-10-29 10:11:41 +00:00
} else {
f = (zend_long)ZSTR_LEN(str) + f;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Normalize substr() behavior Make the behavior of substr(), mb_substr(), iconv_substr() and grapheme_substr() consistent when it comes to the handling of out of bounds offsets. substr() will now always clamp out of bounds offsets to the string boundary. Cases that previously returned false will now return an empty string. This means that substr() itself *always* returns a string now (like mb_substr() already did before.) Closes GH-6182.
2020-09-22 07:56:08 +00:00
} else if ((size_t)f > ZSTR_LEN(str)) {
RETURN_EMPTY_STRING();
}
if (!len_is_null) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (l < 0) {
Optimize substr() edge-case conditions
2018-10-29 10:11:41 +00:00
/* if "length" position is negative, set it to the length
* needed to stop that many chars from the end of the string
*/
if ((size_t)(-l) > ZSTR_LEN(str) - (size_t)f) {
Normalize substr() behavior Make the behavior of substr(), mb_substr(), iconv_substr() and grapheme_substr() consistent when it comes to the handling of out of bounds offsets. substr() will now always clamp out of bounds offsets to the string boundary. Cases that previously returned false will now return an empty string. This means that substr() itself *always* returns a string now (like mb_substr() already did before.) Closes GH-6182.
2020-09-22 07:56:08 +00:00
l = 0;
Optimize substr() edge-case conditions
2018-10-29 10:11:41 +00:00
} else {
l = (zend_long)ZSTR_LEN(str) - f + l;
}
} else if ((size_t)l > ZSTR_LEN(str) - (size_t)f) {
Normalize substr() behavior Make the behavior of substr(), mb_substr(), iconv_substr() and grapheme_substr() consistent when it comes to the handling of out of bounds offsets. substr() will now always clamp out of bounds offsets to the string boundary. Cases that previously returned false will now return an empty string. This means that substr() itself *always* returns a string now (like mb_substr() already did before.) Closes GH-6182.
2020-09-22 07:56:08 +00:00
l = (zend_long)ZSTR_LEN(str) - f;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Optimize substr() edge-case conditions
2018-10-29 10:11:41 +00:00
} else {
l = (zend_long)ZSTR_LEN(str) - f;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
Add helper APIs for maybe-interned string creation Add ZVAL_CHAR/RETVAL_CHAR/RETURN_CHAR as a shortcut for using ZVAL_INTERNED_STRING and ZSTR_CHAR. Add zend_string_init_fast() as a helper for the empty string / one char interned string / zend_string_init() pattern. Also add corresponding ZVAL_STRINGL_FAST etc macros. Closes GH-5684.
2020-06-08 10:45:01 +00:00
if (l == ZSTR_LEN(str)) {
Use interned empty and "one char" strings.
2017-05-18 14:59:59 +00:00
RETURN_STR_COPY(str);
Add helper APIs for maybe-interned string creation Add ZVAL_CHAR/RETVAL_CHAR/RETURN_CHAR as a shortcut for using ZVAL_INTERNED_STRING and ZSTR_CHAR. Add zend_string_init_fast() as a helper for the empty string / one char interned string / zend_string_init() pattern. Also add corresponding ZVAL_STRINGL_FAST etc macros. Closes GH-5684.
2020-06-08 10:45:01 +00:00
} else {
RETURN_STRINGL_FAST(ZSTR_VAL(str) + f, l);
Use interned empty and "one char" strings.
2017-05-18 14:59:59 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Replaces part of a string with another string */
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
PHP_FUNCTION(substr_replace)
{
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
zend_string *str, *repl_str;
HashTable *str_ht, *repl_ht;
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
HashTable *from_ht;
zend_long from_long;
HashTable *len_ht = NULL;
zend_long len_long;
zend_bool len_is_null = 1;
remove comment f and l are the $start and $length, since they come from the user land, they only can be zend_long
2015-06-29 08:56:51 +00:00
zend_long l = 0;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long f;
Fixed memory leak in substr_replace
2014-02-24 07:58:38 +00:00
zend_string *result;
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
HashPosition from_idx, repl_idx, len_idx;
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
zval *tmp_str = NULL, *tmp_repl, *tmp_from = NULL, *tmp_len= NULL;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(3, 4)
Consolidate new union type ZPP macro names They will now follow the canonical order of types. Older macros are left intact due to maintaining BC. Closes GH-6112
2020-09-10 22:23:54 +00:00
Z_PARAM_ARRAY_HT_OR_STR(str_ht, str)
Z_PARAM_ARRAY_HT_OR_STR(repl_ht, repl_str)
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
Z_PARAM_ARRAY_HT_OR_LONG(from_ht, from_long)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_OPTIONAL
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
Z_PARAM_ARRAY_HT_OR_LONG_OR_NULL(len_ht, len_long, len_is_null)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
if (len_is_null) {
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
if (str) {
l = ZSTR_LEN(str);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
} else if (!len_ht) {
l = len_long;
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
}
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
if (str) {
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
if (from_ht) {
zend_argument_type_error(3, "cannot be an array when working on a single string");
RETURN_THROWS();
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
if (len_ht) {
zend_argument_type_error(4, "cannot be an array when working on a single string");
RETURN_THROWS();
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
f = from_long;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
/* if "from" position is negative, count start position from the end
* of the string
*/
if (f < 0) {
f = (zend_long)ZSTR_LEN(str) + f;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
if (f < 0) {
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
f = 0;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
} else if ((size_t)f > ZSTR_LEN(str)) {
f = ZSTR_LEN(str);
}
/* if "length" position is negative, set it to the length
* needed to stop that many chars from the end of the string
*/
if (l < 0) {
l = ((zend_long)ZSTR_LEN(str) - f) + l;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
if (l < 0) {
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
l = 0;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
}
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
if ((size_t)l > ZSTR_LEN(str) || (l < 0 && (size_t)(-l) > ZSTR_LEN(str))) {
l = ZSTR_LEN(str);
}
Fixed bug #40754 (added substr() & substr_replace() overflow checks).
2007-03-09 01:58:34 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
if ((f + l) > (zend_long)ZSTR_LEN(str)) {
l = ZSTR_LEN(str) - f;
}
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
zend_string *tmp_repl_str = NULL;
if (repl_ht) {
repl_idx = 0;
while (repl_idx < repl_ht->nNumUsed) {
tmp_repl = &repl_ht->arData[repl_idx].val;
if (Z_TYPE_P(tmp_repl) != IS_UNDEF) {
break;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
repl_idx++;
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
}
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
if (repl_idx < repl_ht->nNumUsed) {
repl_str = zval_get_tmp_string(tmp_repl, &tmp_repl_str);
} else {
repl_str = STR_EMPTY_ALLOC();
}
}
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
result = zend_string_safe_alloc(1, ZSTR_LEN(str) - l + ZSTR_LEN(repl_str), 0, 0);
Fixed memory leak in substr_replace
2014-02-24 07:58:38 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
memcpy(ZSTR_VAL(result), ZSTR_VAL(str), f);
if (ZSTR_LEN(repl_str)) {
memcpy((ZSTR_VAL(result) + f), ZSTR_VAL(repl_str), ZSTR_LEN(repl_str));
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
}
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
memcpy((ZSTR_VAL(result) + f + ZSTR_LEN(repl_str)), ZSTR_VAL(str) + f + l, ZSTR_LEN(str) - f - l);
ZSTR_VAL(result)[ZSTR_LEN(result)] = '\0';
zend_tmp_string_release(tmp_repl_str);
RETURN_NEW_STR(result);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} else { /* str is array of strings */
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zend_string *str_index = NULL;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t result_len;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_ulong num_index;
Fixed bug #55366: keys lost when using substr_replace an array
2011-08-29 20:38:52 +00:00
Promote substr_replace warnings The implementation here was pretty confused. In reality the only error condition it has right now is that for a string input, from & length cannot be arrays. The fact that the array lengths are the same was probably supposed to be checked for the case of array input, as it wouldn't matter otherwise.
2020-09-22 08:22:43 +00:00
/* TODO
if (!len_is_null && from_ht) {
if (zend_hash_num_elements(from_ht) != zend_hash_num_elements(len_ht)) {
php_error_docref(NULL, E_WARNING, "'start' and 'length' should have the same number of elements");
RETURN_STR_COPY(str);
}
}
*/
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
array_init(return_value);
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
from_idx = len_idx = repl_idx = 0;
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
Merge branch 'PHP-7.4' * PHP-7.4: Fixed bug #79817
2020-07-10 12:07:59 +00:00
ZEND_HASH_FOREACH_KEY_VAL_IND(str_ht, num_index, str_index, tmp_str) {
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_string *tmp_orig_str;
zend_string *orig_str = zval_get_tmp_string(tmp_str, &tmp_orig_str);
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
if (from_ht) {
while (from_idx < from_ht->nNumUsed) {
tmp_from = &from_ht->arData[from_idx].val;
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
if (Z_TYPE_P(tmp_from) != IS_UNDEF) {
break;
}
from_idx++;
}
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
if (from_idx < from_ht->nNumUsed) {
master renames phase 1
2014-08-25 17:24:55 +00:00
f = zval_get_long(tmp_from);
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
if (f < 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
f = (zend_long)ZSTR_LEN(orig_str) + f;
refix the negative zend_long to size_t casts There is no good way to fix this for 32-bit without enormously overcomplicating the logic. Therefore switching back to the previous code and adding the casts to ensure there are no sudden casts of negative to size_t.
2015-06-29 18:33:34 +00:00
if (f < 0) {
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
f = 0;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
} else if (f > (zend_long)ZSTR_LEN(orig_str)) {
f = ZSTR_LEN(orig_str);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
from_idx++;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} else {
f = 0;
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
}
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} else {
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
f = from_long;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
if (f < 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
f = (zend_long)ZSTR_LEN(orig_str) + f;
refix the negative zend_long to size_t casts There is no good way to fix this for 32-bit without enormously overcomplicating the logic. Therefore switching back to the previous code and adding the casts to ensure there are no sudden casts of negative to size_t.
2015-06-29 18:33:34 +00:00
if (f < 0) {
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
f = 0;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
} else if (f > (zend_long)ZSTR_LEN(orig_str)) {
f = ZSTR_LEN(orig_str);
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
}
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
if (len_ht) {
while (len_idx < len_ht->nNumUsed) {
tmp_len = &len_ht->arData[len_idx].val;
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
if (Z_TYPE_P(tmp_len) != IS_UNDEF) {
break;
}
len_idx++;
}
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
if (len_idx < len_ht->nNumUsed) {
master renames phase 1
2014-08-25 17:24:55 +00:00
l = zval_get_long(tmp_len);
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
len_idx++;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
l = ZSTR_LEN(orig_str);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Declare array|int and object-of-class|int types in stubs Closes GH-6081 Co-Authored-By: Nikita Popov <nikic@php.net>
2020-09-14 09:59:32 +00:00
} else if (!len_is_null) {
l = len_long;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
l = ZSTR_LEN(orig_str);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
if (l < 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
l = (ZSTR_LEN(orig_str) - f) + l;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
if (l < 0) {
l = 0;
}
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
}
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if ((f + l) > (zend_long)ZSTR_LEN(orig_str)) {
l = ZSTR_LEN(orig_str) - f;
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result_len = ZSTR_LEN(orig_str) - l;
Optimize substr_replace (this makes it 3-8 times faster depending on the string length).
2003-12-01 23:17:22 +00:00
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
if (repl_ht) {
while (repl_idx < repl_ht->nNumUsed) {
tmp_repl = &repl_ht->arData[repl_idx].val;
if (repl_ht != IS_UNDEF) {
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
break;
}
repl_idx++;
}
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
if (repl_idx < repl_ht->nNumUsed) {
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_string *tmp_repl_str;
zend_string *repl_str = zval_get_tmp_string(tmp_repl, &tmp_repl_str);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result_len += ZSTR_LEN(repl_str);
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
repl_idx++;
pick up the safe alloc pieces from 19866fb76cf4c95d904ebb0e08592cf38303fae9
2016-09-13 09:50:18 +00:00
result = zend_string_safe_alloc(1, result_len, 0, 0);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(result), ZSTR_VAL(orig_str), f);
memcpy((ZSTR_VAL(result) + f), ZSTR_VAL(repl_str), ZSTR_LEN(repl_str));
memcpy((ZSTR_VAL(result) + f + ZSTR_LEN(repl_str)), ZSTR_VAL(orig_str) + f + l, ZSTR_LEN(orig_str) - f - l);
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_tmp_string_release(tmp_repl_str);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} else {
pick up the safe alloc pieces from 19866fb76cf4c95d904ebb0e08592cf38303fae9
2016-09-13 09:50:18 +00:00
result = zend_string_safe_alloc(1, result_len, 0, 0);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(result), ZSTR_VAL(orig_str), f);
memcpy((ZSTR_VAL(result) + f), ZSTR_VAL(orig_str) + f + l, ZSTR_LEN(orig_str) - f - l);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
} else {
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
result_len += ZSTR_LEN(repl_str);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
pick up the safe alloc pieces from 19866fb76cf4c95d904ebb0e08592cf38303fae9
2016-09-13 09:50:18 +00:00
result = zend_string_safe_alloc(1, result_len, 0, 0);
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(result), ZSTR_VAL(orig_str), f);
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
memcpy((ZSTR_VAL(result) + f), ZSTR_VAL(repl_str), ZSTR_LEN(repl_str));
memcpy((ZSTR_VAL(result) + f + ZSTR_LEN(repl_str)), ZSTR_VAL(orig_str) + f + l, ZSTR_LEN(orig_str) - f - l);
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[ZSTR_LEN(result)] = '\0';
Fixed bug #55366: keys lost when using substr_replace an array
2011-08-29 20:38:52 +00:00
Use ZEND_HASH_FOREACH_* macros
2014-04-21 10:20:05 +00:00
if (str_index) {
zval tmp;
Avoid double IS_INTERNED() check
2014-09-19 13:32:50 +00:00
ZVAL_NEW_STR(&tmp, result);
Use ZEND_HASH_FOREACH_* macros
2014-04-21 10:20:05 +00:00
zend_symtable_update(Z_ARRVAL_P(return_value), str_index, &tmp);
Fixed bug #55366: keys lost when using substr_replace an array
2011-08-29 20:38:52 +00:00
} else {
Fixed memory leak in substr_replace
2014-02-24 07:58:38 +00:00
add_index_str(return_value, num_index, result);
Fixed bug #55366: keys lost when using substr_replace an array
2011-08-29 20:38:52 +00:00
}
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_tmp_string_release(tmp_orig_str);
Use ZEND_HASH_FOREACH_* macros
2014-04-21 10:20:05 +00:00
} ZEND_HASH_FOREACH_END();
Extending substr_replace(). Every parameter can be mixed (string/array) also the return type. #see the test file if you need more info.
2003-05-02 10:02:40 +00:00
} /* if */
(PHP substr_replace) New function. @- Added substr_replace() function. (Andrei)
1999-11-16 22:44:13 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Quotes meta characters */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(quotemeta)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *old;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *old_end, *p;
char *q;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
char c;
Fixed memory leak in ext/standard/tests/strings/quotemeta_basic.php
2014-02-25 05:08:16 +00:00
zend_string *str;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(old)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
old_end = ZSTR_VAL(old) + ZSTR_LEN(old);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Return empty str from quotemeta() on empty str
2019-02-11 08:58:10 +00:00
if (ZSTR_LEN(old) == 0) {
RETURN_EMPTY_STRING();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
str = zend_string_safe_alloc(2, ZSTR_LEN(old), 0, 0);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (p = ZSTR_VAL(old), q = ZSTR_VAL(str); p != old_end; p++) {
Made quotemeta() binary-safe. @ Made quotemeta() binary-safe. (Andrei)
1999-12-20 20:43:47 +00:00
c = *p;
ws fixes.
2002-10-02 17:56:04 +00:00
switch (c) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
case '.':
case '\\':
case '+':
case '*':
case '?':
case '[':
case '^':
case ']':
case '$':
case '(':
case ')':
*q++ = '\\';
/* break is missing _intentionally_ */
default:
*q++ = c;
}
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
Fixed memory leak in ext/standard/tests/strings/quotemeta_basic.php
2014-02-25 05:08:16 +00:00
*q = '\0';
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_NEW_STR(zend_string_truncate(str, q - ZSTR_VAL(str), 0));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns ASCII value of character
Added comments about special functions
2016-02-19 12:44:11 +00:00
Warning: This function is special-cased by zend_compile.c and so is bypassed for constant string argument */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(ord)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
change argument parsing
2018-02-17 14:29:59 +00:00
zend_string *str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
change argument parsing
2018-02-17 14:29:59 +00:00
Z_PARAM_STR(str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
change argument parsing
2018-02-17 14:29:59 +00:00
RETURN_LONG((unsigned char) ZSTR_VAL(str)[0]);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Converts ASCII code to a character
Added comments about special functions
2016-02-19 12:44:11 +00:00
Warning: This function is special-cased by zend_compile.c and so is bypassed for constant integer argument */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(chr)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long c;
use new param parsing API
2008-01-22 01:34:24 +00:00
Make chr ZPP failure message consistent with ext/standard
2019-05-08 19:42:02 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Improved chr() function. Use FAST_ZPP, use interned strings.
2014-12-19 02:05:30 +00:00
Z_PARAM_LONG(c)
Update MAY_BE_NULL info for more ext/standard functions
2019-05-28 11:03:41 +00:00
ZEND_PARSE_PARAMETERS_END();
Adjust new chr() param handling to address chr("") calls
2008-01-25 01:31:10 +00:00
Improved chr() function. Use FAST_ZPP, use interned strings.
2014-12-19 02:05:30 +00:00
c &= 0xff;
Add helper APIs for maybe-interned string creation Add ZVAL_CHAR/RETVAL_CHAR/RETURN_CHAR as a shortcut for using ZVAL_INTERNED_STRING and ZSTR_CHAR. Add zend_string_init_fast() as a helper for the empty string / one char interned string / zend_string_init() pattern. Also add corresponding ZVAL_STRINGL_FAST etc macros. Closes GH-5684.
2020-06-08 10:45:01 +00:00
RETURN_CHAR(c);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
- MFH - Making sure this is also using the good zend parsing and added the php_ucfirst function
2008-01-16 03:11:26 +00:00
/* {{{ php_ucfirst
Uppercase the first character of the word in a native string */
Optimized ucfirst(Avoid duping string if possible)
2017-08-06 06:56:16 +00:00
static zend_string* php_ucfirst(zend_string *str)
- MFH - Making sure this is also using the good zend parsing and added the php_ucfirst function
2008-01-16 03:11:26 +00:00
{
Fixed bug #79849 Closes GH-5853.
2020-07-13 12:16:39 +00:00
const unsigned char ch = ZSTR_VAL(str)[0];
unsigned char r = toupper(ch);
if (r == ch) {
Optimized ucfirst(Avoid duping string if possible)
2017-08-06 06:56:16 +00:00
return zend_string_copy(str);
} else {
zend_string *s = zend_string_init(ZSTR_VAL(str), ZSTR_LEN(str), 0);
ZSTR_VAL(s)[0] = r;
return s;
}
- MFH - Making sure this is also using the good zend parsing and added the php_ucfirst function
2008-01-16 03:11:26 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Makes a string's first character uppercase */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(ucfirst)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
fixes to string functions
2014-08-18 13:39:38 +00:00
Z_PARAM_STR(str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(str)) {
fixed return types for ucwords/ucfirst when passed an empty string
2002-06-17 11:37:49 +00:00
RETURN_EMPTY_STRING();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
@- ucfirst()/ucwords() no longer modify arg1. (Thies) (PHP ucfirst,ucwords) no longer modify arg1
1999-11-14 17:20:56 +00:00
Optimized ucfirst(Avoid duping string if possible)
2017-08-06 06:56:16 +00:00
RETURN_STR(php_ucfirst(str));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
/* {{{
Lowercase the first character of the word in a native string */
Optimized ucfirst(Avoid duping string if possible)
2017-08-06 06:56:16 +00:00
static zend_string* php_lcfirst(zend_string *str)
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
{
Optimized ucfirst(Avoid duping string if possible)
2017-08-06 06:56:16 +00:00
unsigned char r = tolower(ZSTR_VAL(str)[0]);
if (r == ZSTR_VAL(str)[0]) {
return zend_string_copy(str);
} else {
zend_string *s = zend_string_init(ZSTR_VAL(str), ZSTR_LEN(str), 0);
ZSTR_VAL(s)[0] = r;
return s;
}
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Make a string's first character lowercase */
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
PHP_FUNCTION(lcfirst)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(str)
ZEND_PARSE_PARAMETERS_END();
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(str)) {
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
RETURN_EMPTY_STRING();
}
Optimized ucfirst(Avoid duping string if possible)
2017-08-06 06:56:16 +00:00
RETURN_STR(php_lcfirst(str));
- MFH (lcfirst()) - Initial test for lcfirst
2008-01-19 19:27:22 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Uppercase the first character of every word in a string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(ucwords)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
char *delims = " \t\r\n\f\v";
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
register char *r;
register const char *r_end;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t delims_len = 6;
Fix for bug #34407 - ucwords and title case Added support for ranges like trim() has
2014-07-02 14:22:11 +00:00
char mask[256];
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Merge branch 'master' into phpng * master: (40 commits) Bug #67609: TLS connections fail behind HTTP proxy Updated NEWS for #67594 Updated NEWS for #67594 Fix #67594 - invisible colon should be stripped off header name Updated NEWS for 34407 Updated NEWS for 34407 Updated NEWS for 34407 Fix for bug #34407 - ucwords and title case fixed broken merged code Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. fixed broken merged code Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. Test output relies on expose_php being on 1.2 is a problematic float to print out Lower the default display precision for this test The test output is dependent on expose_php ini fix makefile in phpize mode fixe output_as_table() when no ext was enabled fix end of stream exception when generating makefile ... Conflicts: ext/standard/http_fopen_wrapper.c ext/standard/string.c sapi/cli/php_cli_server.c
2014-07-14 07:44:53 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
fixes to string functions
2014-08-18 13:39:38 +00:00
Z_PARAM_STR(str)
Merge branch 'master' into phpng * master: (40 commits) Bug #67609: TLS connections fail behind HTTP proxy Updated NEWS for #67594 Updated NEWS for #67594 Fix #67594 - invisible colon should be stripped off header name Updated NEWS for 34407 Updated NEWS for 34407 Updated NEWS for 34407 Fix for bug #34407 - ucwords and title case fixed broken merged code Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. fixed broken merged code Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. Fixed a bug that cannot access custom request header stored in apache_request_headers() though array index. Test output relies on expose_php being on 1.2 is a problematic float to print out Lower the default display precision for this test The test output is dependent on expose_php ini fix makefile in phpize mode fixe output_as_table() when no ext was enabled fix end of stream exception when generating makefile ... Conflicts: ext/standard/http_fopen_wrapper.c ext/standard/string.c sapi/cli/php_cli_server.c
2014-07-14 07:44:53 +00:00
Z_PARAM_OPTIONAL
Z_PARAM_STRING(delims, delims_len)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(str)) {
fixed return types for ucwords/ucfirst when passed an empty string
2002-06-17 11:37:49 +00:00
RETURN_EMPTY_STRING();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
(ucwords) Another try to fix #4748.
2000-06-01 10:07:44 +00:00
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
php_charmask((const unsigned char *) delims, delims_len, mask);
Fix for bug #34407 - ucwords and title case Added support for ranges like trim() has
2014-07-02 14:22:11 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZVAL_STRINGL(return_value, ZSTR_VAL(str), ZSTR_LEN(str));
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
r = Z_STRVAL_P(return_value);
*r = toupper((unsigned char) *r);
master renames phase 1
2014-08-25 17:24:55 +00:00
for (r_end = r + Z_STRLEN_P(return_value) - 1; r < r_end; ) {
Fix for bug #34407 - ucwords and title case Added support for ranges like trim() has
2014-07-02 14:22:11 +00:00
if (mask[(unsigned char)*r++]) {
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
*r = toupper((unsigned char) *r);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strtr */
Move constants into read-only data segment
2017-12-14 19:14:36 +00:00
PHPAPI char *php_strtr(char *str, size_t len, const char *str_from, const char *str_to, size_t trlen)
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
{
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t i;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
if (UNEXPECTED(trlen < 1)) {
Change string->str where possible; string is basic a C++ class, so gdb can't look inside variables named 'string'.
2000-02-08 22:21:43 +00:00
return str;
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
} else if (trlen == 1) {
char ch_from = *str_from;
char ch_to = *str_to;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
for (i = 0; i < len; i++) {
if (str[i] == ch_from) {
str[i] = ch_to;
}
}
} else {
unsigned char xlat[256], j = 0;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
do { xlat[j] = j; } while (++j != 0);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
for (i = 0; i < trlen; i++) {
xlat[(size_t)(unsigned char) str_from[i]] = str_to[i];
}
Fix bug #73648 - integer overflow in substr
2017-01-16 20:11:50 +00:00
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
for (i = 0; i < len; i++) {
str[i] = xlat[(size_t)(unsigned char) str[i]];
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Change string->str where possible; string is basic a C++ class, so gdb can't look inside variables named 'string'.
2000-02-08 22:21:43 +00:00
return str;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strtr_ex */
Move constants into read-only data segment
2017-12-14 19:14:36 +00:00
static zend_string *php_strtr_ex(zend_string *str, const char *str_from, const char *str_to, size_t trlen)
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
{
zend_string *new_str = NULL;
size_t i;
if (UNEXPECTED(trlen < 1)) {
return zend_string_copy(str);
} else if (trlen == 1) {
char ch_from = *str_from;
char ch_to = *str_to;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (i = 0; i < ZSTR_LEN(str); i++) {
if (ZSTR_VAL(str)[i] == ch_from) {
new_str = zend_string_alloc(ZSTR_LEN(str), 0);
memcpy(ZSTR_VAL(new_str), ZSTR_VAL(str), i);
ZSTR_VAL(new_str)[i] = ch_to;
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
break;
}
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (; i < ZSTR_LEN(str); i++) {
ZSTR_VAL(new_str)[i] = (ZSTR_VAL(str)[i] != ch_from) ? ZSTR_VAL(str)[i] : ch_to;
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
}
} else {
unsigned char xlat[256], j = 0;
do { xlat[j] = j; } while (++j != 0);
for (i = 0; i < trlen; i++) {
xlat[(size_t)(unsigned char) str_from[i]] = str_to[i];
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (i = 0; i < ZSTR_LEN(str); i++) {
if (ZSTR_VAL(str)[i] != xlat[(size_t)(unsigned char) ZSTR_VAL(str)[i]]) {
new_str = zend_string_alloc(ZSTR_LEN(str), 0);
memcpy(ZSTR_VAL(new_str), ZSTR_VAL(str), i);
ZSTR_VAL(new_str)[i] = xlat[(size_t)(unsigned char) ZSTR_VAL(str)[i]];
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
break;
}
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (;i < ZSTR_LEN(str); i++) {
ZSTR_VAL(new_str)[i] = xlat[(size_t)(unsigned char) ZSTR_VAL(str)[i]];
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
}
}
if (!new_str) {
return zend_string_copy(str);
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(new_str)[ZSTR_LEN(new_str)] = 0;
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
return new_str;
}
/* }}} */
Reimplement strtr()
2014-04-24 15:14:29 +00:00
/* {{{ php_strtr_array */
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
static void php_strtr_array(zval *return_value, zend_string *input, HashTable *pats)
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
{
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *str = ZSTR_VAL(input);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
size_t slen = ZSTR_LEN(input);
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_ulong num_key;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
zend_string *str_key;
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
size_t len, pos, old_pos;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
int num_keys = 0;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t minlen = 128*1024;
size_t maxlen = 0;
Improve strtr()
2015-05-15 01:03:30 +00:00
HashTable str_hash;
Remove unused var
2015-06-20 02:37:53 +00:00
zval *entry;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *key;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
smart_str result = {0};
Ipmrove strtr() by maintaining a set of characters that may start a matched pattern and avoid zend_hash_find() calls for other paterns.
2015-01-14 12:04:04 +00:00
zend_ulong bitset[256/sizeof(zend_ulong)];
Improve strtr()
2015-05-15 01:03:30 +00:00
zend_ulong *num_bitset;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
typo fixes - https://github.com/vlajos/misspell_fixer
2014-11-19 19:59:31 +00:00
/* we will collect all possible key lengths */
Fixed bug #70667 (strtr() causes invalid writes and a crashes)
2015-10-08 11:30:43 +00:00
num_bitset = ecalloc((slen + sizeof(zend_ulong)) / sizeof(zend_ulong), sizeof(zend_ulong));
Ipmrove strtr() by maintaining a set of characters that may start a matched pattern and avoid zend_hash_find() calls for other paterns.
2015-01-14 12:04:04 +00:00
memset(bitset, 0, sizeof(bitset));
Reimplement strtr()
2014-04-24 15:14:29 +00:00
/* check if original array has numeric keys */
Improve strtr()
2015-05-15 01:03:30 +00:00
ZEND_HASH_FOREACH_STR_KEY(pats, str_key) {
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (UNEXPECTED(!str_key)) {
num_keys = 1;
Refactoring, bugs & leaks
2013-01-08 23:33:14 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
len = ZSTR_LEN(str_key);
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (UNEXPECTED(len < 1)) {
Warn on strtr(["" => "x"]) Previously: * If only ["" => "x"] was present, the original string was returned without warning. * If both ["" => "x"] and at least one more element was present, false was returned without warning. New behavior: * Ignore "" keys in the replacement array (and perform any remaining replacement). * Throw a warning indicating that an empty string replacement has been ignored. Closes GH-4792.
2019-10-07 11:04:06 +00:00
php_error_docref(NULL, E_WARNING, "Ignoring replacement of empty string");
continue;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
} else if (UNEXPECTED(len > slen)) {
/* skip long patterns */
continue;
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
}
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (len > maxlen) {
maxlen = len;
}
if (len < minlen) {
minlen = len;
}
typo fixes - https://github.com/vlajos/misspell_fixer
2014-11-19 19:59:31 +00:00
/* remember possible key length */
Improve strtr()
2015-05-15 01:03:30 +00:00
num_bitset[len / sizeof(zend_ulong)] |= Z_UL(1) << (len % sizeof(zend_ulong));
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
bitset[((unsigned char)ZSTR_VAL(str_key)[0]) / sizeof(zend_ulong)] |= Z_UL(1) << (((unsigned char)ZSTR_VAL(str_key)[0]) % sizeof(zend_ulong));
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
}
Reimplement strtr()
2014-04-24 15:14:29 +00:00
} ZEND_HASH_FOREACH_END();
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
Improve strtr()
2015-05-15 01:03:30 +00:00
if (UNEXPECTED(num_keys)) {
Fixed bug #69872 (uninitialised value in strtr with array)
2015-06-18 14:28:29 +00:00
zend_string *key_used;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
/* we have to rebuild HashTable with numeric keys */
zend_hash_init(&str_hash, zend_hash_num_elements(pats), NULL, NULL, 0);
Fixed bug #79817 Use *_IND macros in a few places in string.c.
2020-07-10 12:06:41 +00:00
ZEND_HASH_FOREACH_KEY_VAL_IND(pats, num_key, str_key, entry) {
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (UNEXPECTED(!str_key)) {
Fixed bug #69872 (uninitialised value in strtr with array)
2015-06-18 14:28:29 +00:00
key_used = zend_long_to_str(num_key);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
len = ZSTR_LEN(key_used);
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (UNEXPECTED(len > slen)) {
/* skip long patterns */
Fixed bug #78612
2019-09-30 13:05:35 +00:00
zend_string_release(key_used);
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
continue;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
}
if (len > maxlen) {
maxlen = len;
}
if (len < minlen) {
minlen = len;
}
typo fixes - https://github.com/vlajos/misspell_fixer
2014-11-19 19:59:31 +00:00
/* remember possible key length */
Improve strtr()
2015-05-15 01:03:30 +00:00
num_bitset[len / sizeof(zend_ulong)] |= Z_UL(1) << (len % sizeof(zend_ulong));
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
bitset[((unsigned char)ZSTR_VAL(key_used)[0]) / sizeof(zend_ulong)] |= Z_UL(1) << (((unsigned char)ZSTR_VAL(key_used)[0]) % sizeof(zend_ulong));
Reimplement strtr()
2014-04-24 15:14:29 +00:00
} else {
Fixed bug #69872 (uninitialised value in strtr with array)
2015-06-18 14:28:29 +00:00
key_used = str_key;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
len = ZSTR_LEN(key_used);
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (UNEXPECTED(len > slen)) {
/* skip long patterns */
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
continue;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
}
}
Fixed bug #69872 (uninitialised value in strtr with array)
2015-06-18 14:28:29 +00:00
zend_hash_add(&str_hash, key_used, entry);
if (UNEXPECTED(!str_key)) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(key_used, 0);
Reimplement strtr()
2014-04-24 15:14:29 +00:00
}
} ZEND_HASH_FOREACH_END();
pats = &str_hash;
}
if (UNEXPECTED(minlen > maxlen)) {
/* return the original string */
if (pats == &str_hash) {
zend_hash_destroy(&str_hash);
}
Improve strtr()
2015-05-15 01:03:30 +00:00
efree(num_bitset);
RETURN_STR_COPY(input);
}
old_pos = pos = 0;
while (pos <= slen - minlen) {
key = str + pos;
Improve strtr (Fixed a bug which the lenth check doesn't work at all)
2015-06-19 14:59:31 +00:00
if (bitset[((unsigned char)key[0]) / sizeof(zend_ulong)] & (Z_UL(1) << (((unsigned char)key[0]) % sizeof(zend_ulong)))) {
Improve strtr()
2015-05-15 01:03:30 +00:00
len = maxlen;
if (len > slen - pos) {
len = slen - pos;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
}
Improve strtr()
2015-05-15 01:03:30 +00:00
while (len >= minlen) {
Improve strtr (Fixed a bug which the lenth check doesn't work at all)
2015-06-19 14:59:31 +00:00
if ((num_bitset[len / sizeof(zend_ulong)] & (Z_UL(1) << (len % sizeof(zend_ulong))))) {
entry = zend_hash_str_find(pats, key, len);
if (entry != NULL) {
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_string *tmp;
zend_string *s = zval_get_tmp_string(entry, &tmp);
Improve strtr (Fixed a bug which the lenth check doesn't work at all)
2015-06-19 14:59:31 +00:00
smart_str_appendl(&result, str + old_pos, pos - old_pos);
smart_str_append(&result, s);
old_pos = pos + len;
pos = old_pos - 1;
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_tmp_string_release(tmp);
Improve strtr (Fixed a bug which the lenth check doesn't work at all)
2015-06-19 14:59:31 +00:00
break;
}
trailing whitespace removal
2015-01-03 09:22:58 +00:00
}
Improve strtr()
2015-05-15 01:03:30 +00:00
len--;
Reimplement strtr()
2014-04-24 15:14:29 +00:00
}
}
Improve strtr()
2015-05-15 01:03:30 +00:00
pos++;
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
}
Improve strtr()
2015-05-15 01:03:30 +00:00
Fixed bug #69144 (strtr not replacing with partly matching replace pairs)
2015-02-28 14:37:46 +00:00
if (result.s) {
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
smart_str_appendl(&result, str + old_pos, slen - old_pos);
smart_str_0(&result);
Use specialized macro for string zval creation
2015-03-12 13:53:51 +00:00
RETVAL_NEW_STR(result.s);
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
} else {
smart_str_free(&result);
Use specialized macro for string zval creation
2015-03-12 13:53:51 +00:00
RETVAL_STR_COPY(input);
Optimize strtr w/ 2nd arg array Fixes bug #63893: poor efficiency of strtr() using array with keys of very different length. The implementation is basically all new, which carries some risk with it. The algorithm is described in "A Fast Algorithm For Multi-Pattern Searching" (1994) by Sun Wu and Udi Manber.
2013-01-07 02:13:11 +00:00
}
Reimplement strtr()
2014-04-24 15:14:29 +00:00
if (pats == &str_hash) {
zend_hash_destroy(&str_hash);
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
}
Improve strtr()
2015-05-15 01:03:30 +00:00
efree(num_bitset);
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_char_to_str_ex */
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
static zend_string* php_char_to_str_ex(zend_string *str, char from, char *to, size_t to_len, int case_sensitivity, zend_long *replace_count)
{
zend_string *result;
size_t char_count = 0;
Fix #79986: str_ireplace bug with diacritics characters `tolower()` returns an `int`, so we must not convert to `char` which may be `signed` and as such may be subject to overflow (actually, implementation defined behavior). Closes GH-6007
2020-08-18 11:24:17 +00:00
int lc_from = 0;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *source, *source_end= ZSTR_VAL(str) + ZSTR_LEN(str);
char *target;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (case_sensitivity) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
char *p = ZSTR_VAL(str), *e = p + ZSTR_LEN(str);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
while ((p = memchr(p, from, (e - p)))) {
char_count++;
p++;
}
} else {
lc_from = tolower(from);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (source = ZSTR_VAL(str); source < source_end; source++) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (tolower(*source) == lc_from) {
char_count++;
}
}
}
if (char_count == 0) {
return zend_string_copy(str);
}
if (to_len > 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result = zend_string_safe_alloc(char_count, to_len - 1, ZSTR_LEN(str), 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result = zend_string_alloc(ZSTR_LEN(str) - char_count, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
target = ZSTR_VAL(result);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (case_sensitivity) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
char *p = ZSTR_VAL(str), *e = p + ZSTR_LEN(str), *s = ZSTR_VAL(str);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
while ((p = memchr(p, from, (e - p)))) {
memcpy(target, s, (p - s));
target += p - s;
memcpy(target, to, to_len);
target += to_len;
p++;
s = p;
if (replace_count) {
*replace_count += 1;
}
}
if (s < e) {
memcpy(target, s, (e - s));
target += e - s;
}
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (source = ZSTR_VAL(str); source < source_end; source++) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (tolower(*source) == lc_from) {
if (replace_count) {
*replace_count += 1;
}
memcpy(target, to, to_len);
target += to_len;
} else {
*target = *source;
target++;
}
}
}
*target = 0;
return result;
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_str_to_str_ex */
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
static zend_string *php_str_to_str_ex(zend_string *haystack,
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *needle, size_t needle_len, const char *str, size_t str_len, zend_long *replace_count)
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
{
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (needle_len < ZSTR_LEN(haystack)) {
Add helper APIs for maybe-interned string creation Add ZVAL_CHAR/RETVAL_CHAR/RETURN_CHAR as a shortcut for using ZVAL_INTERNED_STRING and ZSTR_CHAR. Add zend_string_init_fast() as a helper for the empty string / one char interned string / zend_string_init() pattern. Also add corresponding ZVAL_STRINGL_FAST etc macros. Closes GH-5684.
2020-06-08 10:45:01 +00:00
zend_string *new_str;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *end;
Fix unused variable warnings
2018-04-26 14:53:58 +00:00
const char *p, *r;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
char *e;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (needle_len == str_len) {
new_str = NULL;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
end = ZSTR_VAL(haystack) + ZSTR_LEN(haystack);
for (p = ZSTR_VAL(haystack); (r = (char*)php_memnstr(p, needle, needle_len, end)); p = r + needle_len) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (!new_str) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
new_str = zend_string_init(ZSTR_VAL(haystack), ZSTR_LEN(haystack), 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(new_str) + (r - ZSTR_VAL(haystack)), str, str_len);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
(*replace_count)++;
}
if (!new_str) {
goto nothing_todo;
}
return new_str;
} else {
size_t count = 0;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *o = ZSTR_VAL(haystack);
const char *n = needle;
const char *endp = o + ZSTR_LEN(haystack);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
while ((o = (char*)php_memnstr(o, n, needle_len, endp))) {
o += needle_len;
count++;
}
if (count == 0) {
/* Needle doesn't occur, shortcircuit the actual replacement. */
goto nothing_todo;
}
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
if (str_len > needle_len) {
new_str = zend_string_safe_alloc(count, str_len - needle_len, ZSTR_LEN(haystack), 0);
} else {
new_str = zend_string_alloc(count * (str_len - needle_len) + ZSTR_LEN(haystack), 0);
}
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
Fix unused variable warnings
2018-04-26 14:53:58 +00:00
e = ZSTR_VAL(new_str);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
end = ZSTR_VAL(haystack) + ZSTR_LEN(haystack);
for (p = ZSTR_VAL(haystack); (r = (char*)php_memnstr(p, needle, needle_len, end)); p = r + needle_len) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
memcpy(e, p, r - p);
e += r - p;
memcpy(e, str, str_len);
e += str_len;
(*replace_count)++;
}
if (p < end) {
memcpy(e, p, end - p);
e += end - p;
}
*e = '\0';
return new_str;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
} else if (needle_len > ZSTR_LEN(haystack) || memcmp(ZSTR_VAL(haystack), needle, ZSTR_LEN(haystack))) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
nothing_todo:
return zend_string_copy(haystack);
} else {
(*replace_count)++;
Add helper APIs for maybe-interned string creation Add ZVAL_CHAR/RETVAL_CHAR/RETURN_CHAR as a shortcut for using ZVAL_INTERNED_STRING and ZSTR_CHAR. Add zend_string_init_fast() as a helper for the empty string / one char interned string / zend_string_init() pattern. Also add corresponding ZVAL_STRINGL_FAST etc macros. Closes GH-5684.
2020-06-08 10:45:01 +00:00
return zend_string_init_fast(str, str_len);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_str_to_str_i_ex */
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
static zend_string *php_str_to_str_i_ex(zend_string *haystack, const char *lc_haystack,
zend_string *needle, const char *str, size_t str_len, zend_long *replace_count)
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
{
zend_string *new_str = NULL;
zend_string *lc_needle;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(needle) < ZSTR_LEN(haystack)) {
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *end;
Fix unused variable warnings
2018-04-26 14:53:58 +00:00
const char *p, *r;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
char *e;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(needle) == str_len) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
lc_needle = php_string_tolower(needle);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
end = lc_haystack + ZSTR_LEN(haystack);
for (p = lc_haystack; (r = (char*)php_memnstr(p, ZSTR_VAL(lc_needle), ZSTR_LEN(lc_needle), end)); p = r + ZSTR_LEN(lc_needle)) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (!new_str) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
new_str = zend_string_init(ZSTR_VAL(haystack), ZSTR_LEN(haystack), 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(new_str) + (r - lc_haystack), str, str_len);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
(*replace_count)++;
}
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_needle, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (!new_str) {
goto nothing_todo;
}
return new_str;
} else {
size_t count = 0;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *o = lc_haystack;
const char *n;
const char *endp = o + ZSTR_LEN(haystack);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
lc_needle = php_string_tolower(needle);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
n = ZSTR_VAL(lc_needle);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
while ((o = (char*)php_memnstr(o, n, ZSTR_LEN(lc_needle), endp))) {
o += ZSTR_LEN(lc_needle);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
count++;
}
if (count == 0) {
/* Needle doesn't occur, shortcircuit the actual replacement. */
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_needle, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
goto nothing_todo;
}
Fix bug #73648 - integer overflow in substr
2017-01-16 20:11:50 +00:00
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
if (str_len > ZSTR_LEN(lc_needle)) {
new_str = zend_string_safe_alloc(count, str_len - ZSTR_LEN(lc_needle), ZSTR_LEN(haystack), 0);
} else {
new_str = zend_string_alloc(count * (str_len - ZSTR_LEN(lc_needle)) + ZSTR_LEN(haystack), 0);
}
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
Fix unused variable warnings
2018-04-26 14:53:58 +00:00
e = ZSTR_VAL(new_str);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
end = lc_haystack + ZSTR_LEN(haystack);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (p = lc_haystack; (r = (char*)php_memnstr(p, ZSTR_VAL(lc_needle), ZSTR_LEN(lc_needle), end)); p = r + ZSTR_LEN(lc_needle)) {
memcpy(e, ZSTR_VAL(haystack) + (p - lc_haystack), r - p);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
e += r - p;
memcpy(e, str, str_len);
e += str_len;
(*replace_count)++;
}
if (p < end) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(e, ZSTR_VAL(haystack) + (p - lc_haystack), end - p);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
e += end - p;
}
*e = '\0';
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_needle, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
return new_str;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
} else if (ZSTR_LEN(needle) > ZSTR_LEN(haystack)) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
nothing_todo:
return zend_string_copy(haystack);
} else {
lc_needle = php_string_tolower(needle);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (memcmp(lc_haystack, ZSTR_VAL(lc_needle), ZSTR_LEN(lc_needle))) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_needle, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
goto nothing_todo;
}
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_needle, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
new_str = zend_string_init(str, str_len, 0);
(*replace_count)++;
return new_str;
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_str_to_str */
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
PHPAPI zend_string *php_str_to_str(const char *haystack, size_t length, const char *needle, size_t needle_len, const char *str, size_t str_len)
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
{
zend_string *new_str;
if (needle_len < length) {
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *end;
const char *s, *p;
char *e, *r;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (needle_len == str_len) {
new_str = zend_string_init(haystack, length, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
end = ZSTR_VAL(new_str) + length;
for (p = ZSTR_VAL(new_str); (r = (char*)php_memnstr(p, needle, needle_len, end)); p = r + needle_len) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
memcpy(r, str, str_len);
}
return new_str;
} else {
if (str_len < needle_len) {
new_str = zend_string_alloc(length, 0);
} else {
size_t count = 0;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *o = haystack;
const char *n = needle;
const char *endp = o + length;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
while ((o = (char*)php_memnstr(o, n, needle_len, endp))) {
o += needle_len;
count++;
}
if (count == 0) {
/* Needle doesn't occur, shortcircuit the actual replacement. */
new_str = zend_string_init(haystack, length, 0);
return new_str;
} else {
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
if (str_len > needle_len) {
new_str = zend_string_safe_alloc(count, str_len - needle_len, length, 0);
} else {
new_str = zend_string_alloc(count * (str_len - needle_len) + length, 0);
}
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
}
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
s = e = ZSTR_VAL(new_str);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
end = haystack + length;
for (p = haystack; (r = (char*)php_memnstr(p, needle, needle_len, end)); p = r + needle_len) {
memcpy(e, p, r - p);
e += r - p;
memcpy(e, str, str_len);
e += str_len;
}
if (p < end) {
memcpy(e, p, end - p);
e += end - p;
}
*e = '\0';
Optimize zend_string_realloc() add more specialized versions zend_string_extend() and zend_string_truncate()
2015-03-19 23:02:42 +00:00
new_str = zend_string_truncate(new_str, e - s, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
return new_str;
}
} else if (needle_len > length || memcmp(haystack, needle, length)) {
new_str = zend_string_init(haystack, length, 0);
return new_str;
} else {
new_str = zend_string_init(str, str_len, 0);
return new_str;
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Translates characters in str using given translation tables */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strtr)
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
{
Convert string|array union parameter types Closes GH-4995
2019-12-09 23:03:02 +00:00
zend_string *str, *from_str = NULL;
HashTable *from_ht = NULL;
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
char *to = NULL;
size_t to_len = 0;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
Z_PARAM_STR(str)
Consolidate new union type ZPP macro names They will now follow the canonical order of types. Older macros are left intact due to maintaining BC. Closes GH-6112
2020-09-10 22:23:54 +00:00
Z_PARAM_ARRAY_HT_OR_STR(from_ht, from_str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
Z_PARAM_STRING_OR_NULL(to, to_len)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
if (!to && from_ht == NULL) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_type_error(2, "must be of type array, string given");
Use RETURN_THROWS() after zend_type_error()
2020-01-01 13:23:21 +00:00
RETURN_THROWS();
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
} else if (to && from_str == NULL) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_type_error(2, "must be of type string, array given");
Use RETURN_THROWS() after zend_type_error()
2020-01-01 13:23:21 +00:00
RETURN_THROWS();
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
}
Fix crash when first argument to strtr is empty
2000-12-13 10:22:36 +00:00
/* shortcut for empty string */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(str) == 0) {
Fix crash when first argument to strtr is empty
2000-12-13 10:22:36 +00:00
RETURN_EMPTY_STRING();
}
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
if (!to) {
Convert string|array union parameter types Closes GH-4995
2019-12-09 23:03:02 +00:00
if (zend_hash_num_elements(from_ht) < 1) {
Use specialized macro for string zval creation
2015-03-12 13:53:51 +00:00
RETURN_STR_COPY(str);
Convert string|array union parameter types Closes GH-4995
2019-12-09 23:03:02 +00:00
} else if (zend_hash_num_elements(from_ht) == 1) {
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
zend_long num_key;
minor optimize on strstr
2018-01-10 10:22:59 +00:00
zend_string *str_key, *tmp_str, *replace, *tmp_replace;
zval *entry;
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
Merge branch 'PHP-7.4' * PHP-7.4: Fixed bug #79817
2020-07-10 12:07:59 +00:00
ZEND_HASH_FOREACH_KEY_VAL_IND(from_ht, num_key, str_key, entry) {
minor optimize on strstr
2018-01-10 10:22:59 +00:00
tmp_str = NULL;
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
if (UNEXPECTED(!str_key)) {
minor optimize on strstr
2018-01-10 10:22:59 +00:00
str_key = tmp_str = zend_long_to_str(num_key);
Fix bug #73648 - integer overflow in substr
2017-01-16 20:11:50 +00:00
}
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
replace = zval_get_tmp_string(entry, &tmp_replace);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(str_key) < 1) {
Warn on strtr(["" => "x"]) Previously: * If only ["" => "x"] was present, the original string was returned without warning. * If both ["" => "x"] and at least one more element was present, false was returned without warning. New behavior: * Ignore "" keys in the replacement array (and perform any remaining replacement). * Throw a warning indicating that an empty string replacement has been ignored. Closes GH-4792.
2019-10-07 11:04:06 +00:00
php_error_docref(NULL, E_WARNING, "Ignoring replacement of empty string");
Use specialized macro for string zval creation
2015-03-12 13:53:51 +00:00
RETVAL_STR_COPY(str);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
} else if (ZSTR_LEN(str_key) == 1) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
RETVAL_STR(php_char_to_str_ex(str,
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(str_key)[0],
ZSTR_VAL(replace),
ZSTR_LEN(replace),
Fixed type (case sensetive/insensetive mess)
2015-01-12 19:30:55 +00:00
1,
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
NULL));
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
} else {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
zend_long dummy;
RETVAL_STR(php_str_to_str_ex(str,
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(str_key), ZSTR_LEN(str_key),
ZSTR_VAL(replace), ZSTR_LEN(replace), &dummy));
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
}
minor optimize on strstr
2018-01-10 10:22:59 +00:00
zend_tmp_string_release(tmp_str);
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_tmp_string_release(tmp_replace);
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
return;
} ZEND_HASH_FOREACH_END();
} else {
Convert string|array union parameter types Closes GH-4995
2019-12-09 23:03:02 +00:00
php_strtr_array(return_value, str, from_ht);
Added special optimized cases when strtr() called with zero or one search/replace pair.
2015-01-12 13:51:59 +00:00
}
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
} else {
Avoid string reallocation in strtr()
2015-01-14 17:58:41 +00:00
RETURN_STR(php_strtr_ex(str,
Convert string|array union parameter types Closes GH-4995
2019-12-09 23:03:02 +00:00
ZSTR_VAL(from_str),
New parameter parsing API
2008-06-25 12:16:17 +00:00
to,
Convert string|array union parameter types Closes GH-4995
2019-12-09 23:03:02 +00:00
MIN(ZSTR_LEN(from_str), to_len)));
@- Implemented 2-Arg version of strtr($str,$translation_array). This can be used @ to revert what htmlspecialchars() did. (Thies) (PHP strtr) new 2-Arg version.
1999-11-20 15:53:18 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Reverse a string */
Optimized strrev with SSSE3 since strrev is not wildly used, and SSSE3 is defined default by some compiler, so won't try to do ifunc here.
2018-02-17 15:19:31 +00:00
#if ZEND_INTRIN_SSSE3_NATIVE
#include <tmmintrin.h>
[AArch64] use rev64 to reverse strings The execution time goes from 4.388s down to 0.563s on a Graviton A1 instance for the benchmark: function reverse_strings() { $a = "foo"; for ($i = 0; $i < 100000; $i++) { strrev($a); $a .= "o"; } }
2019-05-03 15:17:41 +00:00
#elif defined(__aarch64__)
#include <arm_neon.h>
Optimized strrev with SSSE3 since strrev is not wildly used, and SSSE3 is defined default by some compiler, so won't try to do ifunc here.
2018-02-17 15:19:31 +00:00
#endif
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(strrev)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *s, *e;
char *p;
Fixed memory leak in strrev
2014-02-24 10:15:40 +00:00
zend_string *n;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(str)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
n = zend_string_alloc(ZSTR_LEN(str), 0);
p = ZSTR_VAL(n);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Optimized strrev with SSSE3 since strrev is not wildly used, and SSSE3 is defined default by some compiler, so won't try to do ifunc here.
2018-02-17 15:19:31 +00:00
s = ZSTR_VAL(str);
e = s + ZSTR_LEN(str);
--e;
#if ZEND_INTRIN_SSSE3_NATIVE
change layout
2018-02-18 11:47:41 +00:00
if (e - s > 15) {
Optimized strrev with SSSE3 since strrev is not wildly used, and SSSE3 is defined default by some compiler, so won't try to do ifunc here.
2018-02-17 15:19:31 +00:00
const __m128i map = _mm_set_epi8(
0, 1, 2, 3,
4, 5, 6, 7,
8, 9, 10, 11,
12, 13, 14, 15);
change layout
2018-02-18 11:47:41 +00:00
do {
Move map out of loop
2018-02-18 11:34:48 +00:00
const __m128i str = _mm_loadu_si128((__m128i *)(e - 15));
_mm_storeu_si128((__m128i *)p, _mm_shuffle_epi8(str, map));
p += 16;
e -= 16;
change layout
2018-02-18 11:47:41 +00:00
} while (e - s > 15);
}
[AArch64] use rev64 to reverse strings The execution time goes from 4.388s down to 0.563s on a Graviton A1 instance for the benchmark: function reverse_strings() { $a = "foo"; for ($i = 0; $i < 100000; $i++) { strrev($a); $a .= "o"; } }
2019-05-03 15:17:41 +00:00
#elif defined(__aarch64__)
if (e - s > 15) {
do {
const uint8x16_t str = vld1q_u8((uint8_t *)(e - 15));
/* Synthesize rev128 with a rev64 + ext. */
const uint8x16_t rev = vrev64q_u8(str);
const uint8x16_t ext = (uint8x16_t)
vextq_u64((uint64x2_t)rev, (uint64x2_t)rev, 1);
vst1q_u8((uint8_t *)p, ext);
p += 16;
e -= 16;
} while (e - s > 15);
}
Optimized strrev with SSSE3 since strrev is not wildly used, and SSSE3 is defined default by some compiler, so won't try to do ifunc here.
2018-02-17 15:19:31 +00:00
#endif
while (e >= s) {
*p++ = *e--;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
*p = '\0';
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETVAL_NEW_STR(n);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_similar_str */
Improve similar_text(), reduce recursive call to php_similar_char() If the longest common substring is the leftmost common substring, there is no need to check the string prefixes for further common substrings, since there are none.
2018-01-13 09:12:57 +00:00
static void php_similar_str(const char *txt1, size_t len1, const char *txt2, size_t len2, size_t *pos1, size_t *pos2, size_t *max, size_t *count)
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
{
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *p, *q;
const char *end1 = (char *) txt1 + len1;
const char *end2 = (char *) txt2 + len2;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t l;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
*max = 0;
Improve similar_text(), reduce recursive call to php_similar_char() If the longest common substring is the leftmost common substring, there is no need to check the string prefixes for further common substrings, since there are none.
2018-01-13 09:12:57 +00:00
*count = 0;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
for (p = (char *) txt1; p < end1; p++) {
for (q = (char *) txt2; q < end2; q++) {
ws fixes
2002-10-10 18:41:24 +00:00
for (l = 0; (p + l < end1) && (q + l < end2) && (p[l] == q[l]); l++);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
if (l > *max) {
*max = l;
Improve similar_text(), reduce recursive call to php_similar_char() If the longest common substring is the leftmost common substring, there is no need to check the string prefixes for further common substrings, since there are none.
2018-01-13 09:12:57 +00:00
*count += 1;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
*pos1 = p - txt1;
*pos2 = q - txt2;
}
}
}
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_similar_char */
master renames phase 3
2014-08-25 18:22:49 +00:00
static size_t php_similar_char(const char *txt1, size_t len1, const char *txt2, size_t len2)
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
{
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t sum;
Improve similar_text(), reduce recursive call to php_similar_char() If the longest common substring is the leftmost common substring, there is no need to check the string prefixes for further common substrings, since there are none.
2018-01-13 09:12:57 +00:00
size_t pos1 = 0, pos2 = 0, max, count;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Improve similar_text(), reduce recursive call to php_similar_char() If the longest common substring is the leftmost common substring, there is no need to check the string prefixes for further common substrings, since there are none.
2018-01-13 09:12:57 +00:00
php_similar_str(txt1, len1, txt2, len2, &pos1, &pos2, &max, &count);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
if ((sum = max)) {
Improve similar_text(), reduce recursive call to php_similar_char() If the longest common substring is the leftmost common substring, there is no need to check the string prefixes for further common substrings, since there are none.
2018-01-13 09:12:57 +00:00
if (pos1 && pos2 && count > 1) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
sum += php_similar_char(txt1, pos1,
style & ws fixes
2003-02-08 15:26:17 +00:00
txt2, pos2);
}
if ((pos1 + max < len1) && (pos2 + max < len2)) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
sum += php_similar_char(txt1 + pos1 + max, len1 - pos1 - max,
_php3/_php3i -> php rename
1999-10-16 16:35:20 +00:00
txt2 + pos2 + max, len2 - pos2 - max);
style & ws fixes
2003-02-08 15:26:17 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
return sum;
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Calculates the similarity between two strings */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
PHP_FUNCTION(similar_text)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *t1, *t2;
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zval *percent = NULL;
- ARG_COUNT(ht) -> ZEND_NUM_ARGS() mega patch
2000-06-05 19:47:54 +00:00
int ac = ZEND_NUM_ARGS();
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t sim;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 3)
Z_PARAM_STR(t1)
Z_PARAM_STR(t2)
Z_PARAM_OPTIONAL
Implement typed properties RFC: https://wiki.php.net/rfc/typed_properties_v2 This is a squash of PR #3734, which is a squash of PR #3313. Co-authored-by: Bob Weinand <bobwei9@hotmail.com> Co-authored-by: Joe Watkins <krakjoe@php.net> Co-authored-by: Dmitry Stogov <dmitry@zend.com>
2019-01-07 11:28:51 +00:00
Z_PARAM_ZVAL(percent)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(t1) + ZSTR_LEN(t2) == 0) {
ws fix
2002-09-11 14:40:52 +00:00
if (ac > 2) {
Use ZEND_TRA_ASSIGN_REF_... macros for by reference arguments of internal functions.
2019-04-24 13:14:48 +00:00
ZEND_TRY_ASSIGN_REF_DOUBLE(percent, 0);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(0);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
sim = php_similar_char(ZSTR_VAL(t1), ZSTR_LEN(t1), ZSTR_VAL(t2), ZSTR_LEN(t2));
ws fix
2002-09-11 14:40:52 +00:00
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
if (ac > 2) {
Use ZEND_TRA_ASSIGN_REF_... macros for by reference arguments of internal functions.
2019-04-24 13:14:48 +00:00
ZEND_TRY_ASSIGN_REF_DOUBLE(percent, sim * 200.0 / (ZSTR_LEN(t1) + ZSTR_LEN(t2)));
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Begin cleanup, api wise...
2001-09-04 06:18:11 +00:00
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(sim);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Escapes all chars mentioned in charlist with backslash. It creates octal representations if asked to backslash characters with 8th bit set or with ASCII<32 (except '\n', '\r', '\t' etc...) */
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
PHP_FUNCTION(addcslashes)
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str, *what;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(str)
Z_PARAM_STR(what)
ZEND_PARSE_PARAMETERS_END();
Add empty string checks to add?slashes
2000-09-17 09:15:48 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(str) == 0) {
Add empty string checks to add?slashes
2000-09-17 09:15:48 +00:00
RETURN_EMPTY_STRING();
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(what) == 0) {
Avoding memory allocation
2018-02-26 08:54:53 +00:00
RETURN_STR_COPY(str);
Add empty string checks to add?slashes
2000-09-17 09:15:48 +00:00
}
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
RETURN_STR(php_addcslashes_str(ZSTR_VAL(str), ZSTR_LEN(str), ZSTR_VAL(what), ZSTR_LEN(what)));
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Escapes single quote, double quotes and backslash characters in a string with backslashes */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(addslashes)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
Z_PARAM_STR(str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
Add empty string checks to add?slashes
2000-09-17 09:15:48 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(str) == 0) {
Add empty string checks to add?slashes
2000-09-17 09:15:48 +00:00
RETURN_EMPTY_STRING();
}
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
RETURN_STR(php_addslashes(str));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Strips backslashes from a string. Uses C-style conventions */
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
PHP_FUNCTION(stripcslashes)
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(str)
ZEND_PARSE_PARAMETERS_END();
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZVAL_STRINGL(return_value, ZSTR_VAL(str), ZSTR_LEN(str));
Use zend_string in case we will optimize them
2015-01-20 02:59:28 +00:00
php_stripcslashes(Z_STR_P(return_value));
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Strips backslashes from a string */
* reverse fsock/string import patches
1999-06-15 21:51:00 +00:00
PHP_FUNCTION(stripslashes)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(str)
ZEND_PARSE_PARAMETERS_END();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZVAL_STRINGL(return_value, ZSTR_VAL(str), ZSTR_LEN(str));
Use zend_string in case we will optimize them
2015-01-20 02:59:28 +00:00
php_stripslashes(Z_STR_P(return_value));
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_stripcslashes */
Use zend_string in case we will optimize them
2015-01-20 02:59:28 +00:00
PHPAPI void php_stripcslashes(zend_string *str)
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
{
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
const char *source, *end;
char *target;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
size_t nlen = ZSTR_LEN(str), i;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
char numtmp[4];
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (source = (char*)ZSTR_VAL(str), end = source + ZSTR_LEN(str), target = ZSTR_VAL(str); source < end; source++) {
Use zend_string in case we will optimize them
2015-01-20 02:59:28 +00:00
if (*source == '\\' && source + 1 < end) {
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
source++;
switch (*source) {
style & ws fixes
2003-02-08 15:26:17 +00:00
case 'n': *target++='\n'; nlen--; break;
case 'r': *target++='\r'; nlen--; break;
case 'a': *target++='\a'; nlen--; break;
case 't': *target++='\t'; nlen--; break;
case 'v': *target++='\v'; nlen--; break;
case 'b': *target++='\b'; nlen--; break;
case 'f': *target++='\f'; nlen--; break;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
case '\\': *target++='\\'; nlen--; break;
style & ws fixes
2003-02-08 15:26:17 +00:00
case 'x':
if (source+1 < end && isxdigit((int)(*(source+1)))) {
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
numtmp[0] = *++source;
ws fixes.
2002-10-02 17:56:04 +00:00
if (source+1 < end && isxdigit((int)(*(source+1)))) {
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
numtmp[1] = *++source;
numtmp[2] = '\0';
nlen-=3;
} else {
numtmp[1] = '\0';
nlen-=2;
}
- Fix warnings and win32 compile
1999-09-07 20:29:31 +00:00
*target++=(char)strtol(numtmp, NULL, 16);
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
break;
}
/* break is left intentionally */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
default:
i=0;
ws fixes.
2002-10-02 17:56:04 +00:00
while (source < end && *source >= '0' && *source <= '7' && i<3) {
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
numtmp[i++] = *source++;
}
if (i) {
numtmp[i]='\0';
- Fix warnings and win32 compile
1999-09-07 20:29:31 +00:00
*target++=(char)strtol(numtmp, NULL, 8);
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
nlen-=i;
source--;
} else {
*target++=*source;
@- Fixed stripcslashes() to remove to strip \ in unknown escapes instead of @ leaving it. (Andrei)
2000-06-09 13:58:10 +00:00
nlen--;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
}
}
} else {
*target++=*source;
}
}
Fix crash in stripcslashes("") (#5628)
2000-07-17 10:22:30 +00:00
ws fixes.
2002-10-02 17:56:04 +00:00
if (nlen != 0) {
Fix crash in stripcslashes("") (#5628)
2000-07-17 10:22:30 +00:00
*target='\0';
}
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_LEN(str) = nlen;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_addcslashes_str */
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
PHPAPI zend_string *php_addcslashes_str(const char *str, size_t len, const char *what, size_t wlength)
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
{
char flags[256];
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
char *target;
const char *source, *end;
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
char c;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t newlen;
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
zend_string *new_str = zend_string_safe_alloc(4, len, 0, 0);
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
php_charmask((const unsigned char *) what, wlength, flags);
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
for (source = str, end = source + len, target = ZSTR_VAL(new_str); source < end; source++) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
c = *source;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
if (flags[(unsigned char)c]) {
style & ws fixes
2003-02-08 15:26:17 +00:00
if ((unsigned char) c < 32 || (unsigned char) c > 126) {
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
*target++ = '\\';
switch (c) {
case '\n': *target++ = 'n'; break;
case '\t': *target++ = 't'; break;
case '\r': *target++ = 'r'; break;
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
case '\a': *target++ = 'a'; break;
case '\v': *target++ = 'v'; break;
case '\b': *target++ = 'b'; break;
case '\f': *target++ = 'f'; break;
style & ws fixes
2003-02-08 15:26:17 +00:00
default: target += sprintf(target, "%03o", (unsigned char) c);
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
}
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
continue;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
Added addcslashes(), stripcslashes(). Reverted addslashes() to the very original behaviour.
1999-09-05 20:55:13 +00:00
*target++ = '\\';
}
*target++ = c;
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
}
*target = 0;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
newlen = target - ZSTR_VAL(new_str);
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
if (newlen < len * 4) {
Optimize zend_string_realloc() add more specialized versions zend_string_extend() and zend_string_truncate()
2015-03-19 23:02:42 +00:00
new_str = zend_string_truncate(new_str, newlen, 0);
Appended second argument of addslashes(), Changed php_addslashes() to add \000 instead of just \0
1999-09-05 00:55:48 +00:00
}
return new_str;
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_addcslashes */
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
PHPAPI zend_string *php_addcslashes(zend_string *str, const char *what, size_t wlength)
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
{
return php_addcslashes_str(ZSTR_VAL(str), ZSTR_LEN(str), what, wlength);
}
/* }}} */
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
/* {{{ php_addslashes */
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
#if ZEND_INTRIN_SSE4_2_NATIVE
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
# include <nmmintrin.h>
# include "Zend/zend_bitset.h"
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
#elif ZEND_INTRIN_SSE4_2_RESOLVER
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
# include <nmmintrin.h>
# include "Zend/zend_bitset.h"
# include "Zend/zend_cpuinfo.h"
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
ZEND_INTRIN_SSE4_2_FUNC_DECL(zend_string *php_addslashes_sse42(zend_string *str));
zend_string *php_addslashes_default(zend_string *str);
If the target attribute is available, attach it explicitly The ifunc attribute might be not available as it is binary format dependent.
2018-02-03 16:15:18 +00:00
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
ZEND_INTRIN_SSE4_2_FUNC_DECL(void php_stripslashes_sse42(zend_string *str));
void php_stripslashes_default(zend_string *str);
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
# if ZEND_INTRIN_SSE4_2_FUNC_PROTO
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
PHPAPI zend_string *php_addslashes(zend_string *str) __attribute__((ifunc("resolve_addslashes")));
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
PHPAPI void php_stripslashes(zend_string *str) __attribute__((ifunc("resolve_stripslashes")));
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
Fix ifunc resolver return type warning
2020-04-28 13:25:35 +00:00
typedef zend_string *(*php_addslashes_func_t)(zend_string *);
typedef void (*php_stripslashes_func_t)(zend_string *);
Add additional no_sanitize_address attributes To fix bug #77447 in release mode as well.
2019-01-18 14:01:51 +00:00
ZEND_NO_SANITIZE_ADDRESS
Mark ifunc resolver functions as unused It looks like clang doesn't see that these are used through __attribute__((ifunc)) and generates a warning.
2019-06-28 14:55:54 +00:00
ZEND_ATTRIBUTE_UNUSED /* clang mistakenly warns about this */
Fix ifunc resolver return type warning
2020-04-28 13:25:35 +00:00
static php_addslashes_func_t resolve_addslashes() {
Optimized base64_encode/decode with SIMD instructions
2018-02-12 12:46:17 +00:00
if (zend_cpu_supports_sse42()) {
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
return php_addslashes_sse42;
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
}
Make ABI of SIMD optimized functions independent of compiler flags Always export these as normal functions and only use function pointers internally if necessary.
2019-02-14 09:21:27 +00:00
return php_addslashes_default;
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
}
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
Add additional no_sanitize_address attributes To fix bug #77447 in release mode as well.
2019-01-18 14:01:51 +00:00
ZEND_NO_SANITIZE_ADDRESS
Mark ifunc resolver functions as unused It looks like clang doesn't see that these are used through __attribute__((ifunc)) and generates a warning.
2019-06-28 14:55:54 +00:00
ZEND_ATTRIBUTE_UNUSED /* clang mistakenly warns about this */
Fix ifunc resolver return type warning
2020-04-28 13:25:35 +00:00
static php_stripslashes_func_t resolve_stripslashes() {
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
if (zend_cpu_supports_sse42()) {
return php_stripslashes_sse42;
}
Make ABI of SIMD optimized functions independent of compiler flags Always export these as normal functions and only use function pointers internally if necessary.
2019-02-14 09:21:27 +00:00
return php_stripslashes_default;
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
}
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
# else /* ZEND_INTRIN_SSE4_2_FUNC_PTR */
Make ABI of SIMD optimized functions independent of compiler flags Always export these as normal functions and only use function pointers internally if necessary.
2019-02-14 09:21:27 +00:00
static zend_string *(*php_addslashes_ptr)(zend_string *str) = NULL;
static void (*php_stripslashes_ptr)(zend_string *str) = NULL;
PHPAPI zend_string *php_addslashes(zend_string *str) {
return php_addslashes_ptr(str);
}
PHPAPI void php_stripslashes(zend_string *str) {
php_stripslashes_ptr(str);
}
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ PHP_MINIT_FUNCTION */
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
PHP_MINIT_FUNCTION(string_intrin)
{
Avoid direct calls to zend_cpu_supports() While the use of zend_cpu_supports_*() is only strictly necessary inside ifunc resolvers, where the cpu state has not been initialized yet, we should prefer the compiler builtins in all cases.
2020-11-27 10:18:10 +00:00
if (zend_cpu_supports_sse42()) {
Make ABI of SIMD optimized functions independent of compiler flags Always export these as normal functions and only use function pointers internally if necessary.
2019-02-14 09:21:27 +00:00
php_addslashes_ptr = php_addslashes_sse42;
php_stripslashes_ptr = php_stripslashes_sse42;
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
} else {
Make ABI of SIMD optimized functions independent of compiler flags Always export these as normal functions and only use function pointers internally if necessary.
2019-02-14 09:21:27 +00:00
php_addslashes_ptr = php_addslashes_default;
php_stripslashes_ptr = php_stripslashes_default;
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
}
return SUCCESS;
}
/* }}} */
# endif
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
#endif
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
#if ZEND_INTRIN_SSE4_2_NATIVE || ZEND_INTRIN_SSE4_2_RESOLVER
# if ZEND_INTRIN_SSE4_2_NATIVE
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
PHPAPI zend_string *php_addslashes(zend_string *str) /* {{{ */
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
# elif ZEND_INTRIN_SSE4_2_RESOLVER
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
zend_string *php_addslashes_sse42(zend_string *str)
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
# endif
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Move useful macro into zend_portability.h
2018-01-24 09:14:28 +00:00
ZEND_SET_ALIGNED(16, static const char slashchars[16]) = "\'\"\\\0";
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
__m128i w128, s128;
uint32_t res = 0;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
/* maximum string length, worst case situation */
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
char *target;
const char *source, *end;
Optimized php_addslashes
2015-01-19 15:46:39 +00:00
size_t offset;
Refactor php_addslashes using zend_string
2014-02-24 04:07:09 +00:00
zend_string *new_str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
(PHP wddx_deserialize) Fixed a problem with deserializing empty values. (php_addslashes) Guard against NULL data.
2000-02-02 21:53:27 +00:00
if (!str) {
Improved zend_string API (Francois Laupretre) Squashed commit of the following: commit d96eab8d79b75ac83d49d49ae4665f948d15a804 Author: Francois Laupretre <francois@tekwire.net> Date: Fri Jun 26 01:23:31 2015 +0200 Use the new 'ZSTR' macros in the rest of the code. Does not change anything to the generated code (thanks to compat macros) but cleaner. commit b3526439104ac7a89a8e0c79dbebf33b22bd01b8 Author: Francois Laupretre <francois@tekwire.net> Date: Thu Jun 25 13:45:06 2015 +0200 Improve zend_string API Add missing methods
2015-06-29 13:44:54 +00:00
return ZSTR_EMPTY_ALLOC();
(PHP wddx_deserialize) Fixed a problem with deserializing empty values. (php_addslashes) Guard against NULL data.
2000-02-02 21:53:27 +00:00
}
Refactor php_addslashes using zend_string
2014-02-24 04:07:09 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
source = ZSTR_VAL(str);
end = source + ZSTR_LEN(str);
Optimized php_addslashes
2015-01-19 15:46:39 +00:00
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
if (ZSTR_LEN(str) > 15) {
slashchars are 16 bytes aligned
2018-03-16 08:01:32 +00:00
w128 = _mm_load_si128((__m128i *)slashchars);
Simplified the codes (the penalty of loading unaligned address is negligible)
2018-02-20 12:56:23 +00:00
do {
s128 = _mm_loadu_si128((__m128i *)source);
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
res = _mm_cvtsi128_si32(_mm_cmpestrm(w128, 4, s128, 16, _SIDD_UBYTE_OPS | _SIDD_CMP_EQUAL_ANY | _SIDD_BIT_MASK));
if (res) {
goto do_escape;
}
Simplified the codes (the penalty of loading unaligned address is negligible)
2018-02-20 12:56:23 +00:00
source += 16;
} while ((end - source) > 15);
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
}
Optimized php_addslashes
2015-01-19 15:46:39 +00:00
while (source < end) {
switch (*source) {
case '\0':
case '\'':
case '\"':
case '\\':
goto do_escape;
default:
source++;
break;
}
}
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
return zend_string_copy(str);
Optimized php_addslashes
2015-01-19 15:46:39 +00:00
do_escape:
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
offset = source - (char *)ZSTR_VAL(str);
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
new_str = zend_string_safe_alloc(2, ZSTR_LEN(str) - offset, offset, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(new_str), ZSTR_VAL(str), offset);
target = ZSTR_VAL(new_str) + offset;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
if (res) {
int pos = 0;
do {
int i, n = zend_ulong_ntz(res);
for (i = 0; i < n; i++) {
*target++ = source[pos + i];
}
pos += n;
*target++ = '\\';
if (source[pos] == '\0') {
*target++ = '0';
} else {
*target++ = source[pos];
}
pos++;
res = res >> (n + 1);
} while (res);
for (; pos < 16; pos++) {
*target++ = source[pos];
}
source += 16;
} else if (end - source > 15) {
slashchars are 16 bytes aligned
2018-03-16 08:01:32 +00:00
w128 = _mm_load_si128((__m128i *)slashchars);
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
}
for (; end - source > 15; source += 16) {
int pos = 0;
Simplified the codes (the penalty of loading unaligned address is negligible)
2018-02-20 12:56:23 +00:00
s128 = _mm_loadu_si128((__m128i *)source);
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
res = _mm_cvtsi128_si32(_mm_cmpestrm(w128, 4, s128, 16, _SIDD_UBYTE_OPS | _SIDD_CMP_EQUAL_ANY | _SIDD_BIT_MASK));
if (res) {
do {
int i, n = zend_ulong_ntz(res);
for (i = 0; i < n; i++) {
*target++ = source[pos + i];
}
pos += n;
*target++ = '\\';
if (source[pos] == '\0') {
*target++ = '0';
} else {
*target++ = source[pos];
}
pos++;
res = res >> (n + 1);
} while (res);
for (; pos < 16; pos++) {
*target++ = source[pos];
}
} else {
_mm_storeu_si128((__m128i*)target, s128);
target += 16;
}
}
- remove magic quotes support, functions are kept (see the NEWS entry for the details) for BC reasons but do not allow to set enable MQ
2011-07-22 11:25:30 +00:00
while (source < end) {
switch (*source) {
case '\0':
*target++ = '\\';
*target++ = '0';
break;
case '\'':
case '\"':
case '\\':
*target++ = '\\';
/* break is missing *intentionally* */
default:
*target++ = *source;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
break;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
source++;
}
*target = '\0';
if (ZSTR_LEN(new_str) - (target - ZSTR_VAL(new_str)) > 16) {
new_str = zend_string_truncate(new_str, target - ZSTR_VAL(new_str), 0);
} else {
ZSTR_LEN(new_str) = target - ZSTR_VAL(new_str);
}
return new_str;
}
/* }}} */
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
#endif
- remove magic quotes support, functions are kept (see the NEWS entry for the details) for BC reasons but do not allow to set enable MQ
2011-07-22 11:25:30 +00:00
vectorize string-escape with NEON On this benchmark: function simple_string_escape() { $a = "test'asd'asd'' asd\'\"asdfasdf"; for($i=0; $i<512; $i++) { $a .= chr($i%256); } for ($i = 0; $i < 100000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= chr($i%256); else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 21.619s to 8.139s (165% speedup) on an A1 Graviton instance. When removing the characters that need escaping, i.e., this benchmark: function simple_string() { $a = "testasdasd asdasdfasdf"; for ($i = 0; $i < 10000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= "test dedeasdf"; else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 2.932s down to 0.516s (468% speedup) on an A1 Graviton instance.
2019-07-10 18:31:27 +00:00
#ifdef __aarch64__
typedef union {
uint8_t mem[16];
uint64_t dw[2];
} quad_word;
static zend_always_inline quad_word aarch64_contains_slash_chars(uint8x16_t x) {
uint8x16_t s0 = vceqq_u8(x, vdupq_n_u8('\0'));
uint8x16_t s1 = vceqq_u8(x, vdupq_n_u8('\''));
uint8x16_t s2 = vceqq_u8(x, vdupq_n_u8('\"'));
uint8x16_t s3 = vceqq_u8(x, vdupq_n_u8('\\'));
uint8x16_t s01 = vorrq_u8(s0, s1);
uint8x16_t s23 = vorrq_u8(s2, s3);
uint8x16_t s0123 = vorrq_u8(s01, s23);
quad_word qw;
vst1q_u8(qw.mem, s0123);
return qw;
}
static zend_always_inline char *aarch64_add_slashes(quad_word res, const char *source, char *target)
{
int i = 0;
for (; i < 16; i++) {
char s = source[i];
if (res.mem[i] == 0)
*target++ = s;
else {
*target++ = '\\';
if (s == '\0')
*target++ = '0';
else
*target++ = s;
}
}
return target;
}
#endif /* __aarch64__ */
Implement intrinsic function resolving per MINIT
2018-01-24 08:33:32 +00:00
#if !ZEND_INTRIN_SSE4_2_NATIVE
# if ZEND_INTRIN_SSE4_2_RESOLVER
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
zend_string *php_addslashes_default(zend_string *str) /* {{{ */
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
# else
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
PHPAPI zend_string *php_addslashes(zend_string *str)
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
# endif
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
{
/* maximum string length, worst case situation */
Use const pointers where appropriate
2018-03-16 19:36:41 +00:00
char *target;
const char *source, *end;
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
size_t offset;
zend_string *new_str;
if (!str) {
return ZSTR_EMPTY_ALLOC();
}
source = ZSTR_VAL(str);
end = source + ZSTR_LEN(str);
vectorize string-escape with NEON On this benchmark: function simple_string_escape() { $a = "test'asd'asd'' asd\'\"asdfasdf"; for($i=0; $i<512; $i++) { $a .= chr($i%256); } for ($i = 0; $i < 100000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= chr($i%256); else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 21.619s to 8.139s (165% speedup) on an A1 Graviton instance. When removing the characters that need escaping, i.e., this benchmark: function simple_string() { $a = "testasdasd asdasdfasdf"; for ($i = 0; $i < 10000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= "test dedeasdf"; else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 2.932s down to 0.516s (468% speedup) on an A1 Graviton instance.
2019-07-10 18:31:27 +00:00
# ifdef __aarch64__
quad_word res = {0};
if (ZSTR_LEN(str) > 15) {
do {
res = aarch64_contains_slash_chars(vld1q_u8((uint8_t *)source));
if (res.dw[0] | res.dw[1])
goto do_escape;
source += 16;
} while ((end - source) > 15);
}
/* Finish the last 15 bytes or less with the scalar loop. */
# endif /* __aarch64__ */
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
while (source < end) {
switch (*source) {
case '\0':
case '\'':
case '\"':
case '\\':
goto do_escape;
default:
source++;
break;
}
}
Changed php_add[c]slashes prototypes (removed should_free argument)
2018-06-28 21:19:16 +00:00
return zend_string_copy(str);
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
do_escape:
offset = source - (char *)ZSTR_VAL(str);
new_str = zend_string_safe_alloc(2, ZSTR_LEN(str) - offset, offset, 0);
memcpy(ZSTR_VAL(new_str), ZSTR_VAL(str), offset);
target = ZSTR_VAL(new_str) + offset;
vectorize string-escape with NEON On this benchmark: function simple_string_escape() { $a = "test'asd'asd'' asd\'\"asdfasdf"; for($i=0; $i<512; $i++) { $a .= chr($i%256); } for ($i = 0; $i < 100000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= chr($i%256); else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 21.619s to 8.139s (165% speedup) on an A1 Graviton instance. When removing the characters that need escaping, i.e., this benchmark: function simple_string() { $a = "testasdasd asdasdfasdf"; for ($i = 0; $i < 10000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= "test dedeasdf"; else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 2.932s down to 0.516s (468% speedup) on an A1 Graviton instance.
2019-07-10 18:31:27 +00:00
# ifdef __aarch64__
if (res.dw[0] | res.dw[1]) {
target = aarch64_add_slashes(res, source, target);
source += 16;
}
for (; end - source > 15; source += 16) {
uint8x16_t x = vld1q_u8((uint8_t *)source);
res = aarch64_contains_slash_chars(x);
if (res.dw[0] | res.dw[1]) {
target = aarch64_add_slashes(res, source, target);
} else {
vst1q_u8((uint8_t*)target, x);
target += 16;
}
}
/* Finish the last 15 bytes or less with the scalar loop. */
# endif /* __aarch64__ */
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
while (source < end) {
switch (*source) {
case '\0':
*target++ = '\\';
*target++ = '0';
break;
case '\'':
case '\"':
case '\\':
*target++ = '\\';
/* break is missing *intentionally* */
default:
*target++ = *source;
break;
}
- remove magic quotes support, functions are kept (see the NEWS entry for the details) for BC reasons but do not allow to set enable MQ
2011-07-22 11:25:30 +00:00
source++;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
*target = '\0';
Optimized php_addslashes
2015-01-19 15:46:39 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(new_str) - (target - ZSTR_VAL(new_str)) > 16) {
new_str = zend_string_truncate(new_str, target - ZSTR_VAL(new_str), 0);
Optimized php_addslashes
2015-01-19 15:46:39 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_LEN(new_str) = target - ZSTR_VAL(new_str);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Refactor php_addslashes using zend_string
2014-02-24 04:07:09 +00:00
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
return new_str;
}
Don't use ifunc if compiling with __SSE4_2__ deinfed
2018-01-17 09:31:10 +00:00
#endif
Optimized php_addslashes with SSE4.2 instructions According to benchmark (https://gist.github.com/laruence/fd0d443d2c5bacca9d8ab99250499956) this brings 30%+ execution time reduced. In the further, this open the door for possible SSE4.2 optimizations in other places.
2018-01-15 13:01:27 +00:00
/* }}} */
Add missing closing fold
2018-01-17 14:08:22 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
/* {{{ php_stripslashes
*
* be careful, this edits the string in-place */
static zend_always_inline char *php_stripslashes_impl(const char *str, char *out, size_t len)
{
vectorize string-escape with NEON On this benchmark: function simple_string_escape() { $a = "test'asd'asd'' asd\'\"asdfasdf"; for($i=0; $i<512; $i++) { $a .= chr($i%256); } for ($i = 0; $i < 100000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= chr($i%256); else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 21.619s to 8.139s (165% speedup) on an A1 Graviton instance. When removing the characters that need escaping, i.e., this benchmark: function simple_string() { $a = "testasdasd asdasdfasdf"; for ($i = 0; $i < 10000; $i++) { if ($a === stripslashes(addslashes($a))) $a .= "test dedeasdf"; else { echo "error at i=".$i."\n"; return; } } } the execution time goes from 2.932s down to 0.516s (468% speedup) on an A1 Graviton instance.
2019-07-10 18:31:27 +00:00
#ifdef __aarch64__
while (len > 15) {
uint8x16_t x = vld1q_u8((uint8_t *)str);
quad_word q;
vst1q_u8(q.mem, vceqq_u8(x, vdupq_n_u8('\\')));
if (q.dw[0] | q.dw[1]) {
int i = 0;
for (; i < 16; i++) {
if (q.mem[i] == 0) {
*out++ = str[i];
continue;
}
i++; /* skip the slash */
char s = str[i];
if (s == '0')
*out++ = '\0';
else
*out++ = s; /* preserve the next character */
}
str += i;
len -= i;
} else {
vst1q_u8((uint8_t*)out, x);
out += 16;
str += 16;
len -= 16;
}
}
/* Finish the last 15 bytes or less with the scalar loop. */
#endif /* __aarch64__ */
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
while (len > 0) {
if (*str == '\\') {
str++; /* skip the slash */
len--;
if (len > 0) {
if (*str == '0') {
*out++='\0';
str++;
} else {
*out++ = *str++; /* preserve the next character */
}
len--;
}
} else {
*out++ = *str++;
len--;
}
}
return out;
}
#if ZEND_INTRIN_SSE4_2_NATIVE || ZEND_INTRIN_SSE4_2_RESOLVER
# if ZEND_INTRIN_SSE4_2_NATIVE
PHPAPI void php_stripslashes(zend_string *str)
# elif ZEND_INTRIN_SSE4_2_RESOLVER
Fix def signature to match declaration
2018-02-21 10:53:38 +00:00
void php_stripslashes_sse42(zend_string *str)
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
# endif
{
const char *s = ZSTR_VAL(str);
char *t = ZSTR_VAL(str);
size_t l = ZSTR_LEN(str);
if (l > 15) {
const __m128i slash = _mm_set1_epi8('\\');
do {
__m128i in = _mm_loadu_si128((__m128i *)s);
__m128i any_slash = _mm_cmpeq_epi8(in, slash);
uint32_t res = _mm_movemask_epi8(any_slash);
if (res) {
int i, n = zend_ulong_ntz(res);
const char *e = s + 15;
l -= n;
for (i = 0; i < n; i++) {
*t++ = *s++;
}
for (; s < e; s++) {
if (*s == '\\') {
s++;
l--;
if (*s == '0') {
*t = '\0';
} else {
*t = *s;
}
} else {
*t = *s;
}
t++;
l--;
}
} else {
_mm_storeu_si128((__m128i *)t, in);
s += 16;
t += 16;
l -= 16;
}
} while (l > 15);
}
t = php_stripslashes_impl(s, t, l);
if (t != (ZSTR_VAL(str) + ZSTR_LEN(str))) {
ZSTR_LEN(str) = t - ZSTR_VAL(str);
ZSTR_VAL(str)[ZSTR_LEN(str)] = '\0';
}
}
#endif
#if !ZEND_INTRIN_SSE4_2_NATIVE
# if ZEND_INTRIN_SSE4_2_RESOLVER
void php_stripslashes_default(zend_string *str) /* {{{ */
# else
PHPAPI void php_stripslashes(zend_string *str)
# endif
{
Use const pointers
2018-03-17 11:45:22 +00:00
const char *t = php_stripslashes_impl(ZSTR_VAL(str), ZSTR_VAL(str), ZSTR_LEN(str));
Optimized php_stripslash with SSSE3 instr
2018-02-21 02:09:11 +00:00
if (t != (ZSTR_VAL(str) + ZSTR_LEN(str))) {
ZSTR_LEN(str) = t - ZSTR_VAL(str);
ZSTR_VAL(str)[ZSTR_LEN(str)] = '\0';
}
}
/* }}} */
#endif
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
#define _HEB_BLOCK_TYPE_ENG 1
#define _HEB_BLOCK_TYPE_HEB 2
style & ws fixes
2003-02-08 15:26:17 +00:00
#define isheb(c) (((((unsigned char) c) >= 224) && (((unsigned char) c) <= 250)) ? 1 : 0)
#define _isblank(c) (((((unsigned char) c) == ' ' || ((unsigned char) c) == '\t')) ? 1 : 0)
#define _isnewline(c) (((((unsigned char) c) == '\n' || ((unsigned char) c) == '\r')) ? 1 : 0)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_str_replace_in_subject */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
static zend_long php_str_replace_in_subject(
zend_string *search_str, HashTable *search_ht, zend_string *replace_str, HashTable *replace_ht,
zend_string *subject_str, zval *result, int case_sensitivity
) {
Fixed bug #79951 One branch did not release tmp_replace_entry_str. Also reduce the scope of some variables.
2020-08-11 08:33:59 +00:00
zval *search_entry;
Merge branch 'PHP-7.4' * PHP-7.4: Fixed bug #79951
2020-08-11 08:35:59 +00:00
zend_string *tmp_result;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
char *replace_value = NULL;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
size_t replace_len = 0;
zend_long replace_count = 0;
zend_string *lc_subject_str = NULL;
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
uint32_t replace_idx;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(subject_str) == 0) {
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
ZVAL_EMPTY_STRING(result);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
return 0;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
/* If search is an array */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
if (search_ht) {
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
/* Duplicate subject string for repeated replacement */
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
zend_string_addref(subject_str);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
if (replace_ht) {
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
replace_idx = 0;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
} else {
/* Set replacement value to the passed one */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
replace_value = ZSTR_VAL(replace_str);
replace_len = ZSTR_LEN(replace_str);
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
/* For each entry in the search array, get the entry */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
ZEND_HASH_FOREACH_VAL_IND(search_ht, search_entry) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* Make sure we're dealing with strings. */
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_string *tmp_search_str;
zend_string *search_str = zval_get_tmp_string(search_entry, &tmp_search_str);
Fixed bug #79951 One branch did not release tmp_replace_entry_str. Also reduce the scope of some variables.
2020-08-11 08:33:59 +00:00
zend_string *replace_entry_str, *tmp_replace_entry_str = NULL;
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
/* If replace is an array. */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
if (replace_ht) {
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
/* Get current entry */
Fixed bug #79951 One branch did not release tmp_replace_entry_str. Also reduce the scope of some variables.
2020-08-11 08:33:59 +00:00
zval *replace_entry = NULL;
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
while (replace_idx < replace_ht->nNumUsed) {
replace_entry = &replace_ht->arData[replace_idx].val;
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
if (Z_TYPE_P(replace_entry) != IS_UNDEF) {
break;
}
replace_idx++;
}
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
if (replace_idx < replace_ht->nNumUsed) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* Make sure we're dealing with strings. */
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
replace_entry_str = zval_get_tmp_string(replace_entry, &tmp_replace_entry_str);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
/* Set replacement value to the one we got from array */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
replace_value = ZSTR_VAL(replace_entry_str);
replace_len = ZSTR_LEN(replace_entry_str);
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
Get rid of old HashTable iteration API
2015-02-10 13:59:34 +00:00
replace_idx++;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
} else {
/* We've run out of replacement strings, so use an empty one. */
- Nuke empty_string. It is a reminanent from the time where RETURN_FALSE() used to return "" and not bool(false). It's not worth keeping it because STR_FREE() and zval_dtor() always have to check for it and it slows down the general case. In addition, it seems that empty_string has been abused quite a lot, and was used not only for setting zval's but generally in PHP code instead of "", which wasn't the intention. Last but not least, nuking empty_string should improve stability as I doubt every place correctly checked if they are not mistakenly erealloc()'ing it or calling efree() on it. NOTE: Some code is probably broken. Each extension maintainer should check and see that my changes are OK. Also, I haven't had time to touch PECL yet. Will try and do it tomorrow.
2004-07-19 07:19:50 +00:00
replace_value = "";
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
replace_len = 0;
}
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fixed #71188 (str_replace converts integers in original $search array to strings)
2015-12-22 03:07:30 +00:00
if (ZSTR_LEN(search_str) == 1) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
zend_long old_replace_count = replace_count;
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
tmp_result = php_char_to_str_ex(subject_str,
Fixed #71188 (str_replace converts integers in original $search array to strings)
2015-12-22 03:07:30 +00:00
ZSTR_VAL(search_str)[0],
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
replace_value,
replace_len,
Fix Bug#24784 single character search keys not respecting case sensitivity/replace count parameters.
2003-07-25 01:03:39 +00:00
case_sensitivity,
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
&replace_count);
if (lc_subject_str && replace_count != old_replace_count) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_subject_str, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
lc_subject_str = NULL;
}
Fixed #71188 (str_replace converts integers in original $search array to strings)
2015-12-22 03:07:30 +00:00
} else if (ZSTR_LEN(search_str) > 1) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (case_sensitivity) {
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
tmp_result = php_str_to_str_ex(subject_str,
Fixed #71188 (str_replace converts integers in original $search array to strings)
2015-12-22 03:07:30 +00:00
ZSTR_VAL(search_str), ZSTR_LEN(search_str),
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
replace_value, replace_len, &replace_count);
} else {
zend_long old_replace_count = replace_count;
if (!lc_subject_str) {
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
lc_subject_str = php_string_tolower(subject_str);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
tmp_result = php_str_to_str_i_ex(subject_str, ZSTR_VAL(lc_subject_str),
Fixed #71188 (str_replace converts integers in original $search array to strings)
2015-12-22 03:07:30 +00:00
search_str, replace_value, replace_len, &replace_count);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (replace_count != old_replace_count) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_subject_str, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
lc_subject_str = NULL;
}
Fix bug #73648 - integer overflow in substr
2017-01-16 20:11:50 +00:00
}
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
} else {
zend_tmp_string_release(tmp_search_str);
Fixed bug #79951 One branch did not release tmp_replace_entry_str. Also reduce the scope of some variables.
2020-08-11 08:33:59 +00:00
zend_tmp_string_release(tmp_replace_entry_str);
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
continue;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_tmp_string_release(tmp_search_str);
Fixed bug #79951 One branch did not release tmp_replace_entry_str. Also reduce the scope of some variables.
2020-08-11 08:33:59 +00:00
zend_tmp_string_release(tmp_replace_entry_str);
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
if (subject_str == tmp_result) {
zend_string_delref(subject_str);
} else {
zend_string_release_ex(subject_str, 0);
subject_str = tmp_result;
if (ZSTR_LEN(subject_str) == 0) {
zend_string_release_ex(subject_str, 0);
ZVAL_EMPTY_STRING(result);
if (lc_subject_str) {
zend_string_release_ex(lc_subject_str, 0);
}
return replace_count;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
Fixed bug #13901.
2001-11-10 23:44:04 +00:00
}
ZEND_HASH_FOREACH_* usage
2014-04-18 17:49:07 +00:00
} ZEND_HASH_FOREACH_END();
str_replace() micro-optimization
2018-07-03 14:47:08 +00:00
ZVAL_STR(result, subject_str);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (lc_subject_str) {
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_subject_str, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
} else {
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
ZEND_ASSERT(search_str);
if (ZSTR_LEN(search_str) == 1) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
ZVAL_STR(result,
php_char_to_str_ex(subject_str,
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
ZSTR_VAL(search_str)[0],
ZSTR_VAL(replace_str),
ZSTR_LEN(replace_str),
Fix Bug#24784 single character search keys not respecting case sensitivity/replace count parameters.
2003-07-25 01:03:39 +00:00
case_sensitivity,
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
&replace_count));
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
} else if (ZSTR_LEN(search_str) > 1) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
if (case_sensitivity) {
ZVAL_STR(result, php_str_to_str_ex(subject_str,
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
ZSTR_VAL(search_str), ZSTR_LEN(search_str),
ZSTR_VAL(replace_str), ZSTR_LEN(replace_str), &replace_count));
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
} else {
Fixed bug #70140 (str_ireplace/php_string_tolower - Arbitrary Code Execution)
2015-07-27 03:17:05 +00:00
lc_subject_str = php_string_tolower(subject_str);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZVAL_STR(result, php_str_to_str_i_ex(subject_str, ZSTR_VAL(lc_subject_str),
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
search_str, ZSTR_VAL(replace_str), ZSTR_LEN(replace_str), &replace_count));
Use zend_string_release_ex() instread of zend_string_release() in places, where we sure about string persistence.
2018-05-28 13:27:12 +00:00
zend_string_release_ex(lc_subject_str, 0);
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
}
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
} else {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
ZVAL_STR_COPY(result, subject_str);
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
return replace_count;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_str_replace_common */
Unify str_replace and str_ireplace using INTERNAL_FUNCTION_PARAM_PASSTHRU -- reduce codebase/maintenance complexity
2003-01-30 20:09:19 +00:00
static void php_str_replace_common(INTERNAL_FUNCTION_PARAMETERS, int case_sensitivity)
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
{
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
zend_string *search_str;
HashTable *search_ht;
zend_string *replace_str;
HashTable *replace_ht;
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
zend_string *subject_str;
HashTable *subject_ht;
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
zval *subject_entry, *zcount = NULL;
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zval result;
zend_string *string_key;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_ulong num_key;
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
zend_long count = 0;
@- Improved str_replace() to accept an array for any parameter - similar @ to preg_replace(). (Andrei)
2001-01-22 21:39:55 +00:00
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_START(3, 4)
Consolidate new union type ZPP macro names They will now follow the canonical order of types. Older macros are left intact due to maintaining BC. Closes GH-6112
2020-09-10 22:23:54 +00:00
Z_PARAM_ARRAY_HT_OR_STR(search_ht, search_str)
Z_PARAM_ARRAY_HT_OR_STR(replace_ht, replace_str)
Z_PARAM_ARRAY_HT_OR_STR(subject_ht, subject_str)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
Z_PARAM_OPTIONAL
Implement typed properties RFC: https://wiki.php.net/rfc/typed_properties_v2 This is a squash of PR #3734, which is a squash of PR #3313. Co-authored-by: Bob Weinand <bobwei9@hotmail.com> Co-authored-by: Joe Watkins <krakjoe@php.net> Co-authored-by: Dmitry Stogov <dmitry@zend.com>
2019-01-07 11:28:51 +00:00
Z_PARAM_ZVAL(zcount)
Fast parameter parsing API This API is experemental. It may be changed or removed. It should be used only for really often used functions. (Keep the original parsing code and wrap usage with #ifndef FAST_ZPP)
2014-07-11 12:32:20 +00:00
ZEND_PARSE_PARAMETERS_END();
- regex cache (needs some work) - str_replace changes from php3
1999-05-12 15:40:15 +00:00
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
/* Make sure we're dealing with strings and do the replacement. */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
if (search_str && replace_ht) {
zend_argument_type_error(2, "must be of type %s when argument #1 ($search) is %s",
search_str ? "string" : "array", search_str ? "a string" : "an array"
);
Use RETURN_THROWS() in various places
2020-01-03 19:44:29 +00:00
RETURN_THROWS();
Allow exceptions in __toString() RFC: https://wiki.php.net/rfc/tostring_exceptions And convert some object to string conversion related recoverable fatal errors into Error exceptions. Improve exception safety of internal code performing string conversions.
2019-02-26 14:32:18 +00:00
}
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
/* if subject is an array */
Support "string or array" in zpp This is one of our more common argument unions. Usage is just prototyped in a few places, certainly not a full conversion. I'm removing the str_replace.phpt test, because aparently it was split up into smaller tests at some point, but the original has not been removed. Closes GH-4970.
2019-12-05 09:28:41 +00:00
if (subject_ht) {
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
array_init(return_value);
/* For each subject entry, convert it to string, then perform replacement
and add the result to the return_value array. */
Merge branch 'PHP-7.4' * PHP-7.4: Fixed bug #79817
2020-07-10 12:07:59 +00:00
ZEND_HASH_FOREACH_KEY_VAL_IND(subject_ht, num_key, string_key, subject_entry) {
Fixed bug #75902 Don't special-case nested arrays/objects in str_replace(), instead perform a string cast on them as well. For arrays, this will always result in the usual conversion warning. This behavior is consistent with preg_replace(). If we didn't want to cast the array to string here, we should instead perform the replacement recursively. Silently copying it is just confusing.
2020-03-10 15:49:17 +00:00
zend_string *tmp_subject_str;
Fixed bug #71969 (str_replace returns an incorrect resulting array after a foreach by reference)
2016-04-06 02:19:24 +00:00
ZVAL_DEREF(subject_entry);
Fixed bug #75902 Don't special-case nested arrays/objects in str_replace(), instead perform a string cast on them as well. For arrays, this will always result in the usual conversion warning. This behavior is consistent with preg_replace(). If we didn't want to cast the array to string here, we should instead perform the replacement recursively. Silently copying it is just confusing.
2020-03-10 15:49:17 +00:00
subject_str = zval_get_tmp_string(subject_entry, &tmp_subject_str);
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
count += php_str_replace_in_subject(search_str, search_ht, replace_str, replace_ht, subject_str, &result, case_sensitivity);
Fixed bug #75902 Don't special-case nested arrays/objects in str_replace(), instead perform a string cast on them as well. For arrays, this will always result in the usual conversion warning. This behavior is consistent with preg_replace(). If we didn't want to cast the array to string here, we should instead perform the replacement recursively. Silently copying it is just confusing.
2020-03-10 15:49:17 +00:00
zend_tmp_string_release(tmp_subject_str);
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
/* Add to return array */
Use ZEND_HASH_FOREACH_* macros
2014-04-21 10:20:05 +00:00
if (string_key) {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
zend_hash_add_new(Z_ARRVAL_P(return_value), string_key, &result);
Use ZEND_HASH_FOREACH_* macros
2014-04-21 10:20:05 +00:00
} else {
Optimized str_[i]replace to aviod reallocations and repeatable conversions to lowercase
2015-01-13 20:23:50 +00:00
zend_hash_index_add_new(Z_ARRVAL_P(return_value), num_key, &result);
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
}
Use ZEND_HASH_FOREACH_* macros
2014-04-21 10:20:05 +00:00
} ZEND_HASH_FOREACH_END();
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
} else { /* if subject is not an array */
Use ZPP instead of custom type checks We can add these types as a native type declaration to stubs as a side-effect. Closes GH-6068
2020-09-04 12:23:43 +00:00
count = php_str_replace_in_subject(search_str, search_ht, replace_str, replace_ht, subject_str, return_value, case_sensitivity);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
if (zcount) {
Use ZEND_TRA_ASSIGN_REF_... macros for by reference arguments of internal functions.
2019-04-24 13:14:48 +00:00
ZEND_TRY_ASSIGN_REF_LONG(zcount, count);
Modify str_replace/str_ireplace to report how many needles were matched. (FR #8218)
2003-04-13 22:59:19 +00:00
}
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Replaces all occurrences of search in haystack with replace */
Unify str_replace and str_ireplace using INTERNAL_FUNCTION_PARAM_PASSTHRU -- reduce codebase/maintenance complexity
2003-01-30 20:09:19 +00:00
PHP_FUNCTION(str_replace)
{
php_str_replace_common(INTERNAL_FUNCTION_PARAM_PASSTHRU, 1);
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Replaces all occurrences of search in haystack with replace / case-insensitive */
Feature Request # 5919 - Addition of str_ireplace() Also removed deprecated BM str replace menthod Also rewrote php_str_to_str to use more processor/memory efficient method (ilia)
2003-01-30 05:00:41 +00:00
PHP_FUNCTION(str_ireplace)
{
Unify str_replace and str_ireplace using INTERNAL_FUNCTION_PARAM_PASSTHRU -- reduce codebase/maintenance complexity
2003-01-30 20:09:19 +00:00
php_str_replace_common(INTERNAL_FUNCTION_PARAM_PASSTHRU, 0);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Converts logical Hebrew text to visual text */
Remove hebrevc() function
2019-11-17 19:56:03 +00:00
PHP_FUNCTION(hebrev)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
Use const pointers
2018-03-17 11:45:22 +00:00
char *str, *heb_str, *target;
const char *tmp;
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
size_t block_start, block_end, block_type, block_length, i;
Fixed compilation warnings
2016-06-23 09:47:06 +00:00
zend_long max_chars=0, char_count;
size_t begin, end, orig_begin;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t str_len;
Fixed momory leak in heberv
2014-02-25 05:04:00 +00:00
zend_string *broken_str;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STRING(str, str_len)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(max_chars)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
if (str_len == 0) {
hebrev/hebrevc: Don't return false for empty string There is nothing special about the empty string in this context, we should pass it through unchanged.
2019-05-28 11:07:03 +00:00
RETURN_EMPTY_STRING();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
New parameter parsing API
2008-06-25 12:16:17 +00:00
tmp = str;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
block_start=block_end=0;
New parameter parsing API
2008-06-25 12:16:17 +00:00
heb_str = (char *) emalloc(str_len+1);
target = heb_str+str_len;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
*target = 0;
target--;
block_length=0;
if (isheb(*tmp)) {
block_type = _HEB_BLOCK_TYPE_HEB;
} else {
block_type = _HEB_BLOCK_TYPE_ENG;
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
do {
ws fixes.
2002-10-02 17:56:04 +00:00
if (block_type == _HEB_BLOCK_TYPE_HEB) {
New parameter parsing API
2008-06-25 12:16:17 +00:00
while ((isheb((int)*(tmp+1)) || _isblank((int)*(tmp+1)) || ispunct((int)*(tmp+1)) || (int)*(tmp+1)=='\n' ) && block_end<str_len-1) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
tmp++;
block_end++;
block_length++;
}
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
for (i = block_start+1; i<= block_end+1; i++) {
*target = str[i-1];
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
switch (*target) {
case '(':
*target = ')';
break;
case ')':
*target = '(';
break;
- Fix bug 11244 (patch by "Tal Peer" <hazgul99@hotmail.com>). # Sorry for the delay :-)
2002-03-26 07:57:07 +00:00
case '[':
*target = ']';
break;
case ']':
*target = '[';
break;
case '{':
*target = '}';
break;
case '}':
*target = '{';
break;
case '<':
*target = '>';
break;
case '>':
*target = '<';
break;
case '\\':
*target = '/';
break;
case '/':
*target = '\\';
break;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
default:
break;
}
target--;
}
block_type = _HEB_BLOCK_TYPE_ENG;
} else {
New parameter parsing API
2008-06-25 12:16:17 +00:00
while (!isheb(*(tmp+1)) && (int)*(tmp+1)!='\n' && block_end < str_len-1) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
tmp++;
block_end++;
block_length++;
}
ws fixes.
2002-10-02 17:56:04 +00:00
while ((_isblank((int)*tmp) || ispunct((int)*tmp)) && *tmp!='/' && *tmp!='-' && block_end > block_start) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
tmp--;
block_end--;
}
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
for (i = block_end+1; i >= block_start+1; i--) {
*target = str[i-1];
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
target--;
}
block_type = _HEB_BLOCK_TYPE_HEB;
}
block_start=block_end+1;
New parameter parsing API
2008-06-25 12:16:17 +00:00
} while (block_end < str_len-1);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
master renames phase 1
2014-08-25 17:24:55 +00:00
broken_str = zend_string_alloc(str_len, 0);
Fixed momory leak in heberv
2014-02-25 05:04:00 +00:00
begin = end = str_len-1;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
target = ZSTR_VAL(broken_str);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
while (1) {
char_count=0;
Fix a couple compile warnings
2014-09-28 19:30:49 +00:00
while ((!max_chars || (max_chars > 0 && char_count < max_chars)) && begin > 0) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
char_count++;
begin--;
Fix unsigned comparisons and remove dead code Fix unsigned comparison Cleanup never executed block Fix unsigned comparison Fix unsigned comparison, diff can't be < 0 Fix unsigned comparison Fix unsigned comparison Remove dead code
2017-11-17 20:35:22 +00:00
if (_isnewline(heb_str[begin])) {
ws fixes.
2002-10-02 17:56:04 +00:00
while (begin > 0 && _isnewline(heb_str[begin-1])) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
begin--;
char_count++;
}
break;
}
}
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
if (max_chars >= 0 && char_count == max_chars) { /* try to avoid breaking words */
size_t new_char_count=char_count, new_begin=begin;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
ws fixes.
2002-10-02 17:56:04 +00:00
while (new_char_count > 0) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (_isblank(heb_str[new_begin]) || _isnewline(heb_str[new_begin])) {
break;
}
new_begin++;
new_char_count--;
}
ws fixes.
2002-10-02 17:56:04 +00:00
if (new_char_count > 0) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
begin=new_begin;
}
}
orig_begin=begin;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
if (_isblank(heb_str[begin])) {
heb_str[begin]='\n';
}
ws fixes.
2002-10-02 17:56:04 +00:00
while (begin <= end && _isnewline(heb_str[begin])) { /* skip leading newlines */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
begin++;
}
ws fixes.
2002-10-02 17:56:04 +00:00
for (i = begin; i <= end; i++) { /* copy content */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
*target = heb_str[i];
target++;
}
ws fixes.
2002-10-02 17:56:04 +00:00
for (i = orig_begin; i <= end && _isnewline(heb_str[i]); i++) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
*target = heb_str[i];
target++;
}
begin=orig_begin;
Fix unsigned comparisons and remove dead code Fix unsigned comparison Cleanup never executed block Fix unsigned comparison Fix unsigned comparison, diff can't be < 0 Fix unsigned comparison Fix unsigned comparison Remove dead code
2017-11-17 20:35:22 +00:00
if (begin == 0) {
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
*target = 0;
break;
}
begin--;
end=begin;
}
efree(heb_str);
Remove hebrevc() function
2019-11-17 19:56:03 +00:00
RETURN_NEW_STR(broken_str);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Converts newlines to HTML line breaks */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
PHP_FUNCTION(nl2br)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
MFH: Implemented feature request #34381 (nl2br() should have an option for XHTML/HTML compatible BR element)
2008-08-14 02:56:23 +00:00
/* in brief this inserts <br /> or <br> before matched regexp \n\r?|\r\n? */
Use const pointers
2018-03-17 11:45:22 +00:00
const char *tmp, *end;
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
Use const pointers
2018-03-17 11:45:22 +00:00
char *target;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t repl_cnt = 0;
MFH: Implemented feature request #34381 (nl2br() should have an option for XHTML/HTML compatible BR element)
2008-08-14 02:56:23 +00:00
zend_bool is_xhtml = 1;
Fixed memory leak in nlbr
2014-02-24 09:40:18 +00:00
zend_string *result;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
FAST ZPP for strtok() and nl2br() Looking at some template-driven code that calls these hundreds of times per request
2015-01-17 16:30:40 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Z_PARAM_BOOL(is_xhtml)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
tmp = ZSTR_VAL(str);
end = ZSTR_VAL(str) + ZSTR_LEN(str);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fix a typo
2009-11-06 09:02:52 +00:00
/* it is really faster to scan twice and allocate mem once instead of scanning once
- nl2br now handles different styles of line endings in one string. (Patch by Boian Bonev <boian@bonev.com>) @- nl2br now handles all styles of line-endings in one string. (Boian, @ Derick)
2001-09-26 19:34:46 +00:00
and constantly reallocing */
New parameter parsing API
2008-06-25 12:16:17 +00:00
while (tmp < end) {
if (*tmp == '\r') {
if (*(tmp+1) == '\n') {
tmp++;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
repl_cnt++;
New parameter parsing API
2008-06-25 12:16:17 +00:00
} else if (*tmp == '\n') {
if (*(tmp+1) == '\r') {
tmp++;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
- nl2br now handles different styles of line endings in one string. (Patch by Boian Bonev <boian@bonev.com>) @- nl2br now handles all styles of line-endings in one string. (Boian, @ Derick)
2001-09-26 19:34:46 +00:00
repl_cnt++;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
tmp++;
- nl2br now handles different styles of line endings in one string. (Patch by Boian Bonev <boian@bonev.com>) @- nl2br now handles all styles of line-endings in one string. (Boian, @ Derick)
2001-09-26 19:34:46 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
- nl2br now handles different styles of line endings in one string. (Patch by Boian Bonev <boian@bonev.com>) @- nl2br now handles all styles of line-endings in one string. (Boian, @ Derick)
2001-09-26 19:34:46 +00:00
if (repl_cnt == 0) {
Unnecessary duplication
2015-07-02 03:44:28 +00:00
RETURN_STR_COPY(str);
- Fix crash bug (patch by <chocolateboy@lunomail.com>)
2001-09-24 08:33:51 +00:00
}
- Really fix nl2br now... it's actaulyl faster now
2001-09-09 12:55:48 +00:00
Fix potential integer overflow in nl2br The buffer size was calculated manually, thus creating integer overflows for very large inputs, e.g. nl2br(str_repeat("\n", 613566757)). The code now uses safe_emalloc, thus making the code throw an error instead of crashing.
2012-07-05 18:31:58 +00:00
{
size_t repl_len = is_xhtml ? (sizeof("<br />") - 1) : (sizeof("<br>") - 1);
MFH: Implemented feature request #34381 (nl2br() should have an option for XHTML/HTML compatible BR element)
2008-08-14 02:56:23 +00:00
Use safe alloc functions when calculations are made on sizes. Fixes bug #71449, bug #71450
2016-01-27 06:10:43 +00:00
result = zend_string_safe_alloc(repl_cnt, repl_len, ZSTR_LEN(str), 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
target = ZSTR_VAL(result);
Fix potential integer overflow in nl2br The buffer size was calculated manually, thus creating integer overflows for very large inputs, e.g. nl2br(str_repeat("\n", 613566757)). The code now uses safe_emalloc, thus making the code throw an error instead of crashing.
2012-07-05 18:31:58 +00:00
}
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
tmp = ZSTR_VAL(str);
fixes to string functions
2014-08-18 13:39:38 +00:00
while (tmp < end) {
switch (*tmp) {
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
case '\r':
case '\n':
*target++ = '<';
*target++ = 'b';
*target++ = 'r';
MFH: Implemented feature request #34381 (nl2br() should have an option for XHTML/HTML compatible BR element)
2008-08-14 02:56:23 +00:00
if (is_xhtml) {
*target++ = ' ';
*target++ = '/';
}
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
*target++ = '>';
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
fixes to string functions
2014-08-18 13:39:38 +00:00
if ((*tmp == '\r' && *(tmp+1) == '\n') || (*tmp == '\n' && *(tmp+1) == '\r')) {
fixed nl2br
2014-08-20 13:32:24 +00:00
*target++ = *tmp++;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
/* lack of a break; is intentional */
default:
fixes to string functions
2014-08-18 13:39:38 +00:00
*target++ = *tmp;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
fixes to string functions
2014-08-18 13:39:38 +00:00
tmp++;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fixed problems with the new nl2br() and beautified the code a little. Closes #19858.
2002-10-11 12:42:01 +00:00
*target = '\0';
- nl2br now handles different styles of line endings in one string. (Patch by Boian Bonev <boian@bonev.com>) @- nl2br now handles all styles of line-endings in one string. (Boian, @ Derick)
2001-09-26 19:34:46 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(result);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Strips HTML and PHP tags from a string */
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
PHP_FUNCTION(strip_tags)
{
Cleanup (2-nd round)
2014-04-15 17:56:30 +00:00
zend_string *buf;
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
zend_string *allow_str = NULL;
HashTable *allow_ht = NULL;
Use const pointers
2018-03-17 11:45:22 +00:00
const char *allowed_tags=NULL;
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
size_t allowed_tags_len=0;
Allow strip_tags with an array of allowed tagnames
2018-12-29 21:10:19 +00:00
smart_str tags_ss = {0};
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Consolidate new union type ZPP macro names They will now follow the canonical order of types. Older macros are left intact due to maintaining BC. Closes GH-6112
2020-09-10 22:23:54 +00:00
Z_PARAM_ARRAY_HT_OR_STR_OR_NULL(allow_ht, allow_str)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
if (allow_ht) {
zval *tmp;
zend_string *tag;
ZEND_HASH_FOREACH_VAL(allow_ht, tmp) {
tag = zval_get_string(tmp);
smart_str_appendc(&tags_ss, '<');
smart_str_append(&tags_ss, tag);
smart_str_appendc(&tags_ss, '>');
zend_string_release(tag);
} ZEND_HASH_FOREACH_END();
if (tags_ss.s) {
smart_str_0(&tags_ss);
allowed_tags = ZSTR_VAL(tags_ss.s);
allowed_tags_len = ZSTR_LEN(tags_ss.s);
Allow strip_tags with an array of allowed tagnames
2018-12-29 21:10:19 +00:00
}
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
} else if (allow_str) {
allowed_tags = ZSTR_VAL(allow_str);
allowed_tags_len = ZSTR_LEN(allow_str);
New parameter parsing API
2008-06-25 12:16:17 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
buf = zend_string_init(ZSTR_VAL(str), ZSTR_LEN(str), 0);
Remove state pointer argument from php_strip_tags
2020-01-27 15:13:36 +00:00
ZSTR_LEN(buf) = php_strip_tags_ex(ZSTR_VAL(buf), ZSTR_LEN(str), allowed_tags, allowed_tags_len, 0);
Allow strip_tags with an array of allowed tagnames
2018-12-29 21:10:19 +00:00
smart_str_free(&tags_ss);
Cleanup (avoid reallocatios and side effects in php_strip_tags)
2015-07-01 16:04:18 +00:00
RETURN_NEW_STR(buf);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
/* }}} */
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
static zend_string *try_setlocale_str(zend_long cat, zend_string *loc) {
Use const pointers
2018-03-17 11:45:22 +00:00
const char *retval;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
if (!strcmp("0", ZSTR_VAL(loc))) {
loc = NULL;
} else {
if (ZSTR_LEN(loc) >= 255) {
php_error_docref(NULL, E_WARNING, "Specified locale name is too long");
return NULL;
setlocale() will now accept multiple locale arguments, first match wins
2002-09-18 07:55:30 +00:00
}
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
}
Added missing safety checks for 2nd parameter of setlocale() function
2007-09-07 02:28:42 +00:00
Fix #77937: preg_match failed On some recent Windows systems, ext\pcre\tests\locales.phpt fails, because 'pt_PT' is accepted by `setlocale()`, but not properly supported by the ctype functions, which are used internally by PCRE2 to build the localized character tables. Since there appears to be no way to properly check whether a given locale is fully supported, but we want to minimize BC impact, we filter out typical Unix locale names, except for a few cases which have already been properly supported on Windows. This way code like setlocale(LC_ALL, 'de_DE.UTF-8', 'de_DE', 'German_Germany.1252'); should work like on older Windows systems. It should be noted that the locale names causing trouble are not (yet) documented as valid names anyway, see <https://docs.microsoft.com/en-us/cpp/c-runtime-library/locale-names-languages-and-country-region-strings?view=vs-2019>.
2019-04-30 14:10:04 +00:00
# ifndef PHP_WIN32
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
retval = setlocale(cat, loc ? ZSTR_VAL(loc) : NULL);
Fix #77937: preg_match failed On some recent Windows systems, ext\pcre\tests\locales.phpt fails, because 'pt_PT' is accepted by `setlocale()`, but not properly supported by the ctype functions, which are used internally by PCRE2 to build the localized character tables. Since there appears to be no way to properly check whether a given locale is fully supported, but we want to minimize BC impact, we filter out typical Unix locale names, except for a few cases which have already been properly supported on Windows. This way code like setlocale(LC_ALL, 'de_DE.UTF-8', 'de_DE', 'German_Germany.1252'); should work like on older Windows systems. It should be noted that the locale names causing trouble are not (yet) documented as valid names anyway, see <https://docs.microsoft.com/en-us/cpp/c-runtime-library/locale-names-languages-and-country-region-strings?view=vs-2019>.
2019-04-30 14:10:04 +00:00
# else
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
if (loc) {
/* BC: don't try /^[a-z]{2}_[A-Z]{2}($|\..*)/ except for /^u[ks]_U[KS]$/ */
char *locp = ZSTR_VAL(loc);
if (ZSTR_LEN(loc) >= 5 && locp[2] == '_'
&& locp[0] >= 'a' && locp[0] <= 'z' && locp[1] >= 'a' && locp[1] <= 'z'
&& locp[3] >= 'A' && locp[3] <= 'Z' && locp[4] >= 'A' && locp[4] <= 'Z'
&& (locp[5] == '\0' || locp[5] == '.')
&& !(locp[0] == 'u' && (locp[1] == 'k' || locp[1] == 's')
&& locp[3] == 'U' && (locp[4] == 'K' || locp[4] == 'S')
&& locp[5] == '\0')
) {
retval = NULL;
Fix #77937: preg_match failed On some recent Windows systems, ext\pcre\tests\locales.phpt fails, because 'pt_PT' is accepted by `setlocale()`, but not properly supported by the ctype functions, which are used internally by PCRE2 to build the localized character tables. Since there appears to be no way to properly check whether a given locale is fully supported, but we want to minimize BC impact, we filter out typical Unix locale names, except for a few cases which have already been properly supported on Windows. This way code like setlocale(LC_ALL, 'de_DE.UTF-8', 'de_DE', 'German_Germany.1252'); should work like on older Windows systems. It should be noted that the locale names causing trouble are not (yet) documented as valid names anyway, see <https://docs.microsoft.com/en-us/cpp/c-runtime-library/locale-names-languages-and-country-region-strings?view=vs-2019>.
2019-04-30 14:10:04 +00:00
} else {
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
retval = setlocale(cat, ZSTR_VAL(loc));
Fix #77937: preg_match failed On some recent Windows systems, ext\pcre\tests\locales.phpt fails, because 'pt_PT' is accepted by `setlocale()`, but not properly supported by the ctype functions, which are used internally by PCRE2 to build the localized character tables. Since there appears to be no way to properly check whether a given locale is fully supported, but we want to minimize BC impact, we filter out typical Unix locale names, except for a few cases which have already been properly supported on Windows. This way code like setlocale(LC_ALL, 'de_DE.UTF-8', 'de_DE', 'German_Germany.1252'); should work like on older Windows systems. It should be noted that the locale names causing trouble are not (yet) documented as valid names anyway, see <https://docs.microsoft.com/en-us/cpp/c-runtime-library/locale-names-languages-and-country-region-strings?view=vs-2019>.
2019-04-30 14:10:04 +00:00
}
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
} else {
retval = setlocale(cat, NULL);
}
Fix #77937: preg_match failed On some recent Windows systems, ext\pcre\tests\locales.phpt fails, because 'pt_PT' is accepted by `setlocale()`, but not properly supported by the ctype functions, which are used internally by PCRE2 to build the localized character tables. Since there appears to be no way to properly check whether a given locale is fully supported, but we want to minimize BC impact, we filter out typical Unix locale names, except for a few cases which have already been properly supported on Windows. This way code like setlocale(LC_ALL, 'de_DE.UTF-8', 'de_DE', 'German_Germany.1252'); should work like on older Windows systems. It should be noted that the locale names causing trouble are not (yet) documented as valid names anyway, see <https://docs.microsoft.com/en-us/cpp/c-runtime-library/locale-names-languages-and-country-region-strings?view=vs-2019>.
2019-04-30 14:10:04 +00:00
# endif
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
zend_update_current_locale();
if (!retval) {
return NULL;
}
if (loc) {
/* Remember if locale was changed */
size_t len = strlen(retval);
BG(locale_changed) = 1;
if (cat == LC_CTYPE || cat == LC_ALL) {
if (BG(ctype_string)) {
zend_string_release_ex(BG(ctype_string), 0);
}
if (len == 1 && *retval == 'C') {
/* C locale is represented as NULL. */
BG(ctype_string) = NULL;
return ZSTR_CHAR('C');
} else if (len == ZSTR_LEN(loc) && !memcmp(ZSTR_VAL(loc), retval, len)) {
BG(ctype_string) = zend_string_copy(loc);
return zend_string_copy(BG(ctype_string));
} else {
BG(ctype_string) = zend_string_init(retval, len, 0);
return zend_string_copy(BG(ctype_string));
Improved setlocale(). Eliminated locale comparison in ext/pcre if it's not necessary.
2014-12-19 15:00:16 +00:00
}
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
} else if (len == ZSTR_LEN(loc) && !memcmp(ZSTR_VAL(loc), retval, len)) {
return zend_string_copy(loc);
Reverted incorrect changes to setlocale()
2017-11-16 15:10:28 +00:00
}
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
}
return zend_string_init(retval, strlen(retval), 0);
}
static zend_string *try_setlocale_zval(zend_long cat, zval *loc_zv) {
zend_string *loc_str = zval_try_get_string(loc_zv);
zend_string *result = try_setlocale_str(cat, loc_str);
zend_string_release_ex(loc_str, 0);
return result;
}
/* {{{ Set locale information */
PHP_FUNCTION(setlocale)
{
zend_long cat;
zval *args = NULL;
int num_args;
ZEND_PARSE_PARAMETERS_START(2, -1)
Z_PARAM_LONG(cat)
Z_PARAM_VARIADIC('+', args, num_args)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
for (uint32_t i = 0; i < num_args; i++) {
if (Z_TYPE(args[i]) == IS_ARRAY) {
zval *elem;
ZEND_HASH_FOREACH_VAL_IND(Z_ARRVAL(args[i]), elem) {
zend_string *result = try_setlocale_zval(cat, elem);
if (EG(exception)) {
RETURN_THROWS();
}
if (result) {
RETURN_STR(result);
}
} ZEND_HASH_FOREACH_END();
setlocale() will now accept multiple locale arguments, first match wins
2002-09-18 07:55:30 +00:00
} else {
Clean up setlocale implementation Factor out the core logic into a separate function and drop the "clever" code that combines iteration through variadic arguments and arrays. This fixes bug #79829 as a side effect.
2020-08-24 15:09:33 +00:00
zend_string *result = try_setlocale_zval(cat, &args[i]);
if (EG(exception)) {
RETURN_THROWS();
}
if (result) {
RETURN_STR(result);
}
setlocale() will now accept multiple locale arguments, first match wins
2002-09-18 07:55:30 +00:00
}
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
setlocale() will now accept multiple locale arguments, first match wins
2002-09-18 07:55:30 +00:00
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
RETURN_FALSE;
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Parses GET/POST/COOKIE data and sets global variables */
A bunch of grunt work updating function entries and declarations.
1999-07-24 22:16:54 +00:00
PHP_FUNCTION(parse_str)
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
{
New parameter parsing API
2008-06-25 12:16:17 +00:00
char *arg;
standardise and clean up string functions parameter parsing
2009-08-13 06:44:50 +00:00
zval *arrayArg = NULL;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
char *res = NULL;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t arglen;
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Require second argument on (mb_)parse_str() This was deprecated in PHP 7.2 as part of https://wiki.php.net/rfc/deprecations_php_7_2.
2019-01-28 12:34:36 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_STRING(arg, arglen)
Implement typed properties RFC: https://wiki.php.net/rfc/typed_properties_v2 This is a squash of PR #3734, which is a squash of PR #3313. Co-authored-by: Bob Weinand <bobwei9@hotmail.com> Co-authored-by: Joe Watkins <krakjoe@php.net> Co-authored-by: Dmitry Stogov <dmitry@zend.com>
2019-01-07 11:28:51 +00:00
Z_PARAM_ZVAL(arrayArg)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
Add additional parameter to parse_str for saving result (thanks to John Bafford <dshadow@zort.net>) @ Added second parameter for parse_str to save result (John Bafford)
2000-09-11 14:50:26 +00:00
Require second argument on (mb_)parse_str() This was deprecated in PHP 7.2 as part of https://wiki.php.net/rfc/deprecations_php_7_2.
2019-01-28 12:34:36 +00:00
arrayArg = zend_try_array_init(arrayArg);
if (!arrayArg) {
Use RETURN_THROWS() in various places
2020-01-03 19:44:29 +00:00
RETURN_THROWS();
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
Require second argument on (mb_)parse_str() This was deprecated in PHP 7.2 as part of https://wiki.php.net/rfc/deprecations_php_7_2.
2019-01-28 12:34:36 +00:00
res = estrndup(arg, arglen);
sapi_module.treat_data(PARSE_STRING, res, arrayArg);
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
/* }}} */
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
#define PHP_TAG_BUF_SIZE 1023
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* {{{ php_tag_find
*
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
* Check if tag is in a set of tags
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
*
* states:
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
*
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
* 0 start tag
* 1 first non-whitespace char seen
*/
Cleanup (avoid reallocatios and side effects in php_strip_tags)
2015-07-01 16:04:18 +00:00
int php_tag_find(char *tag, size_t len, const char *set) {
Use const pointers
2018-03-17 11:45:22 +00:00
char c, *n;
const char *t;
- Not sure if len can be zero but it's better to check it once then rechecking for it every loop and having an extra variable do nothing.
2000-02-13 15:59:32 +00:00
int state=0, done=0;
MFH
2006-12-20 23:36:56 +00:00
char *norm;
Fix unsigned comparisons and remove dead code Fix unsigned comparison Cleanup never executed block Fix unsigned comparison Fix unsigned comparison, diff can't be < 0 Fix unsigned comparison Fix unsigned comparison Remove dead code
2017-11-17 20:35:22 +00:00
if (len == 0) {
MFH
2006-12-20 23:36:56 +00:00
return 0;
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
MFH
2006-12-20 23:36:56 +00:00
norm = emalloc(len+1);
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
n = norm;
t = tag;
c = tolower(*t);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/*
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
normalize the tag removing leading and trailing whitespace
and turn any <a whatever...> into just <a> and any </tag>
into <tag>
*/
ws fixes.
2002-10-02 17:56:04 +00:00
while (!done) {
switch (c) {
case '<':
*(n++) = c;
break;
case '>':
done =1;
break;
default:
if (!isspace((int)c)) {
if (state == 0) {
state=1;
- Fixed bug #53319 (strip_tags() may strip '<br />' incorrectly)
2010-11-16 22:16:44 +00:00
}
Fix #78814: strip_tags allows / in tag name => whitelist bypass When normalizing tags to check whether they are contained in the set of allowable tags, we must not strip slashes, unless they come immediately after the opening `<`, or immediately before the closing `>`.
2019-11-17 13:14:59 +00:00
if (c != '/' || (*(t-1) != '<' && *(t+1) != '>')) {
ws fixes.
2002-10-02 17:56:04 +00:00
*(n++) = c;
}
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
} else {
ws fixes.
2002-10-02 17:56:04 +00:00
if (state == 1)
done=1;
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
}
ws fixes.
2002-10-02 17:56:04 +00:00
break;
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
}
c = tolower(*(++t));
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
*(n++) = '>';
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
*n = '\0';
ws fixes.
2002-10-02 17:56:04 +00:00
if (strstr(set, norm)) {
- Beautify a bit.
2000-02-13 15:56:07 +00:00
done=1;
} else {
done=0;
}
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
efree(norm);
return done;
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
Remove state pointer argument from php_strip_tags
2020-01-27 15:13:36 +00:00
PHPAPI size_t php_strip_tags(char *rbuf, size_t len, const char *allow, size_t allow_len) /* {{{ */
Less hackish support for spaces at the start of tags within strip_tags()
2006-12-18 15:02:16 +00:00
{
Remove state pointer argument from php_strip_tags
2020-01-27 15:13:36 +00:00
return php_strip_tags_ex(rbuf, len, allow, allow_len, 0);
Less hackish support for spaces at the start of tags within strip_tags()
2006-12-18 15:02:16 +00:00
}
fix folding
2007-06-05 13:35:26 +00:00
/* }}} */
Less hackish support for spaces at the start of tags within strip_tags()
2006-12-18 15:02:16 +00:00
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* {{{ php_strip_tags
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
A simple little state-machine to strip out html and php tags
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
State 0 is the output state, State 1 means we are inside a
normal html tag and state 2 means we are inside a php tag.
The state variable is passed in to allow a function like fgetss
to maintain state across calls to the function.
lc holds the last significant character read and br is a bracket
counter.
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
When an allow string is passed in we keep track of the string
in state 1 and when the tag is closed check it against the
allow string to see if we should allow it.
Changed php_strip_tags() to check if <? was XML code.
2001-08-22 02:03:14 +00:00
swm: Added ability to strip <?xml tags without assuming it PHP
code.
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
*/
Remove state pointer argument from php_strip_tags
2020-01-27 15:13:36 +00:00
PHPAPI size_t php_strip_tags_ex(char *rbuf, size_t len, const char *allow, size_t allow_len, zend_bool allow_tag_spaces)
@- Fixed memory corruption in fgetss(), strip_tags() and gzgetss() (Zeev)
2000-02-08 21:29:18 +00:00
{
Use const pointers
2018-03-17 11:45:22 +00:00
char *tbuf, *tp, *rp, c, lc;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
const char *buf, *p, *end;
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
int br, depth=0, in_q = 0;
reduce size of stream struct 32 bytes are spared on 64-bit build
2016-09-15 14:48:29 +00:00
uint8_t state = 0;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
size_t pos;
Reduce (some) compile noise of 'unused variable' and 'may be used uninitialized' warnings.
2013-08-15 03:36:50 +00:00
char *allow_free = NULL;
Fix #70720
2015-12-22 15:25:51 +00:00
char is_xml = 0;
Make state parameter of php_strip_tags passed by reference. Move state tracking to stream structure.
2002-03-20 14:38:13 +00:00
Whitespace
2001-08-11 17:03:37 +00:00
buf = estrndup(rbuf, len);
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
end = buf + len;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
lc = '\0';
p = buf;
rp = rbuf;
br = 0;
ws fixes.
2002-10-02 17:56:04 +00:00
if (allow) {
Cleanup (avoid reallocatios and side effects in php_strip_tags)
2015-07-01 16:04:18 +00:00
allow_free = zend_str_tolower_dup_ex(allow, allow_len);
Avoid misc uninitialized variable warnings
2019-04-12 12:55:37 +00:00
allow = allow_free ? allow_free : allow;
Fixed bug #50847 (strip_tags() removes all tags greater then 1023 bytes long)
2010-02-01 12:59:08 +00:00
tbuf = emalloc(PHP_TAG_BUF_SIZE + 1);
Added optional allowable_tags arguments to strip_tags() and fgetss() to allow you to specify a string of tags that are not to be stripped Could have used flex for this, and Andrew sent me a flex file to do this, but I could do the same thing with only minor additions to the existing state machine and the resulting code is much smaller and tighter.
1999-09-20 15:50:56 +00:00
tp = tbuf;
kill another warning
1999-09-24 15:34:54 +00:00
} else {
tbuf = tp = NULL;
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
state_0:
if (p >= end) {
goto finish;
}
c = *p;
switch (c) {
case '\0':
break;
case '<':
if (in_q) {
break;
}
if (isspace(*(p + 1)) && !allow_tag_spaces) {
*(rp++) = c;
break;
}
lc = '<';
state = 1;
if (allow) {
if (tp - tbuf >= PHP_TAG_BUF_SIZE) {
pos = tp - tbuf;
tbuf = erealloc(tbuf, (tp - tbuf) + PHP_TAG_BUF_SIZE + 1);
tp = tbuf + pos;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
*(tp++) = '<';
}
p++;
goto state_1;
case '>':
if (depth) {
depth--;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
break;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
if (in_q) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
break;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
*(rp++) = c;
break;
default:
*(rp++) = c;
break;
}
p++;
goto state_0;
state_1:
if (p >= end) {
goto finish;
}
c = *p;
switch (c) {
case '\0':
break;
case '<':
if (in_q) {
break;
}
if (isspace(*(p + 1)) && !allow_tag_spaces) {
goto reg_char_1;
}
depth++;
break;
case '>':
if (depth) {
depth--;
break;
}
if (in_q) {
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
break;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
lc = '>';
Merge branch 'PHP-7.2' into PHP-7.3 * PHP-7.2: Update NEWS Fix bug #79037 (global buffer-overflow in `mbfl_filt_conv_big5_wchar`) Fix #79099: OOB read in php_strip_tags_ex Fix #79091: heap use-after-free in session_create_id()
2020-01-21 06:22:02 +00:00
if (is_xml && p >= buf + 1 && *(p -1) == '-') {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
break;
}
in_q = state = is_xml = 0;
if (allow) {
if (tp - tbuf >= PHP_TAG_BUF_SIZE) {
pos = tp - tbuf;
tbuf = erealloc(tbuf, (tp - tbuf) + PHP_TAG_BUF_SIZE + 1);
tp = tbuf + pos;
Fixed bugs #12989 and #12120
2002-09-20 11:48:47 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
*(tp++) = '>';
*tp='\0';
Avoid misc uninitialized variable warnings
2019-04-12 12:55:37 +00:00
if (php_tag_find(tbuf, tp-tbuf, allow)) {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
memcpy(rp, tbuf, tp-tbuf);
rp += tp-tbuf;
}
tp = tbuf;
}
p++;
goto state_0;
case '"':
case '\'':
if (p != buf && (!in_q || *p == in_q)) {
Fixed bug #40432 (strip_tags() fails with greater than in attribute).
2007-02-13 02:16:22 +00:00
if (in_q) {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
in_q = 0;
} else {
in_q = *p;
Fixed bug #40432 (strip_tags() fails with greater than in attribute).
2007-02-13 02:16:22 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
goto reg_char_1;
case '!':
/* JavaScript & Other HTML scripting languages */
Merge branch 'PHP-7.2' into PHP-7.3 * PHP-7.2: Update NEWS Fix bug #79037 (global buffer-overflow in `mbfl_filt_conv_big5_wchar`) Fix #79099: OOB read in php_strip_tags_ex Fix #79091: heap use-after-free in session_create_id()
2020-01-21 06:22:02 +00:00
if (p >= buf + 1 && *(p-1) == '<') {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
state = 3;
lc = c;
p++;
goto state_3;
} else {
goto reg_char_1;
}
break;
case '?':
Merge branch 'PHP-7.2' into PHP-7.3 * PHP-7.2: Update NEWS Fix bug #79037 (global buffer-overflow in `mbfl_filt_conv_big5_wchar`) Fix #79099: OOB read in php_strip_tags_ex Fix #79091: heap use-after-free in session_create_id()
2020-01-21 06:22:02 +00:00
if (p >= buf + 1 && *(p-1) == '<') {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
br=0;
state = 2;
p++;
goto state_2;
} else {
goto reg_char_1;
}
break;
default:
reg_char_1:
if (allow) {
if (tp - tbuf >= PHP_TAG_BUF_SIZE) {
pos = tp - tbuf;
tbuf = erealloc(tbuf, (tp - tbuf) + PHP_TAG_BUF_SIZE + 1);
tp = tbuf + pos;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
*(tp++) = c;
}
break;
}
p++;
goto state_1;
state_2:
if (p >= end) {
goto finish;
}
c = *p;
switch (c) {
case '(':
if (lc != '"' && lc != '\'') {
lc = '(';
br++;
}
break;
case ')':
if (lc != '"' && lc != '\'') {
lc = ')';
br--;
}
break;
case '>':
Fix #78003: strip_tags output change since PHP 7.3 A refactoring of the strip tags state machine[1] missed the special treatment of `depth > 0` when a `>` is encountered in state 2 or 3. We re-add it for BC reasons. [1] <http://git.php.net/?p=php-src.git;a=commit;h=5cf64742773ddbf9af69d962a4d12b567fcf0084>
2019-05-13 11:10:24 +00:00
if (depth) {
depth--;
break;
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
if (in_q) {
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
break;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Merge branch 'PHP-7.2' into PHP-7.3 * PHP-7.2: More checks for php_strip_tags_ex
2020-01-23 06:41:56 +00:00
if (!br && p >= buf + 1 && lc != '\"' && *(p-1) == '?') {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
in_q = state = 0;
tp = tbuf;
p++;
goto state_0;
}
break;
case '"':
case '\'':
Yet another check for php_strip_tags_ex()
2020-01-24 11:45:04 +00:00
if (p >= buf + 1 && *(p-1) != '\\') {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
if (lc == c) {
lc = '\0';
} else if (lc != '\\') {
lc = c;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Fix #78346: strip_tags no longer handling nested php tags When the strip tags state machine has been flattened, an if statement has mistakenly been treated as else if. We fix this, and also simplify a bit right away.
2019-08-02 15:03:20 +00:00
if (p != buf && (!in_q || *p == in_q)) {
Fixed bug #40637 (Regression in strip_tags() in relation so quote handling)
2007-02-27 00:15:21 +00:00
if (in_q) {
in_q = 0;
} else {
in_q = *p;
}
Fixed bug #40432 (strip_tags() fails with greater than in attribute).
2007-02-13 02:16:22 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
break;
case 'l':
case 'L':
/* swm: If we encounter '<?xml' then we shouldn't be in
* state == 2 (PHP). Switch back to HTML.
*/
if (state == 2 && p > buf+4
&& (*(p-1) == 'm' || *(p-1) == 'M')
&& (*(p-2) == 'x' || *(p-2) == 'X')
&& *(p-3) == '?'
&& *(p-4) == '<') {
state = 1; is_xml=1;
p++;
goto state_1;
}
break;
default:
break;
}
p++;
goto state_2;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
state_3:
if (p >= end) {
goto finish;
}
c = *p;
switch (c) {
case '>':
Fix #78003: strip_tags output change since PHP 7.3 A refactoring of the strip tags state machine[1] missed the special treatment of `depth > 0` when a `>` is encountered in state 2 or 3. We re-add it for BC reasons. [1] <http://git.php.net/?p=php-src.git;a=commit;h=5cf64742773ddbf9af69d962a4d12b567fcf0084>
2019-05-13 11:10:24 +00:00
if (depth) {
depth--;
break;
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
if (in_q) {
Fixed bug #7472 as well as many other bugs with strip_tags implementation.
2002-09-21 17:34:06 +00:00
break;
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
in_q = state = 0;
tp = tbuf;
p++;
goto state_0;
case '"':
case '\'':
if (p != buf && *(p-1) != '\\' && (!in_q || *p == in_q)) {
if (in_q) {
in_q = 0;
Fixed bug #23650 (strip_tags() removes hyphens)
2003-05-16 06:19:51 +00:00
} else {
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
in_q = *p;
Fixed buffer overflow issue.
2001-08-22 05:47:11 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
}
break;
case '-':
if (p >= buf + 2 && *(p-1) == '-' && *(p-2) == '!') {
state = 4;
p++;
goto state_4;
}
break;
case 'E':
case 'e':
/* !DOCTYPE exception */
if (p > buf+6
&& (*(p-1) == 'p' || *(p-1) == 'P')
&& (*(p-2) == 'y' || *(p-2) == 'Y')
&& (*(p-3) == 't' || *(p-3) == 'T')
&& (*(p-4) == 'c' || *(p-4) == 'C')
&& (*(p-5) == 'o' || *(p-5) == 'O')
&& (*(p-6) == 'd' || *(p-6) == 'D')) {
state = 1;
p++;
goto state_1;
}
break;
default:
break;
}
p++;
goto state_3;
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
state_4:
while (p < end) {
c = *p;
if (c == '>' && !in_q) {
if (p >= buf + 2 && *(p-1) == '-' && *(p-2) == '-') {
in_q = state = 0;
tp = tbuf;
p++;
goto state_0;
}
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
p++;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
}
Flatten strip_tags state machine (GCC overoptimized this function, producing enormpous amount of code).
2018-07-16 10:22:17 +00:00
finish:
Fixed possible buffer overflow of php_strip_tags
2003-02-24 22:19:36 +00:00
if (rp < rbuf + len) {
*rp = '\0';
}
Use const pointers
2018-03-17 11:45:22 +00:00
efree((void *)buf);
Avoid misc uninitialized variable warnings
2019-04-12 12:55:37 +00:00
if (tbuf) {
ws fixes.
2002-10-02 17:56:04 +00:00
efree(tbuf);
Avoid misc uninitialized variable warnings
2019-04-12 12:55:37 +00:00
}
if (allow_free) {
efree(allow_free);
Added concept of interned strings. All strings constants known at compile time are allocated in a single copy and never changed.
2010-04-20 11:05:54 +00:00
}
Made strip_tags() binary safe
2003-02-24 22:01:12 +00:00
return (size_t)(rp - rbuf);
* fixed some #if/#ifdef issues * hand-patched in php3 changes from 3.0.6 to HEAD in these files: fopen-wrappers.[ch] ext/standard/file.[ch] ext/standard/fsock.[ch] ext/standard/php3_string.h ext/standard/string.c * added some new file/socket macros for more readable code: FP_FGETS(buf,len,sock,fp,issock) FP_FREAD(buf,len,sock,fp,issock) FP_FEOF(sock,fp,issock) FP_FGETC(sock,fp,issock)
1999-06-16 17:06:53 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
First commit of re-structuring phase one. We have started using automake in sub-directories and started to move extension code into ext/<name>. For now, I have moved the "standard" extension (which is quite a mix of everything right now) and the GD extension into their own subdirs in ext/. The configure script now also runs configure in the libzend directory automatically and makes sure php4 and libzend use the same config.cache file. To avoid running configure in libzend, use the --no-recursion option. "make" in php4 also builds libzend now. The Apache module doesn't compile right now, but a fix for that is coming up.
1999-04-17 00:37:12 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Parse a CSV string into an array */
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
PHP_FUNCTION(str_getcsv)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
Allow empty $escape to eschew escaping CSV Albeit CSV is still a widespread data exchange format, it has never been officially standardized. There exists, however, the “informational” RFC 4180[1] which has no notion of escape characters, but rather defines `escaped` as strings enclosed in double-quotes where contained double-quotes have to be doubled. While this concept is supported by PHP's implementation (`$enclosure`), the `$escape` sometimes interferes, so that `fgetcsv()` is unable to correctly parse externally generated CSV, and `fputcsv()` is sometimes generating non-compliant CSV. Since PHP's `$escape` concept is availble for many years, we cannot drop it for BC reasons (even though many consider it as bug). Instead we allow to pass an empty string as `$escape` parameter to the respective functions, which results in ignoring/omitting any escaping, and as such is more inline with RFC 4180. It is noteworthy that this is almost no userland BC break, since formerly most functions did not accept an empty string, and failed in this case. The only exception was `str_getcsv()` which did accept an empty string, and used a backslash as escape character then (which appears to be unintended behavior, anyway). The changed functions are `fputcsv()`, `fgetcsv()` and `str_getcsv()`, and also the `::setCsvControl()`, `::getCsvControl()`, `::fputcsv()`, and `::fgetcsv()` methods of `SplFileObject`. The implementation also changes the type of the escape parameter of the PHP_APIs `php_fgetcsv()` and `php_fputcsv()` from `char` to `int`, where `PHP_CSV_NO_ESCAPE` means to ignore/omit escaping. The parameter accepts the same values as `isalpha()` and friends, i.e. “the value of which shall be representable as an `unsigned char` or shall equal the value of the macro `EOF`. If the argument has any other value, the behavior is undefined.” This is a subtle BC break, since the character `chr(128)` has the value `-1` if `char` is signed, and so likely would be confused with `EOF` when converted to `int`. We consider this BC break to be acceptable, since it's rather unlikely that anybody uses `chr(128)` as escape character, and it easily can be fixed by casting all `escape` arguments to `unsigned char`. This patch implements the feature requests 38301[2] and 51496[3]. [1] <https://tools.ietf.org/html/rfc4180> [2] <https://bugs.php.net/bug.php?id=38301> [3] <https://bugs.php.net/bug.php?id=51496>
2018-12-02 15:28:18 +00:00
char delim = ',', enc = '"';
int esc = (unsigned char) '\\';
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
char *delim_str = NULL, *enc_str = NULL, *esc_str = NULL;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t delim_len = 0, enc_len = 0, esc_len = 0;
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 4)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Z_PARAM_STRING(delim_str, delim_len)
Z_PARAM_STRING(enc_str, enc_len)
Z_PARAM_STRING(esc_str, esc_len)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
delim = delim_len ? delim_str[0] : delim;
enc = enc_len ? enc_str[0] : enc;
Allow empty $escape to eschew escaping CSV Albeit CSV is still a widespread data exchange format, it has never been officially standardized. There exists, however, the “informational” RFC 4180[1] which has no notion of escape characters, but rather defines `escaped` as strings enclosed in double-quotes where contained double-quotes have to be doubled. While this concept is supported by PHP's implementation (`$enclosure`), the `$escape` sometimes interferes, so that `fgetcsv()` is unable to correctly parse externally generated CSV, and `fputcsv()` is sometimes generating non-compliant CSV. Since PHP's `$escape` concept is availble for many years, we cannot drop it for BC reasons (even though many consider it as bug). Instead we allow to pass an empty string as `$escape` parameter to the respective functions, which results in ignoring/omitting any escaping, and as such is more inline with RFC 4180. It is noteworthy that this is almost no userland BC break, since formerly most functions did not accept an empty string, and failed in this case. The only exception was `str_getcsv()` which did accept an empty string, and used a backslash as escape character then (which appears to be unintended behavior, anyway). The changed functions are `fputcsv()`, `fgetcsv()` and `str_getcsv()`, and also the `::setCsvControl()`, `::getCsvControl()`, `::fputcsv()`, and `::fgetcsv()` methods of `SplFileObject`. The implementation also changes the type of the escape parameter of the PHP_APIs `php_fgetcsv()` and `php_fputcsv()` from `char` to `int`, where `PHP_CSV_NO_ESCAPE` means to ignore/omit escaping. The parameter accepts the same values as `isalpha()` and friends, i.e. “the value of which shall be representable as an `unsigned char` or shall equal the value of the macro `EOF`. If the argument has any other value, the behavior is undefined.” This is a subtle BC break, since the character `chr(128)` has the value `-1` if `char` is signed, and so likely would be confused with `EOF` when converted to `int`. We consider this BC break to be acceptable, since it's rather unlikely that anybody uses `chr(128)` as escape character, and it easily can be fixed by casting all `escape` arguments to `unsigned char`. This patch implements the feature requests 38301[2] and 51496[3]. [1] <https://tools.ietf.org/html/rfc4180> [2] <https://bugs.php.net/bug.php?id=38301> [3] <https://bugs.php.net/bug.php?id=51496>
2018-12-02 15:28:18 +00:00
if (esc_str != NULL) {
esc = esc_len ? (unsigned char) esc_str[0] : PHP_CSV_NO_ESCAPE;
}
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
php_fgetcsv(NULL, delim, enc, esc, ZSTR_LEN(str), ZSTR_VAL(str), return_value);
- MFH: Added str_getcsv()
2008-11-02 18:24:34 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns the input string repeat mult times */
(PHP str_repeat) New function.
1999-10-27 22:06:05 +00:00
PHP_FUNCTION(str_repeat)
{
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
zend_string *input_str; /* Input string */
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long mult; /* Multiplier */
Use better data structures (incomplete; able to run bench.php)
2014-02-18 12:27:38 +00:00
zend_string *result; /* Resulting string */
Fixed overflow on 64bit systems in str_repeat() and wordwrap(). Disabled CURLOPT_FOLLOWLOCATION in curl when open_basedir or safe_mode are enabled. # Patches by Stefan E.
2006-08-10 14:40:13 +00:00
size_t result_len; /* Length of the resulting string */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(input_str)
Use new param API in standard
2016-12-31 01:09:35 +00:00
Z_PARAM_LONG(mult)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
New parameter parsing API
2008-06-25 12:16:17 +00:00
if (mult < 0) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(2, "must be greater than or equal to 0");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
(PHP str_repeat) New function.
1999-10-27 22:06:05 +00:00
}
/* Don't waste our time if it's empty */
str_repeat() should be able to handle multipliers of 0 gracefully, IMHO ... Hope no one disagrees. :)
2000-09-05 18:25:58 +00:00
/* ... or if the multiplier is zero */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(input_str) == 0 || mult == 0)
New parameter parsing API
2008-06-25 12:16:17 +00:00
RETURN_EMPTY_STRING();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* Initialize the result string */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
result = zend_string_safe_alloc(ZSTR_LEN(input_str), mult, 0, 0);
result_len = ZSTR_LEN(input_str) * mult;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Fixed a bug in the new implementation of str_repeat()
2002-10-04 17:10:51 +00:00
/* Heavy optimization for situations where input string is 1 byte long */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (ZSTR_LEN(input_str) == 1) {
memset(ZSTR_VAL(result), *ZSTR_VAL(input_str), mult);
Optimized str_repeat() function, it is now 10x faster on 1 byte multipliers and 5-6x faster on multi-byte multipliers.
2002-10-03 13:59:31 +00:00
} else {
Use const pointers
2018-03-17 11:45:22 +00:00
const char *s, *ee;
char *e;
several fixes - - param parsing Z_PARAM_STR vs Z_PARAM_STRING - some functions for new params - etc
2014-08-16 10:55:13 +00:00
ptrdiff_t l=0;
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(result), ZSTR_VAL(input_str), ZSTR_LEN(input_str));
s = ZSTR_VAL(result);
e = ZSTR_VAL(result) + ZSTR_LEN(input_str);
ee = ZSTR_VAL(result) + result_len;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Optimized str_repeat() function, it is now 10x faster on 1 byte multipliers and 5-6x faster on multi-byte multipliers.
2002-10-03 13:59:31 +00:00
while (e<ee) {
l = (e-s) < (ee-e) ? (e-s) : (ee-e);
memmove(e, s, l);
e += l;
}
(PHP str_repeat) New function.
1999-10-27 22:06:05 +00:00
}
Optimized str_repeat() function, it is now 10x faster on 1 byte multipliers and 5-6x faster on multi-byte multipliers.
2002-10-03 13:59:31 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[result_len] = '\0';
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(result);
(PHP str_repeat) New function.
1999-10-27 22:06:05 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns info about what characters are used in input */
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
PHP_FUNCTION(count_chars)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *input;
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
int chars[256];
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long mymode=0;
Use const pointers
2018-03-17 11:45:22 +00:00
const unsigned char *buf;
fixes to string functions
2014-08-18 13:39:38 +00:00
int inx;
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
char retstr[256];
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t retlen=0;
size_t tmp = 0;
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STR(input)
Z_PARAM_OPTIONAL
Use new param API in standard
2016-12-31 01:09:35 +00:00
Z_PARAM_LONG(mymode)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
if (mymode < 0 || mymode > 4) {
Improve some TypeError and ValueError messages Closes GH-5377
2020-04-13 08:41:23 +00:00
zend_argument_value_error(2, "must be between 1 and 4 (inclusive)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
}
New parameter parsing API
2008-06-25 12:16:17 +00:00
Use const pointers
2018-03-17 11:45:22 +00:00
buf = (const unsigned char *) ZSTR_VAL(input);
Whitespace
2001-08-11 17:03:37 +00:00
memset((void*) chars, 0, sizeof(chars));
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
while (tmp < ZSTR_LEN(input)) {
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
chars[*buf]++;
buf++;
fixed count_chars()
2014-08-20 13:51:45 +00:00
tmp++;
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
}
if (mymode < 3) {
array_init(return_value);
}
ws fixes.
2002-10-02 17:56:04 +00:00
for (inx = 0; inx < 256; inx++) {
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
switch (mymode) {
ws fixes. #again :)
2002-10-03 18:15:18 +00:00
case 0:
master renames phase 1
2014-08-25 17:24:55 +00:00
add_index_long(return_value, inx, chars[inx]);
ws fixes. #again :)
2002-10-03 18:15:18 +00:00
break;
case 1:
if (chars[inx] != 0) {
master renames phase 1
2014-08-25 17:24:55 +00:00
add_index_long(return_value, inx, chars[inx]);
ws fixes. #again :)
2002-10-03 18:15:18 +00:00
}
break;
case 2:
if (chars[inx] == 0) {
master renames phase 1
2014-08-25 17:24:55 +00:00
add_index_long(return_value, inx, chars[inx]);
ws fixes. #again :)
2002-10-03 18:15:18 +00:00
}
break;
case 3:
if (chars[inx] != 0) {
retstr[retlen++] = inx;
}
break;
case 4:
if (chars[inx] == 0) {
retstr[retlen++] = inx;
}
break;
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
}
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
if (mymode >= 3 && mymode <= 4) {
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
RETURN_STRINGL(retstr, retlen);
@- implemented count_chars(). (Thies)
1999-12-14 03:52:12 +00:00
}
}
/* }}} */
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ php_strnatcmp */
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
static void php_strnatcmp(INTERNAL_FUNCTION_PARAMETERS, int fold_case)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *s1, *s2;
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(s1)
Z_PARAM_STR(s2)
ZEND_PARSE_PARAMETERS_END();
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(strnatcmp_ex(ZSTR_VAL(s1), ZSTR_LEN(s1),
ZSTR_VAL(s2), ZSTR_LEN(s2),
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
fold_case));
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
PHPAPI int string_natural_compare_function_ex(zval *result, zval *op1, zval *op2, zend_bool case_insensitive) /* {{{ */
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
{
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_string *tmp_str1, *tmp_str2;
zend_string *str1 = zval_get_tmp_string(op1, &tmp_str1);
zend_string *str2 = zval_get_tmp_string(op2, &tmp_str2);
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZVAL_LONG(result, strnatcmp_ex(ZSTR_VAL(str1), ZSTR_LEN(str1), ZSTR_VAL(str2), ZSTR_LEN(str2), case_insensitive));
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
Avoid unnecessary reference-counting on strings.
2017-11-16 14:09:32 +00:00
zend_tmp_string_release(tmp_str1);
zend_tmp_string_release(tmp_str2);
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
return SUCCESS;
}
/* }}} */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
PHPAPI int string_natural_case_compare_function(zval *result, zval *op1, zval *op2) /* {{{ */
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
{
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
return string_natural_compare_function_ex(result, op1, op2, 1);
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
}
/* }}} */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
PHPAPI int string_natural_compare_function(zval *result, zval *op1, zval *op2) /* {{{ */
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
{
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
return string_natural_compare_function_ex(result, op1, op2, 0);
add support for SORT_NATURAL and SORT_FLAG_CASE in array sort functions
2011-08-29 20:23:34 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns the result of string comparison using 'natural' algorithm */
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
PHP_FUNCTION(strnatcmp)
{
php_strnatcmp(INTERNAL_FUNCTION_PARAM_PASSTHRU, 0);
}
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
/* }}} */
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns numeric formatting information based on the current locale */
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
PHP_FUNCTION(localeconv)
{
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zval grouping, mon_grouping;
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
int len, i;
Convert zend_parse_parameters_none() to fast ZPP in standard lib Closes GH-5037
2019-12-29 20:14:36 +00:00
ZEND_PARSE_PARAMETERS_NONE();
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
Remove pointless checks for array_init() return values.
2002-12-05 22:28:02 +00:00
array_init(return_value);
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
array_init(&grouping);
array_init(&mon_grouping);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
{
struct lconv currlocdata;
localeconv_r( &currlocdata );
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
/* Grab the grouping data out of the array */
fix datatype mismatch warnings
2014-10-24 11:12:59 +00:00
len = (int)strlen(currlocdata.grouping);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
ws fixes.
2002-10-02 17:56:04 +00:00
for (i = 0; i < len; i++) {
master renames phase 1
2014-08-25 17:24:55 +00:00
add_index_long(&grouping, i, currlocdata.grouping[i]);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
}
/* Grab the monetary grouping data out of the array */
fix datatype mismatch warnings
2014-10-24 11:12:59 +00:00
len = (int)strlen(currlocdata.mon_grouping);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
ws fixes.
2002-10-02 17:56:04 +00:00
for (i = 0; i < len; i++) {
master renames phase 1
2014-08-25 17:24:55 +00:00
add_index_long(&mon_grouping, i, currlocdata.mon_grouping[i]);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
}
Cleanup (1-st round)
2014-04-15 11:40:40 +00:00
add_assoc_string(return_value, "decimal_point", currlocdata.decimal_point);
add_assoc_string(return_value, "thousands_sep", currlocdata.thousands_sep);
add_assoc_string(return_value, "int_curr_symbol", currlocdata.int_curr_symbol);
add_assoc_string(return_value, "currency_symbol", currlocdata.currency_symbol);
add_assoc_string(return_value, "mon_decimal_point", currlocdata.mon_decimal_point);
add_assoc_string(return_value, "mon_thousands_sep", currlocdata.mon_thousands_sep);
add_assoc_string(return_value, "positive_sign", currlocdata.positive_sign);
add_assoc_string(return_value, "negative_sign", currlocdata.negative_sign);
master renames phase 1
2014-08-25 17:24:55 +00:00
add_assoc_long( return_value, "int_frac_digits", currlocdata.int_frac_digits);
add_assoc_long( return_value, "frac_digits", currlocdata.frac_digits);
add_assoc_long( return_value, "p_cs_precedes", currlocdata.p_cs_precedes);
add_assoc_long( return_value, "p_sep_by_space", currlocdata.p_sep_by_space);
add_assoc_long( return_value, "n_cs_precedes", currlocdata.n_cs_precedes);
add_assoc_long( return_value, "n_sep_by_space", currlocdata.n_sep_by_space);
add_assoc_long( return_value, "p_sign_posn", currlocdata.p_sign_posn);
add_assoc_long( return_value, "n_sign_posn", currlocdata.n_sign_posn);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
}
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
zend_hash_str_update(Z_ARRVAL_P(return_value), "grouping", sizeof("grouping")-1, &grouping);
zend_hash_str_update(Z_ARRVAL_P(return_value), "mon_grouping", sizeof("mon_grouping")-1, &mon_grouping);
Added localeconv() function. Operates similarly to localeconv() defined in locale.h. Should work well under non-ZTS as well as ZTS.
2001-01-14 16:36:30 +00:00
}
vim-6 does folding - clean up a bunch of missing folding tags plus some misguided RINIT and RSHUTDOWN calls in a few fringe extensions
2001-06-05 13:12:10 +00:00
/* }}} */
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns the result of case-insensitive string comparison using 'natural' algorithm */
Added natural comparison/sorting routines using code from Martin Pool. @- Added natural comparison/sorting routines strnatcmp(), strnatcasecmp(), @ natsort(), and natcasesort(). These are useful for comparing and sorting @ strings that contain numbers. Based on the code from Martin Pool @ <mbp@humbug.org.au>. See http://www.linuxcare.com.au/projects/natsort/ @ for more info on natural sorting. (Andrei)
2000-04-12 19:39:02 +00:00
PHP_FUNCTION(strnatcasecmp)
{
php_strnatcmp(INTERNAL_FUNCTION_PARAM_PASSTHRU, 1);
}
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns the number of times a substring occurs in the string */
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
PHP_FUNCTION(substr_count)
{
New parameter parsing API
2008-06-25 12:16:17 +00:00
char *haystack, *needle;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long offset = 0, length = 0;
Accept null lengths for substr functions() If a null $length is passed to any of these functions, behave as if no parameter was passed: - substr() - substr_count() - substr_compare() - iconv_substr()
2019-10-21 18:49:08 +00:00
zend_bool length_is_null = 1;
Fixed bug #75781 (substr_count incorrect result)
2018-01-09 08:20:19 +00:00
zend_long count = 0;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t haystack_len, needle_len;
Use const pointers
2018-03-17 11:45:22 +00:00
const char *p, *endp;
char cmp;
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 4)
Z_PARAM_STRING(haystack, haystack_len)
Z_PARAM_STRING(needle, needle_len)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(offset)
Accept null lengths for substr functions() If a null $length is passed to any of these functions, behave as if no parameter was passed: - substr() - substr_count() - substr_compare() - iconv_substr()
2019-10-21 18:49:08 +00:00
Z_PARAM_LONG_OR_NULL(length, length_is_null)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
if (needle_len == 0) {
Promote a few remaining errors in ext/standard Closes GH-6110
2020-09-10 11:28:18 +00:00
zend_argument_value_error(2, "cannot be empty");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
p = haystack;
endp = p + haystack_len;
if (offset < 0) {
substr_count(): Support negative values for '$offset' and '$length' parameters.
2015-07-23 02:01:23 +00:00
offset += (zend_long)haystack_len;
New parameter parsing API
2008-06-25 12:16:17 +00:00
}
substr_count(): Support negative values for '$offset' and '$length' parameters.
2015-07-23 02:01:23 +00:00
if ((offset < 0) || ((size_t)offset > haystack_len)) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($haystack)");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
New parameter parsing API
2008-06-25 12:16:17 +00:00
}
p += offset;
Accept null lengths for substr functions() If a null $length is passed to any of these functions, behave as if no parameter was passed: - substr() - substr_count() - substr_compare() - iconv_substr()
2019-10-21 18:49:08 +00:00
if (!length_is_null) {
New parameter parsing API
2008-06-25 12:16:17 +00:00
Fixed bug #74041
2017-02-03 16:54:39 +00:00
if (length < 0) {
substr_count(): Support negative values for '$offset' and '$length' parameters.
2015-07-23 02:01:23 +00:00
length += (haystack_len - offset);
Added offset & length parameters to substr_count() function.
2005-06-18 18:23:12 +00:00
}
Fixed bug #74041
2017-02-03 16:54:39 +00:00
if (length < 0 || ((size_t)length > (haystack_len - offset))) {
Promote a few remaining errors in ext/standard Closes GH-6110
2020-09-10 11:28:18 +00:00
zend_argument_value_error(4, "must be contained in argument #1 ($haystack)");
RETURN_THROWS();
Added offset & length parameters to substr_count() function.
2005-06-18 18:23:12 +00:00
}
New parameter parsing API
2008-06-25 12:16:17 +00:00
endp = p + length;
Added offset & length parameters to substr_count() function.
2005-06-18 18:23:12 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
if (needle_len == 1) {
cmp = needle[0];
make substr_count() 30% when counting instances of 1 byte long strings.
2005-06-19 16:31:51 +00:00
while ((p = memchr(p, cmp, endp - p))) {
count++;
p++;
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
}
} else {
Use better data structures (incomplete)
2014-02-10 13:49:21 +00:00
while ((p = (char*)php_memnstr(p, needle, needle_len, endp))) {
New parameter parsing API
2008-06-25 12:16:17 +00:00
p += needle_len;
Optimizations to various php string functions substr_count(), strrev(), nl2br(), php_addslashes() and php_memnstr().
2002-10-09 13:37:02 +00:00
count++;
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
}
}
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(count);
@Added substr_count() from Peter Kovacs. (Andrei) # also some todo stuff
2000-04-26 00:29:59 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
/* }}} */
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Returns input string padded on the left or right to specified length with pad_string */
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
PHP_FUNCTION(str_pad)
{
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
/* Input arguments */
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *input; /* Input string */
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long pad_length; /* Length to pad to */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
/* Helper variables */
Fixed memory leak in str_pad
2014-02-24 09:44:42 +00:00
size_t num_pad_chars; /* Number of padding characters (total - input size) */
's' works with size_t round 2
2014-08-27 15:31:24 +00:00
char *pad_str = " "; /* Pointer to padding string */
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t pad_str_len = 1;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long pad_type_val = STR_PAD_RIGHT; /* The padding type value */
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t i, left_pad=0, right_pad=0;
Fixed memory leak in str_pad
2014-02-24 09:44:42 +00:00
zend_string *result = NULL; /* Resulting string */
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 4)
Z_PARAM_STR(input)
Z_PARAM_LONG(pad_length)
Z_PARAM_OPTIONAL
Z_PARAM_STRING(pad_str, pad_str_len)
Z_PARAM_LONG(pad_type_val)
ZEND_PARSE_PARAMETERS_END();
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
/* If resulting string turns out to be shorter than input string,
we simply copy the input and return. */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (pad_length < 0 || (size_t)pad_length <= ZSTR_LEN(input)) {
Avoding memory allocation
2018-02-26 08:48:49 +00:00
RETURN_STR_COPY(input);
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
}
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
if (pad_str_len == 0) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(3, "must be a non-empty string");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
New parameter parsing API
2008-06-25 12:16:17 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
New parameter parsing API
2008-06-25 12:16:17 +00:00
if (pad_type_val < STR_PAD_LEFT || pad_type_val > STR_PAD_BOTH) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(4, "must be STR_PAD_LEFT, STR_PAD_RIGHT, or STR_PAD_BOTH");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
num_pad_chars = pad_length - ZSTR_LEN(input);
Use zend_string_safe_alloc
2016-04-27 04:23:51 +00:00
result = zend_string_safe_alloc(1, ZSTR_LEN(input), num_pad_chars, 0);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_LEN(result) = 0;
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
/* We need to figure out the left/right padding lengths. */
switch (pad_type_val) {
case STR_PAD_RIGHT:
left_pad = 0;
right_pad = num_pad_chars;
break;
case STR_PAD_LEFT:
left_pad = num_pad_chars;
right_pad = 0;
break;
case STR_PAD_BOTH:
left_pad = num_pad_chars / 2;
right_pad = num_pad_chars - left_pad;
break;
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
}
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
/* First we pad on the left. */
for (i = 0; i < left_pad; i++)
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[ZSTR_LEN(result)++] = pad_str[i % pad_str_len];
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
/* Then we copy the input string. */
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
memcpy(ZSTR_VAL(result) + ZSTR_LEN(result), ZSTR_VAL(input), ZSTR_LEN(input));
ZSTR_LEN(result) += ZSTR_LEN(input);
@- Updated str_pad() to be able to pad on left/right/both sides. (Andrei)
2000-07-08 20:38:23 +00:00
/* Finally, we pad on the right. */
for (i = 0; i < right_pad; i++)
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[ZSTR_LEN(result)++] = pad_str[i % pad_str_len];
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
ZSTR_VAL(result)[ZSTR_LEN(result)] = '\0';
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
Refactored ZVAL flags usage to simplify various checks (e.g. Z_REFCOUNTED(), candidate for GC, etc)
2014-04-03 11:26:23 +00:00
RETURN_NEW_STR(result);
@- Added str_pad() for padding a string with an arbitrary string on left or @ right. (Andrei) Added str_pad() for padding a string with an arbitrary string on left or right. # With all those macros, the code is starting to eerily resemble Perl's guts.
2000-06-06 20:42:33 +00:00
}
Added sscanf() function.
2000-06-06 18:58:15 +00:00
/* }}} */
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Implements an ANSI C compatible sscanf */
Added sscanf() function.
2000-06-06 18:58:15 +00:00
PHP_FUNCTION(sscanf)
{
Use better data structures (incomplete)
2014-02-13 13:54:23 +00:00
zval *args = NULL;
- New parameter parsing API (zend_get_parameters_array_ex--) - Fixed tests
2008-08-16 21:32:41 +00:00
char *str, *format;
fixed strip_tags() and hebrev() to work with big strings
2014-08-28 17:35:37 +00:00
size_t str_len, format_len;
int result, num_args = 0;
cleanup sscanf() just a tiny bit, just so it looks like the rest of the code in the string.c file.
2001-09-03 04:49:36 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, -1)
Z_PARAM_STRING(str, str_len)
Z_PARAM_STRING(format, format_len)
Z_PARAM_VARIADIC('*', args, num_args)
ZEND_PARSE_PARAMETERS_END();
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
result = php_sscanf_internal(str, format, num_args, args, 0, return_value);
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Added sscanf() function.
2000-06-06 18:58:15 +00:00
if (SCAN_ERROR_WRONG_PARAM_COUNT == result) {
(PHP str_pad) Terminate string with a null.
2000-06-12 16:06:27 +00:00
WRONG_PARAM_COUNT;
Added sscanf() function.
2000-06-06 18:58:15 +00:00
}
}
/* }}} */
Fixed folder
2018-03-16 05:23:35 +00:00
/* static zend_string *php_str_rot13(zend_string *str) {{{ */
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
#ifdef __SSE2__
#include <emmintrin.h>
#endif
static zend_string *php_str_rot13(zend_string *str)
{
zend_string *ret;
const char *p, *e;
char *target;
if (UNEXPECTED(ZSTR_LEN(str) == 0)) {
return ZSTR_EMPTY_ALLOC();
}
ret = zend_string_alloc(ZSTR_LEN(str), 0);
p = ZSTR_VAL(str);
e = p + ZSTR_LEN(str);
target = ZSTR_VAL(ret);
#ifdef __SSE2__
if (e - p > 15) {
Avoid using _N in SSE code The _N constant is already defined in OpenSSL. Instead use some more explicit variable names.
2018-04-04 22:06:26 +00:00
const __m128i a_minus_1 = _mm_set1_epi8('a' - 1);
const __m128i m_plus_1 = _mm_set1_epi8('m' + 1);
const __m128i n_minus_1 = _mm_set1_epi8('n' - 1);
const __m128i z_plus_1 = _mm_set1_epi8('z' + 1);
const __m128i A_minus_1 = _mm_set1_epi8('A' - 1);
const __m128i M_plus_1 = _mm_set1_epi8('M' + 1);
const __m128i N_minus_1 = _mm_set1_epi8('N' - 1);
const __m128i Z_plus_1 = _mm_set1_epi8('Z' + 1);
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
const __m128i add = _mm_set1_epi8(13);
const __m128i sub = _mm_set1_epi8(-13);
do {
__m128i in, gt, lt, cmp, delta;
delta = _mm_setzero_si128();
in = _mm_loadu_si128((__m128i *)p);
Avoid using _N in SSE code The _N constant is already defined in OpenSSL. Instead use some more explicit variable names.
2018-04-04 22:06:26 +00:00
gt = _mm_cmpgt_epi8(in, a_minus_1);
lt = _mm_cmplt_epi8(in, m_plus_1);
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
cmp = _mm_and_si128(lt, gt);
if (_mm_movemask_epi8(cmp)) {
cmp = _mm_and_si128(cmp, add);
delta = _mm_or_si128(delta, cmp);
}
Avoid using _N in SSE code The _N constant is already defined in OpenSSL. Instead use some more explicit variable names.
2018-04-04 22:06:26 +00:00
gt = _mm_cmpgt_epi8(in, n_minus_1);
lt = _mm_cmplt_epi8(in, z_plus_1);
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
cmp = _mm_and_si128(lt, gt);
if (_mm_movemask_epi8(cmp)) {
cmp = _mm_and_si128(cmp, sub);
delta = _mm_or_si128(delta, cmp);
}
Avoid using _N in SSE code The _N constant is already defined in OpenSSL. Instead use some more explicit variable names.
2018-04-04 22:06:26 +00:00
gt = _mm_cmpgt_epi8(in, A_minus_1);
lt = _mm_cmplt_epi8(in, M_plus_1);
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
cmp = _mm_and_si128(lt, gt);
if (_mm_movemask_epi8(cmp)) {
cmp = _mm_and_si128(cmp, add);
delta = _mm_or_si128(delta, cmp);
}
Avoid using _N in SSE code The _N constant is already defined in OpenSSL. Instead use some more explicit variable names.
2018-04-04 22:06:26 +00:00
gt = _mm_cmpgt_epi8(in, N_minus_1);
lt = _mm_cmplt_epi8(in, Z_plus_1);
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
cmp = _mm_and_si128(lt, gt);
if (_mm_movemask_epi8(cmp)) {
cmp = _mm_and_si128(cmp, sub);
delta = _mm_or_si128(delta, cmp);
}
in = _mm_add_epi8(in, delta);
_mm_storeu_si128((__m128i *)target, in);
p += 16;
target += 16;
} while (e - p > 15);
}
#endif
while (p < e) {
if (*p >= 'a' && *p <= 'z') {
*target++ = 'a' + (((*p++ - 'a') + 13) % 26);
} else if (*p >= 'A' && *p <= 'Z') {
*target++ = 'A' + (((*p++ - 'A') + 13) % 26);
} else {
*target++ = *p++;
}
}
*target = '\0';
return ret;
}
/* }}} */
Implement filter API for streams. Filters can be stacked onto a stream; more details will follow in docs and on php-dev. Implement "string.rot13" filter Allows the following script: $fp = fopen("file.txt", "r"); stream_filter_prepend($fp, "string.rot13"); // File contents will be subject to a rot13 transformation before // being output. fpassthru($fp); fclose($fp);
2002-08-20 20:47:47 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Perform the rot13 transform on a string */
renamed rot13 to str_rot13()
2001-12-06 21:39:01 +00:00
PHP_FUNCTION(str_rot13)
ported rot13() from php3
2001-12-06 19:02:27 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *arg;
Fix for #18638 (str_rot13 bug)
2002-07-30 10:51:33 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(arg)
ZEND_PARSE_PARAMETERS_END();
- MFH: #36185 str_rot13(NULL) crash
2006-01-28 00:14:01 +00:00
Optimized str_rot13 with SSE2 instructions
2018-03-16 02:52:57 +00:00
RETURN_STR(php_str_rot13(arg));
ported rot13() from php3
2001-12-06 19:02:27 +00:00
}
/* }}} */
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
static void php_string_shuffle(char *str, zend_long len) /* {{{ */
str_shuffle() function added. Like shuffle() for arrays - however the algorithm for creating the permutation is quite simple. More like the implementation of shuffle() for 4.2.1 .
2002-09-25 18:06:05 +00:00
{
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long n_elems, rnd_idx, n_left;
Now str_shuffle() has the randomization characteristics of shuffle(). str_shuffle() won't return anymore the randomized string and thus will be consistent with shuffle(). Since this function is new to 4.3.0 no BC problems. #Thanks to Adam Trachtenberg for providing good example to check the #permuations distribution.
2002-10-08 09:52:59 +00:00
char temp;
/* The implementation is stolen from array_data_shuffle */
/* Thus the characteristics of the randomization are the same */
n_elems = len;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Now str_shuffle() has the randomization characteristics of shuffle(). str_shuffle() won't return anymore the randomized string and thus will be consistent with shuffle(). Since this function is new to 4.3.0 no BC problems. #Thanks to Adam Trachtenberg for providing good example to check the #permuations distribution.
2002-10-08 09:52:59 +00:00
if (n_elems <= 1) {
return;
}
n_left = n_elems;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Now str_shuffle() has the randomization characteristics of shuffle(). str_shuffle() won't return anymore the randomized string and thus will be consistent with shuffle(). Since this function is new to 4.3.0 no BC problems. #Thanks to Adam Trachtenberg for providing good example to check the #permuations distribution.
2002-10-08 09:52:59 +00:00
while (--n_left) {
Remove RAND_RANGE() macro The behavior of RANGE_RANGE() is 7.1 changed completely, from rescaling an already generated number to generating a number itself. Because of this str_shuffle() ended up generating two random numbers on every iteration. To avoid further misuse the function is dropped entirely. Extensions for PHP >= 7.1 should directly call php_mt_rand_range().
2017-12-03 21:35:50 +00:00
rnd_idx = php_mt_rand_range(0, n_left);
Now str_shuffle() has the randomization characteristics of shuffle(). str_shuffle() won't return anymore the randomized string and thus will be consistent with shuffle(). Since this function is new to 4.3.0 no BC problems. #Thanks to Adam Trachtenberg for providing good example to check the #permuations distribution.
2002-10-08 09:52:59 +00:00
if (rnd_idx != n_left) {
temp = str[n_left];
str[n_left] = str[rnd_idx];
str[rnd_idx] = temp;
}
}
str_shuffle() function added. Like shuffle() for arrays - however the algorithm for creating the permutation is quite simple. More like the implementation of shuffle() for 4.2.1 .
2002-09-25 18:06:05 +00:00
}
fix folding
2007-06-05 13:35:26 +00:00
/* }}} */
Now str_shuffle() has the randomization characteristics of shuffle(). str_shuffle() won't return anymore the randomized string and thus will be consistent with shuffle(). Since this function is new to 4.3.0 no BC problems. #Thanks to Adam Trachtenberg for providing good example to check the #permuations distribution.
2002-10-08 09:52:59 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Shuffles string. One permutation of all possible is created */
str_shuffle() function added. Like shuffle() for arrays - however the algorithm for creating the permutation is quite simple. More like the implementation of shuffle() for 4.2.1 .
2002-09-25 18:06:05 +00:00
PHP_FUNCTION(str_shuffle)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *arg;
New parameter parsing API
2008-06-25 12:16:17 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STR(arg)
ZEND_PARSE_PARAMETERS_END();
New parameter parsing API
2008-06-25 12:16:17 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETVAL_STRINGL(ZSTR_VAL(arg), ZSTR_LEN(arg));
master renames phase 1
2014-08-25 17:24:55 +00:00
if (Z_STRLEN_P(return_value) > 1) {
first shot remove TSRMLS_* things
2014-12-13 22:06:14 +00:00
php_string_shuffle(Z_STRVAL_P(return_value), (zend_long) Z_STRLEN_P(return_value));
Returning to the old behavior of str_shuffle() a bit. Now the passed string is left unmodified and the shuffled string is returned.
2002-10-09 10:33:22 +00:00
}
str_shuffle() function added. Like shuffle() for arrays - however the algorithm for creating the permutation is quite simple. More like the implementation of shuffle() for 4.2.1 .
2002-09-25 18:06:05 +00:00
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Counts the number of words inside a string. If format of 1 is specified,
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
then the function will return an array containing all the words
found inside the string. If format of 2 is specified, then the function
will return an associated array where the position of the word is the key
and the word itself is the value.
For the purpose of this function, 'word' is defined as a locale dependent
string containing alphabetic characters, which also may contain, but not start
with "'" and "-" characters.
*/
Renamed word_count to str_word_count to comply with naming conventions. Thanks Andi, for catching this oversight.
2002-10-17 22:44:44 +00:00
PHP_FUNCTION(str_word_count)
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
Use const pointers
2018-03-17 11:45:22 +00:00
char *char_list = NULL, ch[256];
const char *p, *e, *s;
first show to make 's' work with size_t
2014-08-27 13:31:48 +00:00
size_t char_list_len = 0, word_count = 0;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long type = 0;
A more performance friendly version of Andrey's patch.
2005-03-12 17:51:19 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 3)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(type)
Fix UNKNOWN default values in ext/standard Closes GH-6026
2020-08-20 22:42:44 +00:00
Z_PARAM_STRING_OR_NULL(char_list, char_list_len)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_END();
A more performance friendly version of Andrey's patch.
2005-03-12 17:51:19 +00:00
MFH: avoid reading str[-1], add warning when invalid format specified add test
2006-10-11 13:15:30 +00:00
switch(type) {
case 1:
case 2:
array_init(return_value);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(str)) {
Catch empty strings right away.
2006-10-11 14:19:55 +00:00
return;
}
MFH: avoid reading str[-1], add warning when invalid format specified add test
2006-10-11 13:15:30 +00:00
break;
case 0:
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(str)) {
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(0);
Catch empty strings right away.
2006-10-11 14:19:55 +00:00
}
MFH: avoid reading str[-1], add warning when invalid format specified add test
2006-10-11 13:15:30 +00:00
/* nothing to be done */
break;
default:
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(2, "must be a valid format value");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
}
Fixed bug #35427 (str_word_count() handles '-' incorrectly).
2005-11-29 16:14:19 +00:00
Catch empty strings right away.
2006-10-11 14:19:55 +00:00
if (char_list) {
Constify char * arguments of APIs Closes GH-5676.
2020-06-07 09:01:19 +00:00
php_charmask((const unsigned char *) char_list, char_list_len, ch);
Catch empty strings right away.
2006-10-11 14:19:55 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
p = ZSTR_VAL(str);
e = ZSTR_VAL(str) + ZSTR_LEN(str);
Catch empty strings right away.
2006-10-11 14:19:55 +00:00
Fixed bug #35427 (str_word_count() handles '-' incorrectly).
2005-11-29 16:14:19 +00:00
/* first character cannot be ' or -, unless explicitly allowed by the user */
if ((*p == '\'' && (!char_list || !ch['\''])) || (*p == '-' && (!char_list || !ch['-']))) {
p++;
}
/* last character cannot be -, unless explicitly allowed by the user */
Catch empty strings right away.
2006-10-11 14:19:55 +00:00
if (*(e - 1) == '-' && (!char_list || !ch['-'])) {
Fixed bug #35427 (str_word_count() handles '-' incorrectly).
2005-11-29 16:14:19 +00:00
e--;
}
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
while (p < e) {
Fixed bug #35427 (str_word_count() handles '-' incorrectly).
2005-11-29 16:14:19 +00:00
s = p;
MFH: fix #43863 (str_word_count() breaks on cyrillic "ya" in locale cp1251) patch by phprus at gmail dot com
2008-01-16 08:34:33 +00:00
while (p < e && (isalpha((unsigned char)*p) || (char_list && ch[(unsigned char)*p]) || *p == '\'' || *p == '-')) {
Fixed bug #35427 (str_word_count() handles '-' incorrectly).
2005-11-29 16:14:19 +00:00
p++;
}
if (p > s) {
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
switch (type)
{
case 1:
Cleanup (1-st round)
2014-04-15 11:40:40 +00:00
add_next_index_stringl(return_value, s, p - s);
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
break;
case 2:
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
add_index_stringl(return_value, (s - ZSTR_VAL(str)), s, p - s);
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
break;
default:
word_count++;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
break;
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
}
Fixed bug in str_word_count() when charlist if specified and "word" starts with a character found inside the charlist.
2005-07-23 19:44:13 +00:00
}
Fixed bug #35427 (str_word_count() handles '-' incorrectly).
2005-11-29 16:14:19 +00:00
p++;
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
}
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
if (!type) {
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(word_count);
Added word_count() function that allows counting of words inside a string. The function also allows the user to retrieve all the words from a string.
2002-10-17 03:27:19 +00:00
}
}
/* }}} */
str_shuffle() function added. Like shuffle() for arrays - however the algorithm for creating the permutation is quite simple. More like the implementation of shuffle() for 4.2.1 .
2002-09-25 18:06:05 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Convert a string to an array. If split_length is specified, break the string down into chunks each split_length characters long. */
style & ws fixes
2003-02-08 15:26:17 +00:00
PHP_FUNCTION(str_split)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *str;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long split_length = 1;
Use const pointers
2018-03-17 11:45:22 +00:00
const char *p;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t n_reg_segments;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 2)
Z_PARAM_STR(str)
Z_PARAM_OPTIONAL
Z_PARAM_LONG(split_length)
ZEND_PARSE_PARAMETERS_END();
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
if (split_length <= 0) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(2, "must be greater than 0");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (0 == ZSTR_LEN(str) || (size_t)split_length >= ZSTR_LEN(str)) {
fixes to string functions
2014-08-18 13:39:38 +00:00
array_init_size(return_value, 1);
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
add_next_index_stringl(return_value, ZSTR_VAL(str), ZSTR_LEN(str));
Slight optimization in str_split() when split length is the same or greater then the string length.
2004-10-20 22:44:43 +00:00
return;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
array_init_size(return_value, (uint32_t)(((ZSTR_LEN(str) - 1) / split_length) + 1));
fixes to string functions
2014-08-18 13:39:38 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
n_reg_segments = ZSTR_LEN(str) / split_length;
p = ZSTR_VAL(str);
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
while (n_reg_segments-- > 0) {
Cleanup (1-st round)
2014-04-15 11:40:40 +00:00
add_next_index_stringl(return_value, p, split_length);
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
p += split_length;
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (p != (ZSTR_VAL(str) + ZSTR_LEN(str))) {
add_next_index_stringl(return_value, p, (ZSTR_VAL(str) + ZSTR_LEN(str) - p));
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
}
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Search a string for any of a set of characters */
Added strpbrk(), which is essentially a wrapper around C's strpbrk function that allows searching through a string for a character list.
2003-02-11 22:47:26 +00:00
PHP_FUNCTION(strpbrk)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *haystack, *char_list;
Use const pointers
2018-03-17 11:45:22 +00:00
const char *haystack_ptr, *cl_ptr;
fix bug #55871 - Interruption in substr_replace()
2012-01-02 00:47:57 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(2, 2)
Z_PARAM_STR(haystack)
Z_PARAM_STR(char_list)
Remove most uses of ZEND_PARSE_PARAMETERS_END_EX() As ZPP now throws, it makes no sense to specify an explicit return value.
2019-10-07 08:02:18 +00:00
ZEND_PARSE_PARAMETERS_END();
Added strpbrk(), which is essentially a wrapper around C's strpbrk function that allows searching through a string for a character list.
2003-02-11 22:47:26 +00:00
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
if (!ZSTR_LEN(char_list)) {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(2, "must be a non-empty string");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Added strpbrk(), which is essentially a wrapper around C's strpbrk function that allows searching through a string for a character list.
2003-02-11 22:47:26 +00:00
}
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
for (haystack_ptr = ZSTR_VAL(haystack); haystack_ptr < (ZSTR_VAL(haystack) + ZSTR_LEN(haystack)); ++haystack_ptr) {
for (cl_ptr = ZSTR_VAL(char_list); cl_ptr < (ZSTR_VAL(char_list) + ZSTR_LEN(char_list)); ++cl_ptr) {
MFH: Fix bug #60801 (strpbrk() mishandles NUL byte). (Trunk commit: r322934).
2012-03-02 03:39:04 +00:00
if (*cl_ptr == *haystack_ptr) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_STRINGL(haystack_ptr, (ZSTR_VAL(haystack) + ZSTR_LEN(haystack) - haystack_ptr));
MFH: Fix bug #60801 (strpbrk() mishandles NUL byte). (Trunk commit: r322934).
2012-03-02 03:39:04 +00:00
}
}
Added strpbrk(), which is essentially a wrapper around C's strpbrk function that allows searching through a string for a character list.
2003-02-11 22:47:26 +00:00
}
MFH: Fix bug #60801 (strpbrk() mishandles NUL byte). (Trunk commit: r322934).
2012-03-02 03:39:04 +00:00
RETURN_FALSE;
Added strpbrk(), which is essentially a wrapper around C's strpbrk function that allows searching through a string for a character list.
2003-02-11 22:47:26 +00:00
}
/* }}} */
Added str_split() function. This function can be used to break down a string into an array.
2003-02-07 21:36:18 +00:00
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Binary safe optionally case insensitive comparison of 2 strings from an offset, up to length characters */
Added substr_compare().
2003-10-30 00:49:33 +00:00
PHP_FUNCTION(substr_compare)
{
fixes to string functions
2014-08-18 13:39:38 +00:00
zend_string *s1, *s2;
master renames phase 1
2014-08-25 17:24:55 +00:00
zend_long offset, len=0;
Fix bug #55451 Make substr_compare ignore the length if it's NULL. This allows to use the last parameter (case_insensitivity) with the default length.
2016-08-03 08:29:12 +00:00
zend_bool len_is_default=1;
Added substr_compare().
2003-10-30 00:49:33 +00:00
zend_bool cs=0;
master renames phase 3
2014-08-25 18:22:49 +00:00
size_t cmp_len;
Added substr_compare().
2003-10-30 00:49:33 +00:00
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(3, 5)
Z_PARAM_STR(s1)
Z_PARAM_STR(s2)
Z_PARAM_LONG(offset)
Z_PARAM_OPTIONAL
Add Z_PARAM_LONG_OR_NULL macro
2019-10-21 20:04:57 +00:00
Z_PARAM_LONG_OR_NULL(len, len_is_default)
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
Z_PARAM_BOOL(cs)
Remove most uses of ZEND_PARSE_PARAMETERS_END_EX() As ZPP now throws, it makes no sense to specify an explicit return value.
2019-10-07 08:02:18 +00:00
ZEND_PARSE_PARAMETERS_END();
Added substr_compare().
2003-10-30 00:49:33 +00:00
Fix bug #55451 Make substr_compare ignore the length if it's NULL. This allows to use the last parameter (case_insensitivity) with the default length.
2016-08-03 08:29:12 +00:00
if (!len_is_default && len <= 0) {
Fixed bug 67043
2014-04-08 23:33:55 +00:00
if (len == 0) {
master renames phase 1
2014-08-25 17:24:55 +00:00
RETURN_LONG(0L);
Fixed bug 67043
2014-04-08 23:33:55 +00:00
} else {
Improve argument error messages in ext/standard Closes GH-5198
2020-02-24 19:19:20 +00:00
zend_argument_value_error(4, "must be greater than or equal to 0");
Use RETURN_THROWS() when an exception is thrown Closes GH-5036
2019-12-29 19:54:57 +00:00
RETURN_THROWS();
Fixed bug 67043
2014-04-08 23:33:55 +00:00
}
Added substr_compare().
2003-10-30 00:49:33 +00:00
}
MFH: Fixed possible GLOBALS variable override when register_globals are ON. MFH: Fixed possible register_globals toggle via parse_str(). MFH: Fixed negative offset handling in substr_compare() function.
2005-09-28 22:39:52 +00:00
if (offset < 0) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
offset = ZSTR_LEN(s1) + offset;
fix possible substr_compare() crash add new tests
2006-04-25 12:48:42 +00:00
offset = (offset < 0) ? 0 : offset;
}
Fixed bug #77853
2019-04-08 09:40:50 +00:00
if ((size_t)offset > ZSTR_LEN(s1)) {
Promote a few remaining errors in ext/standard Closes GH-6110
2020-09-10 11:28:18 +00:00
zend_argument_value_error(3, "must be contained in argument #1 ($main_str)");
RETURN_THROWS();
MFH: Fixed possible GLOBALS variable override when register_globals are ON. MFH: Fixed possible register_globals toggle via parse_str(). MFH: Fixed negative offset handling in substr_compare() function.
2005-09-28 22:39:52 +00:00
}
Fixed compilation warnings
2016-06-23 09:47:06 +00:00
cmp_len = len ? (size_t)len : MAX(ZSTR_LEN(s2), (ZSTR_LEN(s1) - offset));
Added substr_compare().
2003-10-30 00:49:33 +00:00
if (!cs) {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(zend_binary_strncmp(ZSTR_VAL(s1) + offset, (ZSTR_LEN(s1) - offset), ZSTR_VAL(s2), ZSTR_LEN(s2), cmp_len));
Added substr_compare().
2003-10-30 00:49:33 +00:00
} else {
Use ZSTR_ API to access zend_string elements (this is just renaming without semantick changes).
2015-06-30 01:05:24 +00:00
RETURN_LONG(zend_binary_strncasecmp_l(ZSTR_VAL(s1) + offset, (ZSTR_LEN(s1) - offset), ZSTR_VAL(s2), ZSTR_LEN(s2), cmp_len));
Added substr_compare().
2003-10-30 00:49:33 +00:00
}
}
/* }}} */
Move utf8_encode and utf8_decode to ext/standard
2016-10-13 22:33:33 +00:00
/* {{{ */
static zend_string *php_utf8_encode(const char *s, size_t len)
{
size_t pos = len;
zend_string *str;
unsigned char c;
str = zend_string_safe_alloc(len, 2, 0, 0);
ZSTR_LEN(str) = 0;
while (pos > 0) {
/* The lower 256 codepoints of Unicode are identical to Latin-1,
* so we don't need to do any mapping here. */
c = (unsigned char)(*s);
if (c < 0x80) {
ZSTR_VAL(str)[ZSTR_LEN(str)++] = (char) c;
/* We only account for the single-byte and two-byte cases because
* we're only dealing with the first 256 Unicode codepoints. */
} else {
ZSTR_VAL(str)[ZSTR_LEN(str)++] = (0xc0 | (c >> 6));
ZSTR_VAL(str)[ZSTR_LEN(str)++] = (0x80 | (c & 0x3f));
}
pos--;
s++;
}
ZSTR_VAL(str)[ZSTR_LEN(str)] = '\0';
str = zend_string_truncate(str, ZSTR_LEN(str), 0);
return str;
}
/* }}} */
/* {{{ */
static zend_string *php_utf8_decode(const char *s, size_t len)
{
size_t pos = 0;
unsigned int c;
zend_string *str;
str = zend_string_alloc(len, 0);
ZSTR_LEN(str) = 0;
while (pos < len) {
int status = FAILURE;
c = php_next_utf8_char((const unsigned char*)s, (size_t) len, &pos, &status);
/* The lower 256 codepoints of Unicode are identical to Latin-1,
* so we don't need to do any mapping here beyond replacing non-Latin-1
* characters. */
if (status == FAILURE || c > 0xFFU) {
c = '?';
}
ZSTR_VAL(str)[ZSTR_LEN(str)++] = c;
}
ZSTR_VAL(str)[ZSTR_LEN(str)] = '\0';
if (ZSTR_LEN(str) < len) {
str = zend_string_truncate(str, ZSTR_LEN(str), 0);
}
return str;
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Encodes an ISO-8859-1 string to UTF-8 */
Move utf8_encode and utf8_decode to ext/standard
2016-10-13 22:33:33 +00:00
PHP_FUNCTION(utf8_encode)
{
char *arg;
size_t arg_len;
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STRING(arg, arg_len)
ZEND_PARSE_PARAMETERS_END();
Move utf8_encode and utf8_decode to ext/standard
2016-10-13 22:33:33 +00:00
RETURN_STR(php_utf8_encode(arg, arg_len));
}
/* }}} */
Remove proto comments from C files Closes GH-5758
2020-07-01 13:32:55 +00:00
/* {{{ Converts a UTF-8 encoded string to ISO-8859-1 */
Move utf8_encode and utf8_decode to ext/standard
2016-10-13 22:33:33 +00:00
PHP_FUNCTION(utf8_decode)
{
char *arg;
size_t arg_len;
Use new param API in standard/string
2016-12-31 00:40:45 +00:00
ZEND_PARSE_PARAMETERS_START(1, 1)
Z_PARAM_STRING(arg, arg_len)
ZEND_PARSE_PARAMETERS_END();
Move utf8_encode and utf8_decode to ext/standard
2016-10-13 22:33:33 +00:00
RETURN_STR(php_utf8_decode(arg, arg_len));
}
/* }}} */
Reference in New Issue Copy Permalink