Add UConverter class (ICU's UConverter API)

RFC at http://wiki.php.net/rfc/uconverter

ext/intl/config.m4

@@ -34,6 +34,7 @@ if test "$PHP_INTL" != "no"; then
     common/common_error.c \
 	common/common_enum.cpp \
 	common/common_date.cpp \
+    converter/converter.c \
     formatter/formatter.c \
     formatter/formatter_main.c \
     formatter/formatter_class.c \
@@ -86,6 +87,7 @@ if test "$PHP_INTL" != "no"; then
     idn/idn.c \
     $icu_spoof_src, $ext_shared,,$ICU_INCS -Wno-write-strings)
   PHP_ADD_BUILD_DIR($ext_builddir/collator)
+  PHP_ADD_BUILD_DIR($ext_builddir/converter)
   PHP_ADD_BUILD_DIR($ext_builddir/common)
   PHP_ADD_BUILD_DIR($ext_builddir/formatter)
   PHP_ADD_BUILD_DIR($ext_builddir/normalizer)

ext/intl/config.w32

@@ -26,6 +26,9 @@ if (PHP_INTL != "no") {
 				common_enum.cpp \
 				common_date.cpp \
 				", "intl");
+		ADD_SOURCES(configure_module_dirname + "/converter", "\
+				converter.c \
+				", "intl");
 		ADD_SOURCES(configure_module_dirname + "/formatter", "\
 				formatter.c \
 				formatter_attr.c \

ext/intl/converter/converter.h

@@ -0,0 +1,28 @@
+/*
+ +----------------------------------------------------------------------+
+ | PHP Version 5                                                        |
+ +----------------------------------------------------------------------+
+ | This source file is subject to version 3.01 of the PHP license,      |
+ | that is bundled with this package in the file LICENSE, and is        |
+ | available through the world-wide-web at the following url:           |
+ | http://www.php.net/license/3_01.txt                                  |
+ | 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.               |
+ +----------------------------------------------------------------------+
+ | Authors: Sara Golemon <pollita@php.net>                              |
+ +----------------------------------------------------------------------+
+*/
+
+#ifndef PHP_INTL_CONVERTER_H
+#define PHP_INTL_CONVERTER_H
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include "php.h"
+
+int php_converter_minit(INIT_FUNC_ARGS);
+
+#endif /* PHP_INTL_CONVERTER_H */

ext/intl/php_intl.c

@@ -34,6 +34,8 @@
 #include "collator/collator_create.h"
 #include "collator/collator_error.h"
 
+#include "converter/converter.h"
+
 #include "formatter/formatter.h"
 #include "formatter/formatter_class.h"
 #include "formatter/formatter_attr.h"
@@ -986,6 +988,9 @@ PHP_MINIT_FUNCTION( intl )
 	/* Global error handling. */
 	intl_error_init( NULL TSRMLS_CC );
 
+	/* 'Converter' class for codepage conversions */
+	php_converter_minit(INIT_FUNC_ARGS_PASSTHRU);
+
 	return SUCCESS;
 }
 /* }}} */

ext/intl/tests/uconverter_enum.phpt

@@ -0,0 +1,21 @@
+--TEST--
+UConverter Enumerations
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--FILE--
+<?php
+$avail = UConverter::getAvailable();
+var_dump(count($avail) > 100);
+var_dump(in_array('UTF-7', $avail));
+var_dump(in_array('CESU-8', $avail));
+var_dump(in_array('ISO-8859-1', $avail));
+
+$latin1 = UConverter::getAliases('latin1');
+var_dump(in_array('ISO-8859-1', $latin1));
+
+--EXPECT--
+bool(true)
+bool(true)
+bool(true)
+bool(true)
+bool(true)

ext/intl/tests/uconverter_func_basic.phpt

@@ -0,0 +1,17 @@
+--TEST--
+Basic UConverter::transcode() usage
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--FILE--
+<?php
+var_dump(UConverter::transcode("This is an ascii string", 'utf-8', 'latin1'));
+// urlencode so that non-ascii shows up parsable in phpt file
+var_dump(urlencode(UConverter::transcode("Espa\xF1ol", 'utf-8', 'latin1')));
+var_dump(urlencode(UConverter::transcode("Stra\xDFa",  'utf-8', 'latin1')));
+
+var_dump(bin2hex(UConverter::transcode("\xE4", 'utf-8', 'koi8-r')));
+--EXPECT--
+string(23) "This is an ascii string"
+string(12) "Espa%C3%B1ol"
+string(11) "Stra%C3%9Fa"
+string(4) "d094"

ext/intl/tests/uconverter_func_subst.phpt

@@ -0,0 +1,31 @@
+--TEST--
+Basic UConverter::convert() w/ Subsitution
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--INI--
+intl.use_exceptions=false
+--FILE--
+<?php
+foreach(array('?','','??') as $subst) {
+  $opts = array('to_subst' => $subst);
+  $ret = UConverter::transcode("This is an ascii string", 'ascii', 'utf-8', $opts);
+  if ($ret === NULL) {
+    echo "Error: ", intl_get_error_message(), "\n";
+  } else {
+    var_dump($ret);
+  }
+  $ret = UConverter::transcode("Snowman: (\xE2\x98\x83)", 'ascii', 'utf-8', $opts);
+  if ($ret === NULL) {
+    echo "Error: ", intl_get_error_message(), "\n";
+  } else {
+    var_dump($ret);
+  }
+}
+
+--EXPECTF--
+string(23) "This is an ascii string"
+string(12) "Snowman: (?)"
+Error: transcode() returned error 1: U_ILLEGAL_ARGUMENT_ERROR: U_ILLEGAL_ARGUMENT_ERROR
+Error: transcode() returned error 1: U_ILLEGAL_ARGUMENT_ERROR: U_ILLEGAL_ARGUMENT_ERROR
+Error: transcode() returned error 1: U_ILLEGAL_ARGUMENT_ERROR: U_ILLEGAL_ARGUMENT_ERROR
+Error: transcode() returned error 1: U_ILLEGAL_ARGUMENT_ERROR: U_ILLEGAL_ARGUMENT_ERROR

ext/intl/tests/uconverter_oop_algo.phpt

@@ -0,0 +1,18 @@
+--TEST--
+UConverter Algorithmic converters
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--FILE--
+<?php
+$c = new UConverter('utf-8', 'latin1');
+var_dump(UConverter::LATIN_1 === $c->getSourceType());
+var_dump(UConverter::UTF8    === $c->getDestinationType());
+
+$c = new UConverter('koi8-r', 'utf-32be');
+var_dump(UConverter::UTF32_BigEndian === $c->getSourceType());
+var_dump(UConverter::SBCS            === $c->getDestinationType());
+--EXPECT--
+bool(true)
+bool(true)
+bool(true)
+bool(true)

ext/intl/tests/uconverter_oop_basic.phpt

@@ -0,0 +1,21 @@
+--TEST--
+Basic UConverter::convert() usage
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--FILE--
+<?php
+$c = new UConverter('utf-8', 'latin1');
+var_dump($c->convert("This is an ascii string"));
+// urlencode so that non-ascii shows up parsable in phpt file
+var_dump(urlencode($c->convert("Espa\xF1ol"))); // U+00F1 LATIN SMALL LETTER N WITH TILDE
+var_dump(urlencode($c->convert("Stra\xDFa")));  // U+00DF LATIN SMALL LETTER SHARP S
+var_dump(urlencode($c->convert("Stra\xC3\x9Fa", true))); // Reverse prior op
+
+$k = new UConverter('utf-8', 'koi8-r');
+var_dump(bin2hex($k->convert("\xE4"))); // U+0414 CYRILLIC CAPITAL LETTER DE
+--EXPECT--
+string(23) "This is an ascii string"
+string(12) "Espa%C3%B1ol"
+string(11) "Stra%C3%9Fa"
+string(8) "Stra%DFa"
+string(4) "d094"

ext/intl/tests/uconverter_oop_callback.phpt

@@ -0,0 +1,52 @@
+--TEST--
+UConverter::convert() w/ Callback Reasons
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--FILE--
+<?php
+class MyConverter extends UConverter {
+  /**
+   * Called during conversion from source encoding to internal UChar representation
+   */
+  public function toUCallback($reason, $source, $codeUnits, &$error) {
+    echo "toUCallback(", UConverter::reasonText($reason), ", ...)\n";
+    return parent::toUCallback($reason, $source, $codeUnits, $error);
+  }
+
+  /**
+   * Called during conversion from internal UChar to destination encoding
+   */
+  public function fromUCallback($reason, $source, $codePoint, &$error) {
+    echo "fromUCallback(", UConverter::reasonText($reason), ", ...)\n";
+    return parent::fromUCallback($reason, $source, $codePoint, $error);
+  }
+
+}
+
+$c = new MyConverter('ascii', 'utf-8');
+foreach(array("regular", "irregul\xC1\xA1r", "\xC2\xA1unsupported!") as $word) {
+  $c->convert($word);
+}
+--EXPECT--
+toUCallback(REASON_RESET, ...)
+toUCallback(REASON_RESET, ...)
+fromUCallback(REASON_RESET, ...)
+fromUCallback(REASON_RESET, ...)
+toUCallback(REASON_RESET, ...)
+toUCallback(REASON_ILLEGAL, ...)
+toUCallback(REASON_RESET, ...)
+toUCallback(REASON_ILLEGAL, ...)
+fromUCallback(REASON_RESET, ...)
+fromUCallback(REASON_UNASSIGNED, ...)
+fromUCallback(REASON_RESET, ...)
+fromUCallback(REASON_UNASSIGNED, ...)
+toUCallback(REASON_RESET, ...)
+toUCallback(REASON_RESET, ...)
+fromUCallback(REASON_RESET, ...)
+fromUCallback(REASON_UNASSIGNED, ...)
+fromUCallback(REASON_RESET, ...)
+fromUCallback(REASON_UNASSIGNED, ...)
+toUCallback(REASON_CLOSE, ...)
+fromUCallback(REASON_CLOSE, ...)
+toUCallback(REASON_CLOSE, ...)
+fromUCallback(REASON_CLOSE, ...)

ext/intl/tests/uconverter_oop_callback_return.phpt

@@ -0,0 +1,40 @@
+--TEST--
+UConverter::convert() w/ Callback Return Values
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--FILE--
+<?php
+class MyConverter extends UConverter {
+  public function toUCallback($reason, $source, $codeUnits, &$error) {
+    $error = U_ZERO_ERROR;
+    switch ($codeUnits) {
+      case "\x80": return NULL;
+      case "\x81": return 'a';
+      case "\x82": return ord('b');
+      case "\x83": return array('c');
+    }
+  }
+
+  /**
+   * Called during conversion from internal UChar to destination encoding
+   */
+  public function fromUCallback($reason, $source, $codePoint, &$error) {
+    $error = U_ZERO_ERROR;
+    switch ($codePoint) {
+      case 0x00F1: return "A";
+      case 0x00F2: return ord("B");
+      case 0x00F3: return array("C");
+      case 0x00F4: return NULL;
+    }
+  }
+
+}
+
+$c = new MyConverter('ascii', 'utf-8');
+// This line will trigger toUCallback
+var_dump($c->convert("\x80\x81\x82\x83"));
+// This line will trigger fromUCallback
+var_dump($c->convert("\xC3\xB1\xC3\xB2\xC3\xB3\xC3\xB4"));
+--EXPECT--
+string(3) "abc"
+string(3) "ABC"

ext/intl/tests/uconverter_oop_subst.phpt

@@ -0,0 +1,24 @@
+--TEST--
+Basic UConverter::convert() w/ Subsitution
+--SKIPIF--
+<?php if( !extension_loaded( 'intl' ) ) print 'skip'; ?>
+--INI--
+intl.use_exceptions=false
+--FILE--
+<?php
+$c = new UConverter('ascii', 'utf-8');
+
+foreach(array('?','','<unknown>') as $subst) {
+  if (!$c->setSubstChars($subst)) {
+    echo "**Disallowed\n";
+    continue;
+  }
+  var_dump($c->convert("This is an ascii string"));
+  var_dump($c->convert("Snowman: (\xE2\x98\x83)"));
+}
+
+--EXPECT--
+string(23) "This is an ascii string"
+string(12) "Snowman: (?)"
+**Disallowed
+**Disallowed