php-src/ext/json/utf8_decode.c

180 lines
4.8 KiB
C
Raw Normal View History

/* utf8_decode.c */
/* 2005-12-25 */
/*
Copyright (c) 2005 JSON.org
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
The Software shall be used for Good, not Evil.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
*/
#include "utf8_decode.h"
/*
Very Strict UTF-8 Decoder
UTF-8 is a multibyte character encoding of Unicode. A character can be
represented by 1-4 bytes. The bit pattern of the first byte indicates the
number of continuation bytes.
Most UTF-8 decoders tend to be lenient, attempting to recover as much
information as possible, even from badly encoded input. This UTF-8
decoder is not lenient. It will reject input which does not include
proper continuation bytes. It will reject aliases (or suboptimal
codings). It will reject surrogates. (Surrogate encoding should only be
used with UTF-16.)
Code Contination Minimum Maximum
0xxxxxxx 0 0 127
10xxxxxx error
110xxxxx 1 128 2047
1110xxxx 2 2048 65535 excluding 55296 - 57343
11110xxx 3 65536 1114111
11111xxx error
*/
/*
Get the next byte. It returns UTF8_END if there are no more bytes.
*/
static int
get(json_utf8_decode *utf8)
{
int c;
if (utf8->the_index >= utf8->the_length) {
return UTF8_END;
}
c = utf8->the_input[utf8->the_index] & 0xFF;
utf8->the_index += 1;
return c;
}
/*
Get the 6-bit payload of the next continuation byte.
Return UTF8_ERROR if it is not a contination byte.
*/
static int
cont(json_utf8_decode *utf8)
{
int c = get(utf8);
return ((c & 0xC0) == 0x80) ? (c & 0x3F) : UTF8_ERROR;
}
/*
Initialize the UTF-8 decoder. The decoder is not reentrant,
*/
void
utf8_decode_init(json_utf8_decode *utf8, char p[], int length)
{
utf8->the_index = 0;
utf8->the_input = p;
utf8->the_length = length;
utf8->the_char = 0;
utf8->the_byte = 0;
}
/*
Get the current byte offset. This is generally used in error reporting.
*/
int
utf8_decode_at_byte(json_utf8_decode *utf8)
{
return utf8->the_byte;
}
/*
Get the current character offset. This is generally used in error reporting.
The character offset matches the byte offset if the text is strictly ASCII.
*/
int
utf8_decode_at_character(json_utf8_decode *utf8)
{
return utf8->the_char > 0 ? utf8->the_char - 1 : 0;
}
/*
Extract the next character.
Returns: the character (between 0 and 1114111)
or UTF8_END (the end)
or UTF8_ERROR (error)
*/
int
utf8_decode_next(json_utf8_decode *utf8)
{
int c; /* the first byte of the character */
int r; /* the result */
if (utf8->the_index >= utf8->the_length) {
return utf8->the_index == utf8->the_length ? UTF8_END : UTF8_ERROR;
}
utf8->the_byte = utf8->the_index;
utf8->the_char += 1;
c = get(utf8);
/*
Zero continuation (0 to 127)
*/
if ((c & 0x80) == 0) {
return c;
}
/*
One contination (128 to 2047)
*/
if ((c & 0xE0) == 0xC0) {
int c1 = cont(utf8);
if (c1 < 0) {
return UTF8_ERROR;
}
r = ((c & 0x1F) << 6) | c1;
return r >= 128 ? r : UTF8_ERROR;
}
/*
Two continuation (2048 to 55295 and 57344 to 65535)
*/
if ((c & 0xF0) == 0xE0) {
int c1 = cont(utf8);
int c2 = cont(utf8);
if (c1 < 0 || c2 < 0) {
return UTF8_ERROR;
}
r = ((c & 0x0F) << 12) | (c1 << 6) | c2;
return r >= 2048 && (r < 55296 || r > 57343) ? r : UTF8_ERROR;
}
/*
Three continuation (65536 to 1114111)
*/
if ((c & 0xF8) == 0xF0) {
int c1 = cont(utf8);
int c2 = cont(utf8);
int c3 = cont(utf8);
if (c1 < 0 || c2 < 0 || c3 < 0) {
return UTF8_ERROR;
}
r = ((c & 0x0F) << 18) | (c1 << 12) | (c2 << 6) | c3;
return r >= 65536 && r <= 1114111 ? r : UTF8_ERROR;
}
return UTF8_ERROR;
}