CVS update: php3/functions

Date: Wednesday October 28, 1998 @ 0:27 Author: ssb Update of /repository/php3/functions In directory asf:/u2/tmp/cvs-serv15755/functions Modified Files: xml.c php3_xml.h Log Message: added utf8_encode() and utf8_decode() Index: php3/functions/xml.c diff -c php3/functions/xml.c:1.13 php3/functions/xml.c:1.14 *** php3/functions/xml.c:1.13 Tue Oct 27 22:35:49 1998 --- php3/functions/xml.c Wed Oct 28 00:27:47 1998 *************** *** 27,33 **** +----------------------------------------------------------------------+ */ ! /* $Id: xml.c,v 1.13 1998/10/28 03:35:49 ssb Exp $ */ #define IS_EXT_MODULE #if COMPILE_DL # if PHP_31 --- 27,33 ---- +----------------------------------------------------------------------+ */ ! /* $Id: xml.c,v 1.14 1998/10/28 05:27:47 ssb Exp $ */ #define IS_EXT_MODULE #if COMPILE_DL # if PHP_31 *************** *** 60,66 **** */ /* Long-term TODO list: - * - Support other character sets than ISO-8859-1 * - Fix the expat library so you can install your own memory manager * functions */ --- 60,65 ---- *************** *** 111,119 **** static char *php3i_pvalstrcpy(pval *); static void xml_destroy_parser(xml_parser *); static void xml_set_handler(char **, pval *); inline static char xml_decode_iso_8859_1(unsigned short); inline static char xml_decode_us_ascii(unsigned short); ! static char *xml_treat_utf8(const XML_Char *, int, int *, const XML_Char *); static pval *xml_call_handler(xml_parser *, char *, int, pval **); static pval *php3i_longpval(long); static pval *php3i_stringpval(const char *); --- 110,121 ---- static char *php3i_pvalstrcpy(pval *); static void xml_destroy_parser(xml_parser *); static void xml_set_handler(char **, pval *); + inline static unsigned short xml_encode_iso_8859_1(unsigned char); inline static char xml_decode_iso_8859_1(unsigned short); + inline static unsigned short xml_encode_us_ascii(char); inline static char xml_decode_us_ascii(unsigned short); ! static XML_Char *xml_utf8_encode(const char *, int, int *, const XML_Char *); ! static char *xml_utf8_decode(const XML_Char *, int, int *, const XML_Char *); static pval *xml_call_handler(xml_parser *, char *, int, pval **); static pval *php3i_longpval(long); static pval *php3i_stringpval(const char *); *************** *** 150,155 **** --- 152,159 ---- PHP_FE(xml_parser_free, NULL) PHP_FE(xml_parser_set_option, NULL) PHP_FE(xml_parser_get_option, NULL) + PHP_FE(utf8_encode, NULL) + PHP_FE(utf8_decode, NULL) {NULL, NULL, NULL} }; *************** *** 168,177 **** * the encoding is currently done internally by expat/xmltok. */ xml_encoding xml_encodings[] = { ! { "ISO-8859-1", xml_decode_iso_8859_1, NULL }, ! { "US-ASCII", xml_decode_us_ascii, NULL }, ! { "UTF-8", NULL, NULL }, ! { NULL, NULL, NULL } }; /* }}} */ --- 172,181 ---- * the encoding is currently done internally by expat/xmltok. */ xml_encoding xml_encodings[] = { ! { "ISO-8859-1", xml_decode_iso_8859_1, xml_encode_iso_8859_1 }, ! { "US-ASCII", xml_decode_us_ascii, xml_encode_us_ascii }, ! { "UTF-8", NULL, NULL }, ! { NULL, NULL, NULL } }; /* }}} */ *************** *** 395,400 **** --- 399,414 ---- } /* }}} */ + /* {{{ xml_encode_iso_8859_1() */ + + inline static unsigned short + xml_encode_iso_8859_1(unsigned char c) + { + php3_printf("c=%d ", c); + return (unsigned short)c; + } + + /* }}} */ /* {{{ xml_decode_iso_8859_1() */ inline static char *************** *** 404,409 **** --- 418,432 ---- } /* }}} */ + /* {{{ xml_encode_us_ascii() */ + + inline static unsigned short + xml_encode_us_ascii(char c) + { + return (unsigned short)c; + } + + /* }}} */ /* {{{ xml_decode_us_ascii() */ inline static char *************** *** 413,418 **** --- 436,443 ---- } /* }}} */ + /* {{{ xml_get_encoding() */ + static xml_encoding * xml_get_encoding(const XML_Char *name) { *************** *** 425,434 **** enc++; } return NULL; } ! /* {{{ xml_treat_utf8 */ static char * ! xml_treat_utf8(const XML_Char *s, int len, int *newlen, const XML_Char *encoding) { int pos = len; char *newbuf = emalloc(len); --- 450,515 ---- enc++; } return NULL; + } + + /* }}} */ + /* {{{ xml_utf8_encode */ + static XML_Char * + xml_utf8_encode(const char *s, int len, int *newlen, const XML_Char *encoding) + { + int pos = len; + char *newbuf; + unsigned short c; + unsigned short (*encoder)(char) = NULL; + xml_encoding *enc = xml_get_encoding(encoding); + + *newlen = 0; + if (enc) { + encoder = enc->encoding_function; + } else { + /* If the target encoding was unknown, fail */ + return NULL; + } + if (encoder == NULL) { + /* If no encoder function was specified, return the data as-is. + */ + newbuf = emalloc(len); + memcpy(newbuf, s, len); + *newlen = len; + return newbuf; + } + /* This is the theoretical max (will never get beyond len * 2 as long + * as we are converting from single-byte characters, though) */ + newbuf = emalloc(len * 4); + while (pos > 0) { + c = encoder ? encoder((unsigned char)(*s)) : (unsigned short)(*s); + if (c < 0x80) { + newbuf[(*newlen)++] = c; + } else if (c < 0x800) { + newbuf[(*newlen)++] = (0xc0 | (c >> 6)); + newbuf[(*newlen)++] = (0x80 | (c & 0x3f)); + } else if (c < 0x10000) { + newbuf[(*newlen)++] = (0xe0 | (c >> 12)); + newbuf[(*newlen)++] = (0xc0 | ((c >> 6) & 0x3f)); + newbuf[(*newlen)++] = (0x80 | (c & 0x3f)); + } else if (c < 0x200000) { + newbuf[(*newlen)++] = (0xf0 | (c >> 18)); + newbuf[(*newlen)++] = (0xe0 | ((c >> 12) & 0x3f)); + newbuf[(*newlen)++] = (0xc0 | ((c >> 6) & 0x3f)); + newbuf[(*newlen)++] = (0x80 | (c & 0x3f)); + } + pos--; + s++; + } + if (*newlen < len * 4) { + newbuf = erealloc(newbuf, *newlen); + } + return newbuf; } ! /* }}} */ ! /* {{{ xml_utf8_decode */ static char * ! xml_utf8_decode(const XML_Char *s, int len, int *newlen, const XML_Char *encoding) { int pos = len; char *newbuf = emalloc(len); *************** *** 449,459 **** return newbuf; } while (pos > 0) { - #ifdef XML_UNICODE /* defined if we are using 16-bit Unicode characters */ - c = (*s); - len--; - s++; - #else c = (unsigned char)(*s); if (c >= 0xf0) { /* four bytes encoded, 21 bits */ c = ((s[0]&7)<<18) | ((s[1]&63)<<12) | ((s[2]&63)<<6) | (s[3]&63); --- 530,535 ---- *************** *** 471,477 **** s++; pos--; } - #endif newbuf[*newlen] = decoder ? decoder(c) : c; ++*newlen; } --- 547,552 ---- *************** *** 521,527 **** len = php3i_xmlcharlen(s); } ret->type = IS_STRING; ! ret->value.str.val = xml_treat_utf8(s, len, &ret->value.str.len, encoding); return ret; } --- 596,602 ---- len = php3i_xmlcharlen(s); } ret->type = IS_STRING; ! ret->value.str.val = xml_utf8_decode(s, len, &ret->value.str.len, encoding); return ret; } *************** *** 1216,1221 **** --- 1291,1338 ---- break; } RETVAL_FALSE; + } + + /* }}} */ + /* {{{ string utf8_encode(string data) */ + + PHP_FUNCTION(utf8_encode) + { + pval *arg; + XML_Char *encoded; + int len; + XML_TLS_VARS; + + if (ARG_COUNT(ht) != 1 || getParameters(ht, 1, &arg) == FAILURE) { + WRONG_PARAM_COUNT; + } + convert_to_string(arg); + encoded = xml_utf8_encode(arg->value.str.val, arg->value.str.len, &len, "ISO-8859-1"); + if (encoded == NULL) { + RETURN_FALSE; + } + RETVAL_STRINGL(encoded, len, 0); + } + + /* }}} */ + /* {{{ string utf8_decode(string data) */ + + PHP_FUNCTION(utf8_decode) + { + pval *arg; + XML_Char *decoded; + int len; + XML_TLS_VARS; + + if (ARG_COUNT(ht) != 1 || getParameters(ht, 1, &arg) == FAILURE) { + WRONG_PARAM_COUNT; + } + convert_to_string(arg); + decoded = xml_utf8_decode(arg->value.str.val, arg->value.str.len, &len, "ISO-8859-1"); + if (decoded == NULL) { + RETURN_FALSE; + } + RETVAL_STRINGL(decoded, len, 0); } /* }}} */ Index: php3/functions/php3_xml.h diff -c php3/functions/php3_xml.h:1.6 php3/functions/php3_xml.h:1.7 *** php3/functions/php3_xml.h:1.6 Tue Oct 27 22:35:50 1998 --- php3/functions/php3_xml.h Wed Oct 28 00:27:47 1998 *************** *** 37,42 **** --- 37,46 ---- #include <xml/xmltok.h> #include <xml/xmlparse.h> + #ifdef XML_UNICODE + # error "UTF-16 Unicode support not implemented!" + #endif + typedef struct { int le_xml_parser; XML_Char *default_encoding; *************** *** 96,101 **** --- 100,107 ---- PHP_FUNCTION(xml_parser_free); PHP_FUNCTION(xml_parser_set_option); PHP_FUNCTION(xml_parser_get_option); + PHP_FUNCTION(utf8_encode); + PHP_FUNCTION(utf8_decode); #else /* !HAVE_LIBEXPAT */ -- PHP Development Mailing List http://www.php.net/ To unsubscribe send an empty message to php-dev-unsubscribe@lists.php.net For help: php-dev-help@lists.php.net