Project

General

Profile

wb-2_10_x / branches / main / framework / functions-utf8.php @ 10

1
<?php
2

    
3
// $Id: functions-utf8.php 2 2017-07-02 15:14:29Z Manuela $
4

    
5
/*
6

7
 Website Baker Project <http://www.websitebaker.org/>
8
 Copyright (C) 2004-2009, Ryan Djurovich
9

10
 Website Baker is free software; you can redistribute it and/or modify
11
 it under the terms of the GNU General Public License as published by
12
 the Free Software Foundation; either version 2 of the License, or
13
 (at your option) any later version.
14

15
 Website Baker is distributed in the hope that it will be useful,
16
 but WITHOUT ANY WARRANTY; without even the implied warranty of
17
 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
18
 GNU General Public License for more details.
19

20
 You should have received a copy of the GNU General Public License
21
 along with Website Baker; if not, write to the Free Software
22
 Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
23

24
*/
25

    
26
/*
27
 * A part of this file is based on 'utf8.php' from the DokuWiki-project.
28
 * (http://www.splitbrain.org/projects/dokuwiki):
29
 **
30
 * UTF8 helper functions
31
 * @license    LGPL (http://www.gnu.org/copyleft/lesser.html)
32
 * @author     Andreas Gohr <andi@splitbrain.org>
33
 **
34
 * modified for use with Website Baker
35
 * from thorn, Jan. 2008
36
 *
37
 * most of the original functions appeared to be to slow with large strings, so i replaced them with my own ones
38
 * thorn, Mar. 2008
39
 */
40

    
41
// Functions we use in Website Baker:
42
//   entities_to_7bit()
43
//   entities_to_umlauts2()
44
//   umlauts_to_entities2()
45
/* -------------------------------------------------------- */
46
// Must include code to stop this file being accessed directly
47
if(!defined('WB_PATH')) {
48
    require_once(dirname(__FILE__).'/globalExceptionHandler.php');
49
    throw new IllegalFileException();
50
}
51
/* -------------------------------------------------------- */
52
if (function_exists('functions-utf8')){return;}
53
/*
54
 * check for mb_string support
55
 */
56
//define('UTF8_NOMBSTRING',1); // uncomment this to forbid use of mb_string-functions
57
if(!defined('UTF8_MBSTRING')){
58
  if(function_exists('mb_substr') && !defined('UTF8_NOMBSTRING')){
59
    define('UTF8_MBSTRING',1);
60
  }else{
61
    define('UTF8_MBSTRING',0);
62
  }
63
}
64

    
65
if(UTF8_MBSTRING){ mb_internal_encoding('UTF-8'); }
66

    
67
require_once(WB_PATH.'/framework/charsets_table.php');
68

    
69
/*
70
 * Checks if a string contains 7bit ASCII only
71
 *
72
 * @author thorn
73
 */
74
function utf8_isASCII($str){
75
    if(preg_match('/[\x80-\xFF]/', $str))
76
        return false;
77
    else
78
        return true;
79
}
80

    
81
/*
82
 * Tries to detect if a string is in Unicode encoding
83
 *
84
 * @author <bmorel@ssi.fr>
85
 * @link   http://www.php.net/manual/en/function.utf8-encode.php
86
 */
87
function utf8_check($Str) {
88
 for ($i=0; $i<strlen($Str); $i++) {
89
  $b = ord($Str[$i]);
90
  if ($b < 0x80) continue; # 0bbbbbbb
91
  elseif (($b & 0xE0) == 0xC0) $n=1; # 110bbbbb
92
  elseif (($b & 0xF0) == 0xE0) $n=2; # 1110bbbb
93
  elseif (($b & 0xF8) == 0xF0) $n=3; # 11110bbb
94
  elseif (($b & 0xFC) == 0xF8) $n=4; # 111110bb
95
  elseif (($b & 0xFE) == 0xFC) $n=5; # 1111110b
96
  else return false; # Does not match any model
97
  for ($j=0; $j<$n; $j++) { # n bytes matching 10bbbbbb follow ?
98
   if ((++$i == strlen($Str)) || ((ord($Str[$i]) & 0xC0) != 0x80))
99
   return false;
100
  }
101
 }
102
 return true;
103
}
104

    
105
/*
106
 * Romanize a non-latin string
107
 *
108
 * @author Andreas Gohr <andi@splitbrain.org>
109
 */
110
function utf8_romanize($string){
111
  if(utf8_isASCII($string)) return $string; //nothing to do
112

    
113
  global $UTF8_ROMANIZATION;
114
  return strtr($string,$UTF8_ROMANIZATION);
115
}
116

    
117
/*
118
 * Removes special characters (nonalphanumeric) from a UTF-8 string
119
 *
120
 * This function adds the controlchars 0x00 to 0x19 to the array of
121
 * stripped chars (they are not included in $UTF8_SPECIAL_CHARS2)
122
 *
123
 * @author Andreas Gohr <andi@splitbrain.org>
124
 * @param  string $string     The UTF8 string to strip of special chars
125
 * @param  string $repl       Replace special with this string
126
 * @param  string $additional Additional chars to strip (used in regexp char class)
127
 */
128
function utf8_stripspecials($string,$repl='',$additional=''){
129
  global $UTF8_SPECIAL_CHARS2;
130

    
131
  static $specials = null;
132
  if(is_null($specials)){
133
    $specials = preg_quote($UTF8_SPECIAL_CHARS2, '/');
134
  }
135

    
136
  return preg_replace('/['.$additional.'\x00-\x19'.$specials.']/u',$repl,$string);
137
}
138

    
139
/*
140
 * added functions - thorn
141
 */
142

    
143
/*
144
 * faster replacement for utf8_entities_to_umlauts()
145
 * not all features of utf8_entities_to_umlauts() --> utf8_unhtml() are supported!
146
 * @author thorn
147
 */
148
function utf8_fast_entities_to_umlauts($str) {
149
    if(UTF8_MBSTRING) {
150
        // we need this for use with mb_convert_encoding
151
        $str = str_replace(array('&amp;','&gt;','&lt;','&quot;','&#039;','&nbsp;'), array('&amp;amp;','&amp;gt;','&amp;lt;','&amp;quot;','&amp;#39;','&amp;nbsp;'), $str);
152
        // we need two mb_convert_encoding()-calls - is this a bug?
153
        // mb_convert_encoding("ö&ouml;", 'UTF-8', 'HTML-ENTITIES'); // with string in utf-8-encoding doesn't work. Result: "öö"
154
        // Work-around: convert all umlauts to entities first ("ö&ouml;"->"&ouml;&ouml;"), then all entities to umlauts ("&ouml;&ouml;"->"öö")
155
        return(mb_convert_encoding(mb_convert_encoding($str, 'HTML-ENTITIES', 'UTF-8'),'UTF-8', 'HTML-ENTITIES'));
156
    } else {
157
        global $named_entities;global $numbered_entities;
158
        $str = str_replace($named_entities, $numbered_entities, $str);
159
        $str = preg_replace("/&#([0-9]+);/e", "code_to_utf8($1)", $str);
160
    }
161
    return($str);
162
}
163
// support-function for utf8_fast_entities_to_umlauts()
164
function code_to_utf8($num) {
165
    if ($num <= 0x7F) {
166
        return chr($num);
167
    } elseif ($num <= 0x7FF) {
168
        return chr(($num >> 6) + 192) . chr(($num & 63) + 128);
169
    } elseif ($num <= 0xFFFF) {
170
         return chr(($num >> 12) + 224) . chr((($num >> 6) & 63) + 128) . chr(($num & 63) + 128);
171
    } elseif ($num <= 0x1FFFFF) {
172
        return chr(($num >> 18) + 240) . chr((($num >> 12) & 63) + 128) . chr((($num >> 6) & 63) + 128) . chr(($num & 63) + 128);
173
    }
174
    return "?";
175
}
176

    
177
/*
178
 * faster replacement for utf8_umlauts_to_entities()
179
 * not all features of utf8_umlauts_to_entities() --> utf8_tohtml() are supported!
180
 * @author thorn
181
 */
182
function utf8_fast_umlauts_to_entities($string, $named_entities=true) {
183
    if(UTF8_MBSTRING)
184
        return(mb_convert_encoding($string, 'HTML-ENTITIES', 'UTF-8'));
185
    else {
186
        global $named_entities;global $numbered_entities;
187
        $new = "";
188
        $i=0;
189
        $len=strlen($string);
190
        if($len==0) return $string;
191
        do {
192
            if(ord($string{$i}) <= 127) $ud = $string{$i++};
193
            elseif(ord($string{$i}) <= 223) $ud = (ord($string{$i++})-192)*64 + (ord($string{$i++})-128);
194
            elseif(ord($string{$i}) <= 239) $ud = (ord($string{$i++})-224)*4096 + (ord($string{$i++})-128)*64 + (ord($string{$i++})-128);
195
            elseif(ord($string{$i}) <= 247) $ud = (ord($string{$i++})-240)*262144 + (ord($string{$i++})-128)*4096 + (ord($string{$i++})-128)*64 + (ord($string{$i++})-128);
196
            else $ud = ord($string{$i++}); // error!
197
            if($ud > 127) {
198
                $new .= "&#$ud;";
199
            } else {
200
                $new .= $ud;
201
            }
202
        } while($i < $len);
203
        $string = $new;
204
        if($named_entities)
205
            $string = str_replace($numbered_entities, $named_entities, $string);
206
    }
207
    return($string);
208
}
209

    
210
/*
211
 * Converts from various charsets to UTF-8
212
 *
213
 * Will convert a string from various charsets to UTF-8.
214
 * HTML-entities may be converted, too.
215
 * In case of error the returned string is unchanged, and a message is emitted.
216
 * Supported charsets are:
217
 * direct: iso_8859_1 iso_8859_2 iso_8859_3 iso_8859_4 iso_8859_5
218
 *         iso_8859_6 iso_8859_7 iso_8859_8 iso_8859_9 iso_8859_10 iso_8859_11
219
 * mb_convert_encoding: all wb charsets (except those from 'direct'); but not GB2312
220
 * iconv:  all wb charsets (except those from 'direct')
221
 *
222
 * @param  string  A string in supported encoding
223
 * @param  string  The charset to convert from, defaults to DEFAULT_CHARSET
224
 * @return string  A string in UTF-8-encoding, with all entities decoded, too.
225
 *                 String is unchanged in case of error.
226
 * @author thorn
227
 */
228
function charset_to_utf8($str, $charset_in=DEFAULT_CHARSET, $decode_entities=true) {
229
    global $iso_8859_2_to_utf8, $iso_8859_3_to_utf8, $iso_8859_4_to_utf8, $iso_8859_5_to_utf8, $iso_8859_6_to_utf8, $iso_8859_7_to_utf8, $iso_8859_8_to_utf8, $iso_8859_9_to_utf8, $iso_8859_10_to_utf8, $iso_8859_11_to_utf8;
230
    $charset_in = strtoupper($charset_in);
231
    if ($charset_in == "") { $charset_in = 'UTF-8'; }
232
    $wrong_ISO8859 = false;
233
    $converted = false;
234

    
235
    if((!function_exists('iconv') && !UTF8_MBSTRING && ($charset_in=='BIG5' || $charset_in=='ISO-2022-JP' || $charset_in=='ISO-2022-KR')) || (!function_exists('iconv') && $charset_in=='GB2312')) {
236
        // Nothing we can do here :-(
237
        // Charset is one of those obscure ISO-2022... or BIG5, GB2312 or something
238
        // and we can't use mb_convert_encoding() or iconv();
239
        // Emit an error-message.
240
        trigger_error("Can't convert from $charset_in without mb_convert_encoding() or iconv(). Use UTF-8 instead.", E_USER_WARNING);
241
        return($str);
242
    }
243

    
244
    // check if we have UTF-8 or a plain ASCII string
245
    if($charset_in == 'UTF-8' || utf8_isASCII($str)) {
246
        // we have utf-8. Just replace HTML-entities and return
247
        if($decode_entities && preg_match('/&[#0-9a-zA-Z]+;/',$str))
248
            return(utf8_fast_entities_to_umlauts($str));
249
        else // nothing to do
250
            return($str);
251
    }
252
    
253
    // Convert $str to utf8
254
    if(substr($charset_in,0,8) == 'ISO-8859') {
255
        switch($charset_in) {
256
            case 'ISO-8859-1': $str=utf8_encode($str); break;
257
            case 'ISO-8859-2': $str=strtr($str, $iso_8859_2_to_utf8); break;
258
            case 'ISO-8859-3': $str=strtr($str, $iso_8859_3_to_utf8); break;
259
            case 'ISO-8859-4': $str=strtr($str, $iso_8859_4_to_utf8); break;
260
            case 'ISO-8859-5': $str=strtr($str, $iso_8859_5_to_utf8); break;
261
            case 'ISO-8859-6': $str=strtr($str, $iso_8859_6_to_utf8); break;
262
            case 'ISO-8859-7': $str=strtr($str, $iso_8859_7_to_utf8); break;
263
            case 'ISO-8859-8': $str=strtr($str, $iso_8859_8_to_utf8); break;
264
            case 'ISO-8859-9': $str=strtr($str, $iso_8859_9_to_utf8); break;
265
            case 'ISO-8859-10': $str=strtr($str, $iso_8859_10_to_utf8); break;
266
            case 'ISO-8859-11': $str=strtr($str, $iso_8859_11_to_utf8); break;
267
            default: $wrong_ISO8859 = true;
268
        }
269
        if(!$wrong_ISO8859)
270
            $converted = true;
271
    }
272
    if(!$converted && UTF8_MBSTRING && $charset_in != 'GB2312') {
273
        // $charset is neither UTF-8 nor a known ISO-8859...
274
        // Try mb_convert_encoding() - but there's no GB2312 encoding in php's mb_* functions
275
        $str = mb_convert_encoding($str, 'UTF-8', $charset_in);
276
        $converted = true;
277
    } elseif(!$converted) { // Try iconv
278
        if(function_exists('iconv')) {
279
            $str = iconv($charset_in, 'UTF-8', $str);
280
            $converted = true;
281
        }
282
    }
283
    if($converted) {
284
        // we have utf-8, now replace HTML-entities and return
285
        if($decode_entities && preg_match('/&[#0-9a-zA-Z]+;/',$str))
286
            $str = utf8_fast_entities_to_umlauts($str);
287
        return($str);
288
    }
289
    
290
    // Nothing we can do here :-(
291
    // Charset is one of those obscure ISO-2022... or BIG5, GB2312 or something
292
    // and we can't use mb_convert_encoding() or iconv();
293
    // Emit an error-message.
294
    trigger_error("Can't convert from $charset_in without mb_convert_encoding() or iconv(). Use UTF-8 instead.", E_USER_WARNING);
295
    
296
    return $str;
297
}
298

    
299
/*
300
 * Converts from UTF-8 to various charsets
301
 *
302
 * Will convert a string from UTF-8 to various charsets.
303
 * HTML-entities will not! be converted.
304
 * In case of error the returned string is unchanged, and a message is emitted.
305
 * Supported charsets are:
306
 * direct: iso_8859_1 iso_8859_2 iso_8859_3 iso_8859_4 iso_8859_5
307
 *         iso_8859_6 iso_8859_7 iso_8859_8 iso_8859_9 iso_8859_10 iso_8859_11
308
 * mb_convert_encoding: all wb charsets (except those from 'direct'); but not GB2312
309
 * iconv:  all wb charsets (except those from 'direct')
310
 *
311
 * @param  string  An UTF-8 encoded string
312
 * @param  string  The charset to convert to, defaults to DEFAULT_CHARSET
313
 * @return string  A string in a supported encoding, with all entities decoded, too.
314
 *                 String is unchanged in case of error.
315
 * @author thorn
316
 */
317
function utf8_to_charset($str, $charset_out=DEFAULT_CHARSET) {
318
    global $utf8_to_iso_8859_2, $utf8_to_iso_8859_3, $utf8_to_iso_8859_4, $utf8_to_iso_8859_5, $utf8_to_iso_8859_6, $utf8_to_iso_8859_7, $utf8_to_iso_8859_8, $utf8_to_iso_8859_9, $utf8_to_iso_8859_10, $utf8_to_iso_8859_11;
319
    $charset_out = strtoupper($charset_out);
320
    $wrong_ISO8859 = false;
321
    $converted = false;
322

    
323
    if((!function_exists('iconv') && !UTF8_MBSTRING && ($charset_out=='BIG5' || $charset_out=='ISO-2022-JP' || $charset_out=='ISO-2022-KR')) || (!function_exists('iconv') && $charset_out=='GB2312')) {
324
        // Nothing we can do here :-(
325
        // Charset is one of those obscure ISO-2022... or BIG5, GB2312 or something
326
        // and we can't use mb_convert_encoding() or iconv();
327
        // Emit an error-message.
328
        trigger_error("Can't convert into $charset_out without mb_convert_encoding() or iconv(). Use UTF-8 instead.", E_USER_WARNING);
329
        return($str);
330
    }
331
    
332
    // the string comes from charset_to_utf8(), so we can skip this
333
    // replace HTML-entities first
334
    //if(preg_match('/&[#0-9a-zA-Z]+;/',$str))
335
    //    $str = utf8_entities_to_umlauts($str);
336
    
337
    // check if we need to convert
338
    if($charset_out == 'UTF-8' || utf8_isASCII($str)) {
339
        // Nothing to do. Just return
340
            return($str);
341
    }
342
    
343
    // Convert $str to $charset_out
344
    if(substr($charset_out,0,8) == 'ISO-8859') {
345
        switch($charset_out) {
346
            case 'ISO-8859-1': $str=utf8_decode($str); break;
347
            case 'ISO-8859-2': $str=strtr($str, $utf8_to_iso_8859_2); break;
348
            case 'ISO-8859-3': $str=strtr($str, $utf8_to_iso_8859_3); break;
349
            case 'ISO-8859-4': $str=strtr($str, $utf8_to_iso_8859_4); break;
350
            case 'ISO-8859-5': $str=strtr($str, $utf8_to_iso_8859_5); break;
351
            case 'ISO-8859-6': $str=strtr($str, $utf8_to_iso_8859_6); break;
352
            case 'ISO-8859-7': $str=strtr($str, $utf8_to_iso_8859_7); break;
353
            case 'ISO-8859-8': $str=strtr($str, $utf8_to_iso_8859_8); break;
354
            case 'ISO-8859-9': $str=strtr($str, $utf8_to_iso_8859_9); break;
355
            case 'ISO-8859-10': $str=strtr($str, $utf8_to_iso_8859_10); break;
356
            case 'ISO-8859-11': $str=strtr($str, $utf8_to_iso_8859_11); break;
357
            default: $wrong_ISO8859 = true;
358
        }
359
        if(!$wrong_ISO8859)
360
            $converted = true;
361
    }
362
    if(!$converted && UTF8_MBSTRING && $charset_out != 'GB2312') {
363
        // $charset is neither UTF-8 nor a known ISO-8859...
364
        // Try mb_convert_encoding() - but there's no GB2312 encoding in php's mb_* functions
365
        $str = mb_convert_encoding($str, $charset_out, 'UTF-8');
366
        $converted = true;
367
    } elseif(!$converted) { // Try iconv
368
        if(function_exists('iconv')) {
369
            $str = iconv('UTF-8', $charset_out, $str);
370
            $converted = true;
371
        }
372
    }
373
    if($converted) {
374
        return($str);
375
    }
376
    
377
    // Nothing we can do here :-(
378
    // Charset is one of those obscure ISO-2022... or BIG5, GB2312 or something
379
    // and we can't use mb_convert_encoding() or iconv();
380
    // Emit an error-message.
381
    trigger_error("Can't convert into $charset_out without mb_convert_encoding() or iconv(). Use UTF-8 instead.", E_USER_WARNING);
382
    
383
    return $str;
384
}
385

    
386
/*
387
 * convert Filenames to ASCII
388
 *
389
 * Convert all non-ASCII characters and all HTML-entities to their plain 7bit equivalents
390
 * Characters without an equivalent will be converted to hex-values.
391
 * The name entities_to_7bit() is somewhat misleading, but kept for compatibility-reasons.
392
 *
393
 * @param  string  Filename to convert (all encodings from charset_to_utf8() are allowed)
394
 * @return string  ASCII encoded string, to use as filename in wb's page_filename() and media_filename
395
 * @author thorn
396
 */
397
function entities_to_7bit($str) {
398
    // convert to UTF-8
399
    $str = charset_to_utf8($str);
400
    if(!utf8_check($str))
401
        return($str);
402
    // replace some specials
403
    $str = utf8_stripspecials($str, '_');
404
    // translate non-ASCII characters to ASCII
405
    $str = utf8_romanize($str);
406
    // missed some? - Many UTF-8-chars can't be romanized
407
    // convert to HTML-entities, and replace entites by hex-numbers
408
    $str = utf8_fast_umlauts_to_entities($str, false);
409
    $str = str_replace('&#039;', '&apos;', $str);
410
    $str = preg_replace_callback('/&#([0-9]+);/', function($matches) {return dechex($matches[1]);}, $str);
411
    // maybe there are some &gt; &lt; &apos; &quot; &amp; &nbsp; left, replace them too
412
    $str = str_replace(array('&gt;', '&lt;', '&apos;', '\'', '&quot;', '&amp;'), '', $str);
413
    return($str);
414
}
415

    
416
/*
417
 * Convert a string from mixed html-entities/umlauts to pure $charset_out-umlauts
418
 * 
419
 * Will replace all numeric and named entities except
420
 * &gt; &lt; &apos; &quot; &#039; &nbsp;
421
 * @author thorn
422
 */
423
function entities_to_umlauts2($string, $charset_out=DEFAULT_CHARSET) {
424
    $string = charset_to_utf8($string, DEFAULT_CHARSET, true);
425
    //if(utf8_check($string)) // this check is to much time-consuming (this may fail only if AddDefaultCharset is set)
426
        $string = utf8_to_charset($string, $charset_out);
427
    return ($string);
428
}
429

    
430
/*
431
 * Convert a string from mixed html-entities/umlauts to pure ASCII with HTML-entities
432
 * 
433
 * Will convert a string in $charset_in encoding to a pure ASCII string with HTML-entities.
434
 * @author thorn
435
 */
436
function umlauts_to_entities2($string, $charset_in=DEFAULT_CHARSET) {
437
    $string = charset_to_utf8($string, $charset_in, false);
438
    //if(utf8_check($string)) // this check is to much time-consuming (this may fail only if AddDefaultCharset is set)
439
        $string = utf8_fast_umlauts_to_entities($string, false);
440
    return($string);
441
}
442