Move prefs init above plugin stuff, make life easier for plugins hooked on config_ove...
[squirrelmail.git] / include / languages.php
1 <?php
2
3 /**
4 * SquirrelMail internationalization functions
5 *
6 * This file contains variuos functions that are needed to do
7 * internationalization of SquirrelMail.
8 *
9 * Internally the output character set is used. Other characters are
10 * encoded using Unicode entities according to HTML 4.0.
11 *
12 * Before 1.5.2 functions were stored in functions/i18n.php. Script is moved
13 * because it executes some code in order to detect functions supported by
14 * existing PHP installation and implements fallback functions when required
15 * functions are not available. Scripts in functions/ directory should not
16 * setup anything when they are loaded.
17 * @copyright &copy; 1999-2007 The SquirrelMail Project Team
18 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
19 * @version $Id$
20 * @package squirrelmail
21 * @subpackage i18n
22 */
23
24
25 /**
26 * Wrapper for textdomain(), bindtextdomain() and
27 * bind_textdomain_codeset() primarily intended for
28 * plugins when changing into their own text domain
29 * and back again.
30 *
31 * Note that if plugins using this function have
32 * their translation files located in the SquirrelMail
33 * locale directory, the second argument is optional.
34 *
35 * @param string $domain_name The name of the text domain
36 * (usually the plugin name, or
37 * "squirrelmail") being switched to.
38 * @param string $directory The directory that contains
39 * all translations for the domain
40 * (OPTIONAL; default is SquirrelMail
41 * locale directory).
42 *
43 * @return string The name of the text domain that was set
44 * *BEFORE* it is changed herein - NOTE that
45 * this differs from PHP's textdomain()
46 *
47 * @since 1.5.2 and 1.4.10
48 */
49 function sq_change_text_domain($domain_name, $directory='') {
50
51 static $domains_already_seen = array();
52 global $gettext_domain;
53 $return_value = $gettext_domain;
54
55 // only need to call bindtextdomain() once
56 //
57 if (in_array($domain_name, $domains_already_seen)) {
58 sq_textdomain($domain_name);
59 return $return_value;
60 }
61
62 $domains_already_seen[] = $domain_name;
63
64 if (empty($directory)) $directory = SM_PATH . 'locale/';
65
66 sq_bindtextdomain($domain_name, $directory);
67 sq_textdomain($domain_name);
68
69 return $return_value;
70 }
71
72 /**
73 * Gettext bindtextdomain wrapper.
74 *
75 * Wrapper solves differences between php versions in order to provide
76 * ngettext support. Should be used if translation uses ngettext
77 * functions.
78 *
79 * This also provides a bind_textdomain_codeset call to make sure the
80 * domain's encoding will not be overridden.
81 *
82 * @since 1.4.10 and 1.5.1
83 * @param string $domain gettext domain name
84 * @param string $dir directory that contains all translations (OPTIONAL;
85 * if not specified, defaults to SquirrelMail locale
86 * directory)
87 * @return string path to translation directory
88 */
89 function sq_bindtextdomain($domain,$dir='') {
90 global $l10n, $gettext_flags, $sm_notAlias;
91
92 if (empty($dir)) $dir = SM_PATH . 'locale/';
93
94 if ($gettext_flags==7) {
95 // gettext extension without ngettext
96 if (substr($dir, -1) != '/') $dir .= '/';
97 $mofile=$dir . $sm_notAlias . '/LC_MESSAGES/' . $domain . '.mo';
98 $input = new FileReader($mofile);
99 $l10n[$domain] = new gettext_reader($input);
100 }
101
102 $dir=bindtextdomain($domain,$dir);
103
104 // set codeset in order to avoid gettext charset conversions
105 if (function_exists('bind_textdomain_codeset')
106 && isset($languages[$sm_notAlias]['CHARSET'])) {
107
108 // Japanese translation uses different internal charset
109 if ($sm_notAlias == 'ja_JP') {
110 bind_textdomain_codeset ($domain_name, 'EUC-JP');
111 } else {
112 bind_textdomain_codeset ($domain_name, $languages[$sm_notAlias]['CHARSET']);
113 }
114
115 }
116
117 return $dir;
118 }
119
120 /**
121 * Gettext textdomain wrapper.
122 * Makes sure that gettext_domain global is modified.
123 * @since 1.5.1
124 * @param string $name gettext domain name
125 * @return string gettext domain name
126 */
127 function sq_textdomain($domain) {
128 global $gettext_domain;
129 $gettext_domain=textdomain($domain);
130 return $gettext_domain;
131 }
132
133 /**
134 * php setlocale function wrapper
135 *
136 * From php 4.3.0 it is possible to use arrays in order to set locale.
137 * php gettext extension works only when locale is set. This wrapper
138 * function allows to use more than one locale name.
139 *
140 * @param int $category locale category name. Use php named constants
141 * (LC_ALL, LC_COLLATE, LC_CTYPE, LC_MONETARY, LC_NUMERIC, LC_TIME)
142 * @param mixed $locale option contains array with possible locales or string with one locale
143 * @return string name of set locale or false, if all locales fail.
144 * @since 1.5.1 and 1.4.5
145 * @see http://www.php.net/setlocale
146 */
147 function sq_setlocale($category,$locale) {
148 if (is_string($locale)) {
149 // string with only one locale
150 $ret = setlocale($category,$locale);
151 } elseif (! check_php_version(4,3)) {
152 // older php version (second setlocale argument must be string)
153 $ret=false;
154 $index=0;
155 while ( ! $ret && $index<count($locale)) {
156 $ret=setlocale($category,$locale[$index]);
157 $index++;
158 }
159 } else {
160 // php 4.3.0 or better, use entire array
161 $ret=setlocale($category,$locale);
162 }
163
164 /* safety checks */
165 if (preg_match("/^.*\/.*\/.*\/.*\/.*\/.*$/",$ret)) {
166 /**
167 * Welcome to We-Don't-Follow-Own-Fine-Manual department
168 * OpenBSD 3.8, 3.9-current and maybe later versions
169 * return invalid response to setlocale command.
170 * SM bug report #1427512.
171 */
172 $ret = false;
173 }
174 return $ret;
175 }
176
177 /**
178 * Converts string from given charset to charset, that can be displayed by user translation.
179 *
180 * Function by default returns html encoded strings, if translation uses different encoding.
181 * If Japanese translation is used - function returns string converted to euc-jp
182 * If iconv or recode functions are enabled and translation uses utf-8 - function returns utf-8 encoded string.
183 * If $charset is not supported - function returns unconverted string.
184 *
185 * sanitizing of html tags is also done by this function.
186 *
187 * @param string $charset
188 * @param string $string Text to be decoded
189 * @param boolean $force_decode converts string to html without $charset!=$default_charset check.
190 * Argument is available since 1.5.1 and 1.4.5.
191 * @param boolean $save_html disables htmlspecialchars() in order to preserve
192 * html formating. Use with care. Available since 1.5.1
193 * @return string decoded string
194 */
195 function charset_decode ($charset, $string, $force_decode=false, $save_html=false) {
196 global $languages, $squirrelmail_language, $default_charset;
197 global $use_php_recode, $use_php_iconv, $aggressive_decoding;
198
199 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
200 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
201 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode', $string);
202 }
203
204 $charset = strtolower($charset);
205
206 set_my_charset();
207
208 // Variables that allow to use functions without function_exist() calls
209 if (! isset($use_php_recode) || $use_php_recode=="" ) {
210 $use_php_recode=false; }
211 if (! isset($use_php_iconv) || $use_php_iconv=="" ) {
212 $use_php_iconv=false; }
213
214 // Don't do conversion if charset is the same.
215 if ( ! $force_decode && $charset == strtolower($default_charset) )
216 return ($save_html ? $string : htmlspecialchars($string));
217
218 // catch iso-8859-8-i thing
219 if ( $charset == "iso-8859-8-i" )
220 $charset = "iso-8859-8";
221
222 /*
223 * Recode converts html special characters automatically if you use
224 * 'charset..html' decoding. There is no documented way to put -d option
225 * into php recode function call.
226 */
227 if ( $use_php_recode ) {
228 if ( $default_charset == "utf-8" ) {
229 // other charsets can be converted to utf-8 without loss.
230 // and output string is smaller
231 $string = recode_string($charset . "..utf-8",$string);
232 return ($save_html ? $string : htmlspecialchars($string));
233 } else {
234 $string = recode_string($charset . "..html",$string);
235 // recode does not convert single quote, htmlspecialchars does.
236 $string = str_replace("'", '&#039;', $string);
237 // undo html specialchars
238 if ($save_html)
239 $string=str_replace(array('&amp;','&quot;','&lt;','&gt;'),
240 array('&','"','<','>'),$string);
241 return $string;
242 }
243 }
244
245 // iconv functions does not have html target and can be used only with utf-8
246 if ( $use_php_iconv && $default_charset=='utf-8') {
247 $string = iconv($charset,$default_charset,$string);
248 return ($save_html ? $string : htmlspecialchars($string));
249 }
250
251 // If we don't use recode and iconv, we'll do it old way.
252
253 /* All HTML special characters are 7 bit and can be replaced first */
254 if (! $save_html) $string = htmlspecialchars ($string);
255
256 /* controls cpu and memory intensive decoding cycles */
257 if (! isset($aggressive_decoding) || $aggressive_decoding=="" ) {
258 $aggressive_decoding=false; }
259
260 $decode=fixcharset($charset);
261 $decodefile=SM_PATH . 'functions/decode/' . $decode . '.php';
262 if ($decode != 'index' && file_exists($decodefile)) {
263 include_once($decodefile);
264 // send $save_html argument to decoding function. needed for iso-2022-xx decoding.
265 $ret = call_user_func('charset_decode_'.$decode, $string, $save_html);
266 } else {
267 $ret = $string;
268 }
269 return( $ret );
270 }
271
272 /**
273 * Converts html string to given charset
274 * @since 1.5.1 and 1.4.4
275 * @param string $string
276 * @param string $charset
277 * @param boolean $htmlencode keep htmlspecialchars encoding
278 * @return string
279 */
280 function charset_encode($string,$charset,$htmlencode=true) {
281 global $default_charset;
282
283 $encode=fixcharset($charset);
284 $encodefile=SM_PATH . 'functions/encode/' . $encode . '.php';
285 if ($encode != 'index' && file_exists($encodefile)) {
286 include_once($encodefile);
287 $ret = call_user_func('charset_encode_'.$encode, $string);
288 } elseif(file_exists(SM_PATH . 'functions/encode/us_ascii.php')) {
289 // function replaces all 8bit html entities with question marks.
290 // it is used when other encoding functions are unavailable
291 include_once(SM_PATH . 'functions/encode/us_ascii.php');
292 $ret = charset_encode_us_ascii($string);
293 } else {
294 /**
295 * fix for yahoo users that remove all us-ascii related things
296 */
297 $ret = $string;
298 }
299
300 /**
301 * Undo html special chars, some places (like compose form) have
302 * own sanitizing functions and don't need html symbols.
303 * Undo chars only after encoding in order to prevent conversion of
304 * html entities in plain text emails.
305 */
306 if (! $htmlencode ) {
307 $ret = str_replace(array('&amp;','&gt;','&lt;','&quot;'),array('&','>','<','"'),$ret);
308 }
309 return( $ret );
310 }
311
312 /**
313 * Combined decoding and encoding functions
314 *
315 * If conversion is done to charset different that utf-8, unsupported symbols
316 * will be replaced with question marks.
317 * @since 1.5.1 and 1.4.4
318 * @param string $in_charset initial charset
319 * @param string $string string that has to be converted
320 * @param string $out_charset final charset
321 * @param boolean $htmlencode keep htmlspecialchars encoding
322 * @return string converted string
323 */
324 function charset_convert($in_charset,$string,$out_charset,$htmlencode=true) {
325 $string=charset_decode($in_charset,$string,true);
326 $string=sqi18n_convert_entities($string);
327 $string=charset_encode($string,$out_charset,$htmlencode);
328 return $string;
329 }
330
331 /**
332 * Makes charset name suitable for decoding cycles
333 *
334 * @since 1.5.0 and 1.4.4
335 * @param string $charset Name of charset
336 * @return string $charset Adjusted name of charset
337 */
338 function fixcharset($charset) {
339 /* remove minus and characters that might be used in paths from charset
340 * name in order to be able to use it in function names and include calls.
341 */
342 $charset=preg_replace("/[-:.\/\\\]/",'_',$charset);
343
344 // OE ks_c_5601_1987 > cp949
345 $charset=str_replace('ks_c_5601_1987','cp949',$charset);
346 // Moz x-euc-tw > euc-tw
347 $charset=str_replace('x_euc','euc',$charset);
348 // Moz x-windows-949 > cp949
349 $charset=str_replace('x_windows_','cp',$charset);
350
351 // windows-125x and cp125x charsets
352 $charset=str_replace('windows_','cp',$charset);
353
354 // ibm > cp
355 $charset=str_replace('ibm','cp',$charset);
356
357 // iso-8859-8-i -> iso-8859-8
358 // use same cycle until I'll find differences
359 $charset=str_replace('iso_8859_8_i','iso_8859_8',$charset);
360
361 return $charset;
362 }
363
364 /**
365 * Set up the language to be output
366 * if $do_search is true, then scan the browser information
367 * for a possible language that we know
368 *
369 * Function sets system locale environment (LC_ALL, LANG, LANGUAGE),
370 * gettext translation bindings and html header information.
371 *
372 * Function returns error codes, if there is some fatal error.
373 * 0 = no error,
374 * 1 = mbstring support is not present,
375 * 2 = mbstring support is not present, user's translation reverted to en_US.
376 *
377 * @param string $sm_language translation used by user's interface
378 * @param bool $do_search use browser's preferred language detection functions. Defaults to false.
379 * @param bool $default set $sm_language to $squirrelmail_default_language if language detection fails or language is not set. Defaults to false.
380 * @return int function execution error codes.
381 */
382 function set_up_language($sm_language, $do_search = false, $default = false) {
383
384 static $SetupAlready = 0;
385 global $use_gettext, $languages,
386 $squirrelmail_language, $squirrelmail_default_language, $default_charset,
387 $sm_notAlias, $username, $data_dir;
388
389 if ($SetupAlready) {
390 return;
391 }
392
393 $SetupAlready = TRUE;
394 sqgetGlobalVar('HTTP_ACCEPT_LANGUAGE', $accept_lang, SQ_SERVER);
395
396 /**
397 * If function is asked to detect preferred language
398 * OR squirrelmail default language is set to empty string
399 * AND
400 * squirrelmail language ($sm_language) is empty string
401 * (not set in user's prefs and no cookie with language info)
402 * AND
403 * browser provides list of preferred languages
404 * THEN
405 * get preferred language from HTTP_ACCEPT_LANGUAGE header
406 */
407 if (($do_search || empty($squirrelmail_default_language)) &&
408 ! $sm_language &&
409 isset($accept_lang)) {
410 // TODO: use more than one language, if first language is not available
411 // FIXME: function assumes that string contains two or more characters.
412 // FIXME: some languages use 5 chars
413 $sm_language = substr($accept_lang, 0, 2);
414 }
415
416 /**
417 * If language preference is not set OR script asks to use default language
418 * AND
419 * default squirrelmail language is not set to empty string
420 * THEN
421 * use default squirrelmail language value from configuration.
422 */
423 if ((!$sm_language||$default) &&
424 ! empty($squirrelmail_default_language)) {
425 $squirrelmail_language = $squirrelmail_default_language;
426 $sm_language = $squirrelmail_default_language;
427 }
428
429 /** provide failsafe language when detection fails */
430 if (! $sm_language) $sm_language='en_US';
431
432 $sm_notAlias = $sm_language;
433
434 // Catching removed translation
435 // System reverts to English translation if user prefs contain translation
436 // that is not available in $languages array
437 if (!isset($languages[$sm_notAlias])) {
438 $sm_notAlias="en_US";
439 }
440
441 while (isset($languages[$sm_notAlias]['ALIAS'])) {
442 $sm_notAlias = $languages[$sm_notAlias]['ALIAS'];
443 }
444
445 if ( isset($sm_language) &&
446 $use_gettext &&
447 $sm_language != '' &&
448 isset($languages[$sm_notAlias]['CHARSET']) ) {
449 sq_bindtextdomain( 'squirrelmail', SM_PATH . 'locale/' );
450 sq_textdomain( 'squirrelmail' );
451
452 // Use LOCALE key, if it is set.
453 if (isset($languages[$sm_notAlias]['LOCALE'])){
454 $longlocale=$languages[$sm_notAlias]['LOCALE'];
455 } else {
456 $longlocale=$sm_notAlias;
457 }
458
459 // try setting locale
460 $retlocale=sq_setlocale(LC_ALL, $longlocale);
461
462 // check if locale is set and assign that locale to $longlocale
463 // in order to use it in putenv calls.
464 if (! is_bool($retlocale)) {
465 $longlocale=$retlocale;
466 } elseif (is_array($longlocale)) {
467 // setting of all locales failed.
468 // we need string instead of array used in LOCALE key.
469 $longlocale=$sm_notAlias;
470 }
471
472 if ( !((bool)ini_get('safe_mode')) &&
473 getenv( 'LC_ALL' ) != $longlocale ) {
474 putenv( "LC_ALL=$longlocale" );
475 putenv( "LANG=$longlocale" );
476 putenv( "LANGUAGE=$longlocale" );
477 putenv( "LC_NUMERIC=C" );
478 if ($sm_notAlias=='tr_TR') putenv( "LC_CTYPE=C" );
479 }
480 // Workaround for plugins that use numbers with floating point
481 // It might be removed if plugins use correct decimal delimiters
482 // according to locale settings.
483 setlocale(LC_NUMERIC, 'C');
484 // Workaround for specific Turkish strtolower/strtoupper rules.
485 // Many functions expect English conversion rules.
486 if ($sm_notAlias=='tr_TR') setlocale(LC_CTYPE,'C');
487
488 /**
489 * Set text direction/alignment variables
490 * When language environment is setup, scripts can use these globals
491 * without accessing $languages directly and making checks for optional
492 * array key.
493 */
494 global $text_direction, $left_align, $right_align;
495 if (isset($languages[$sm_notAlias]['DIR']) &&
496 $languages[$sm_notAlias]['DIR'] == 'rtl') {
497 /**
498 * Text direction
499 * @global string $text_direction
500 */
501 $text_direction='rtl';
502 /**
503 * Left alignment
504 * @global string $left_align
505 */
506 $left_align='right';
507 /**
508 * Right alignment
509 * @global string $right_align
510 */
511 $right_align='left';
512 } else {
513 $text_direction='ltr';
514 $left_align='left';
515 $right_align='right';
516 }
517
518 $squirrelmail_language = $sm_notAlias;
519 if ($squirrelmail_language == 'ja_JP') {
520 header ('Content-Type: text/html; charset=EUC-JP');
521 if (!function_exists('mb_internal_encoding')) {
522 // Error messages can't be displayed here
523 $error = 1;
524 // Revert to English if possible.
525 if (function_exists('setPref') && $username!='' && $data_dir!="") {
526 setPref($data_dir, $username, 'language', "en_US");
527 $error = 2;
528 }
529 // stop further execution in order not to get php errors on mb_internal_encoding().
530 return $error;
531 }
532 if (function_exists('mb_language')) {
533 mb_language('Japanese');
534 }
535 mb_internal_encoding('EUC-JP');
536 mb_http_output('pass');
537 } elseif ($squirrelmail_language == 'en_US') {
538 header( 'Content-Type: text/html; charset=' . $default_charset );
539 } else {
540 header( 'Content-Type: text/html; charset=' . $languages[$sm_notAlias]['CHARSET'] );
541 }
542 /**
543 * mbstring.func_overload fix (#929644).
544 *
545 * php mbstring extension can replace standard string functions with their multibyte
546 * equivalents. See http://www.php.net/ref.mbstring#mbstring.overload. This feature
547 * was added in php v.4.2.0
548 *
549 * Some SquirrelMail functions work with 8bit strings in bytes. If interface is forced
550 * to use mbstring functions and mbstring internal encoding is set to multibyte charset,
551 * interface can't trust regular string functions. Due to mbstring overloading design
552 * limits php scripts can't control this setting.
553 *
554 * This hack should fix some issues related to 8bit strings in passwords. Correct fix is
555 * to disable mbstring overloading. Japanese translation uses different internal encoding.
556 */
557 if ($squirrelmail_language != 'ja_JP' &&
558 function_exists('mb_internal_encoding') &&
559 check_php_version(4,2,0) &&
560 (int)ini_get('mbstring.func_overload')!=0) {
561 mb_internal_encoding('pass');
562 }
563 }
564 return 0;
565 }
566
567 /**
568 * Sets default_charset variable according to the one that is used by user's translations.
569 *
570 * Function changes global $default_charset variable in order to be sure, that it
571 * contains charset used by user's translation. Sanity of $squirrelmail_language
572 * and $default_charset combination is also tested.
573 *
574 * There can be a $default_charset setting in the
575 * config.php file, but the user may have a different language
576 * selected for a user interface. This function checks the
577 * language selected by the user and tags the outgoing messages
578 * with the appropriate charset corresponding to the language
579 * selection. This is "more right" (tm), than just stamping the
580 * message blindly with the system-wide $default_charset.
581 */
582 function set_my_charset(){
583 global $data_dir, $username, $default_charset, $languages, $squirrelmail_language;
584
585 $my_language = getPref($data_dir, $username, 'language');
586 if (!$my_language) {
587 $my_language = $squirrelmail_language ;
588 }
589 // Catch removed translation
590 if (!isset($languages[$my_language])) {
591 $my_language="en_US";
592 }
593 while (isset($languages[$my_language]['ALIAS'])) {
594 $my_language = $languages[$my_language]['ALIAS'];
595 }
596 $my_charset = $languages[$my_language]['CHARSET'];
597 if ($my_language!='en_US') {
598 $default_charset = $my_charset;
599 }
600 }
601
602 /**
603 * Replaces non-braking spaces inserted by some browsers with regular space
604 *
605 * This function can be used to replace non-braking space symbols
606 * that are inserted in forms by some browsers instead of normal
607 * space symbol.
608 *
609 * @param string $string Text that needs to be cleaned
610 * @param string $charset Charset used in text
611 * @return string Cleaned text
612 */
613 function cleanup_nbsp($string,$charset) {
614
615 // reduce number of case statements
616 if (stristr('iso-8859-',substr($charset,0,9))){
617 $output_charset="iso-8859-x";
618 }
619 if (stristr('windows-125',substr($charset,0,11))){
620 $output_charset="cp125x";
621 }
622 if (stristr('koi8',substr($charset,0,4))){
623 $output_charset="koi8-x";
624 }
625 if (! isset($output_charset)){
626 $output_charset=strtolower($charset);
627 }
628
629 // where is non-braking space symbol
630 switch($output_charset):
631 case "iso-8859-x":
632 case "cp125x":
633 case "iso-2022-jp":
634 $nbsp="\xA0";
635 break;
636 case "koi8-x":
637 $nbsp="\x9A";
638 break;
639 case "utf-8":
640 $nbsp="\xC2\xA0";
641 break;
642 default:
643 // don't change string if charset is unmatched
644 return $string;
645 endswitch;
646
647 // return space instead of non-braking space.
648 return str_replace($nbsp,' ',$string);
649 }
650
651 /**
652 * Function informs if it is safe to convert given charset to the one that is used by user.
653 *
654 * It is safe to use conversion only if user uses utf-8 encoding and when
655 * converted charset is similar to the one that is used by user.
656 *
657 * @param string $input_charset Charset of text that needs to be converted
658 * @return bool is it possible to convert to user's charset
659 */
660 function is_conversion_safe($input_charset) {
661 global $languages, $sm_notAlias, $default_charset, $lossy_encoding;
662
663 if (isset($lossy_encoding) && $lossy_encoding )
664 return true;
665
666 // convert to lower case
667 $input_charset = strtolower($input_charset);
668
669 // Is user's locale Unicode based ?
670 if ( $default_charset == "utf-8" ) {
671 return true;
672 }
673
674 // Charsets that are similar
675 switch ($default_charset) {
676 case "windows-1251":
677 if ( $input_charset == "iso-8859-5" ||
678 $input_charset == "koi8-r" ||
679 $input_charset == "koi8-u" ) {
680 return true;
681 } else {
682 return false;
683 }
684 case "windows-1257":
685 if ( $input_charset == "iso-8859-13" ||
686 $input_charset == "iso-8859-4" ) {
687 return true;
688 } else {
689 return false;
690 }
691 case "iso-8859-4":
692 if ( $input_charset == "iso-8859-13" ||
693 $input_charset == "windows-1257" ) {
694 return true;
695 } else {
696 return false;
697 }
698 case "iso-8859-5":
699 if ( $input_charset == "windows-1251" ||
700 $input_charset == "koi8-r" ||
701 $input_charset == "koi8-u" ) {
702 return true;
703 } else {
704 return false;
705 }
706 case "iso-8859-13":
707 if ( $input_charset == "iso-8859-4" ||
708 $input_charset == "windows-1257" ) {
709 return true;
710 } else {
711 return false;
712 }
713 case "koi8-r":
714 if ( $input_charset == "windows-1251" ||
715 $input_charset == "iso-8859-5" ||
716 $input_charset == "koi8-u" ) {
717 return true;
718 } else {
719 return false;
720 }
721 case "koi8-u":
722 if ( $input_charset == "windows-1251" ||
723 $input_charset == "iso-8859-5" ||
724 $input_charset == "koi8-r" ) {
725 return true;
726 } else {
727 return false;
728 }
729 default:
730 return false;
731 }
732 }
733
734 /**
735 * Converts html character entities to numeric entities
736 *
737 * SquirrelMail encoding functions work only with numeric entities.
738 * This function fixes issues with decoding functions that might convert
739 * some symbols to character entities. Issue is specific to PHP recode
740 * extension decoding. Function is used internally in charset_convert()
741 * function.
742 * @param string $str string that might contain html character entities
743 * @return string string with character entities converted to decimals.
744 * @since 1.5.2
745 */
746 function sqi18n_convert_entities($str) {
747
748 $entities = array(
749 // Latin 1
750 '&nbsp;' => '&#160;',
751 '&iexcl;' => '&#161;',
752 '&cent;' => '&#162;',
753 '&pound;' => '&#163;',
754 '&curren;' => '&#164;',
755 '&yen;' => '&#165;',
756 '&brvbar;' => '&#166;',
757 '&sect;' => '&#167;',
758 '&uml;' => '&#168;',
759 '&copy;' => '&#169;',
760 '&ordf;' => '&#170;',
761 '&laquo;' => '&#171;',
762 '&not;' => '&#172;',
763 '&shy;' => '&#173;',
764 '&reg;' => '&#174;',
765 '&macr;' => '&#175;',
766 '&deg;' => '&#176;',
767 '&plusmn;' => '&#177;',
768 '&sup2;' => '&#178;',
769 '&sup3;' => '&#179;',
770 '&acute;' => '&#180;',
771 '&micro;' => '&#181;',
772 '&para;' => '&#182;',
773 '&middot;' => '&#183;',
774 '&cedil;' => '&#184;',
775 '&sup1;' => '&#185;',
776 '&ordm;' => '&#186;',
777 '&raquo;' => '&#187;',
778 '&frac14;' => '&#188;',
779 '&frac12;' => '&#189;',
780 '&frac34;' => '&#190;',
781 '&iquest;' => '&#191;',
782 '&Agrave;' => '&#192;',
783 '&Aacute;' => '&#193;',
784 '&Acirc;' => '&#194;',
785 '&Atilde;' => '&#195;',
786 '&Auml;' => '&#196;',
787 '&Aring;' => '&#197;',
788 '&AElig;' => '&#198;',
789 '&Ccedil;' => '&#199;',
790 '&Egrave;' => '&#200;',
791 '&Eacute;' => '&#201;',
792 '&Ecirc;' => '&#202;',
793 '&Euml;' => '&#203;',
794 '&Igrave;' => '&#204;',
795 '&Iacute;' => '&#205;',
796 '&Icirc;' => '&#206;',
797 '&Iuml;' => '&#207;',
798 '&ETH;' => '&#208;',
799 '&Ntilde;' => '&#209;',
800 '&Ograve;' => '&#210;',
801 '&Oacute;' => '&#211;',
802 '&Ocirc;' => '&#212;',
803 '&Otilde;' => '&#213;',
804 '&Ouml;' => '&#214;',
805 '&times;' => '&#215;',
806 '&Oslash;' => '&#216;',
807 '&Ugrave;' => '&#217;',
808 '&Uacute;' => '&#218;',
809 '&Ucirc;' => '&#219;',
810 '&Uuml;' => '&#220;',
811 '&Yacute;' => '&#221;',
812 '&THORN;' => '&#222;',
813 '&szlig;' => '&#223;',
814 '&agrave;' => '&#224;',
815 '&aacute;' => '&#225;',
816 '&acirc;' => '&#226;',
817 '&atilde;' => '&#227;',
818 '&auml;' => '&#228;',
819 '&aring;' => '&#229;',
820 '&aelig;' => '&#230;',
821 '&ccedil;' => '&#231;',
822 '&egrave;' => '&#232;',
823 '&eacute;' => '&#233;',
824 '&ecirc;' => '&#234;',
825 '&euml;' => '&#235;',
826 '&igrave;' => '&#236;',
827 '&iacute;' => '&#237;',
828 '&icirc;' => '&#238;',
829 '&iuml;' => '&#239;',
830 '&eth;' => '&#240;',
831 '&ntilde;' => '&#241;',
832 '&ograve;' => '&#242;',
833 '&oacute;' => '&#243;',
834 '&ocirc;' => '&#244;',
835 '&otilde;' => '&#245;',
836 '&ouml;' => '&#246;',
837 '&divide;' => '&#247;',
838 '&oslash;' => '&#248;',
839 '&ugrave;' => '&#249;',
840 '&uacute;' => '&#250;',
841 '&ucirc;' => '&#251;',
842 '&uuml;' => '&#252;',
843 '&yacute;' => '&#253;',
844 '&thorn;' => '&#254;',
845 '&yuml;' => '&#255;',
846 // Latin Extended-A
847 '&OElig;' => '&#338;',
848 '&oelig;' => '&#339;',
849 '&Scaron;' => '&#352;',
850 '&scaron;' => '&#353;',
851 '&Yuml;' => '&#376;',
852 // Spacing Modifier Letters
853 '&circ;' => '&#710;',
854 '&tilde;' => '&#732;',
855 // General Punctuation
856 '&ensp;' => '&#8194;',
857 '&emsp;' => '&#8195;',
858 '&thinsp;' => '&#8201;',
859 '&zwnj;' => '&#8204;',
860 '&zwj;' => '&#8205;',
861 '&lrm;' => '&#8206;',
862 '&rlm;' => '&#8207;',
863 '&ndash;' => '&#8211;',
864 '&mdash;' => '&#8212;',
865 '&lsquo;' => '&#8216;',
866 '&rsquo;' => '&#8217;',
867 '&sbquo;' => '&#8218;',
868 '&ldquo;' => '&#8220;',
869 '&rdquo;' => '&#8221;',
870 '&bdquo;' => '&#8222;',
871 '&dagger;' => '&#8224;',
872 '&Dagger;' => '&#8225;',
873 '&permil;' => '&#8240;',
874 '&lsaquo;' => '&#8249;',
875 '&rsaquo;' => '&#8250;',
876 '&euro;' => '&#8364;',
877 // Latin Extended-B
878 '&fnof;' => '&#402;',
879 // Greek
880 '&Alpha;' => '&#913;',
881 '&Beta;' => '&#914;',
882 '&Gamma;' => '&#915;',
883 '&Delta;' => '&#916;',
884 '&Epsilon;' => '&#917;',
885 '&Zeta;' => '&#918;',
886 '&Eta;' => '&#919;',
887 '&Theta;' => '&#920;',
888 '&Iota;' => '&#921;',
889 '&Kappa;' => '&#922;',
890 '&Lambda;' => '&#923;',
891 '&Mu;' => '&#924;',
892 '&Nu;' => '&#925;',
893 '&Xi;' => '&#926;',
894 '&Omicron;' => '&#927;',
895 '&Pi;' => '&#928;',
896 '&Rho;' => '&#929;',
897 '&Sigma;' => '&#931;',
898 '&Tau;' => '&#932;',
899 '&Upsilon;' => '&#933;',
900 '&Phi;' => '&#934;',
901 '&Chi;' => '&#935;',
902 '&Psi;' => '&#936;',
903 '&Omega;' => '&#937;',
904 '&alpha;' => '&#945;',
905 '&beta;' => '&#946;',
906 '&gamma;' => '&#947;',
907 '&delta;' => '&#948;',
908 '&epsilon;' => '&#949;',
909 '&zeta;' => '&#950;',
910 '&eta;' => '&#951;',
911 '&theta;' => '&#952;',
912 '&iota;' => '&#953;',
913 '&kappa;' => '&#954;',
914 '&lambda;' => '&#955;',
915 '&mu;' => '&#956;',
916 '&nu;' => '&#957;',
917 '&xi;' => '&#958;',
918 '&omicron;' => '&#959;',
919 '&pi;' => '&#960;',
920 '&rho;' => '&#961;',
921 '&sigmaf;' => '&#962;',
922 '&sigma;' => '&#963;',
923 '&tau;' => '&#964;',
924 '&upsilon;' => '&#965;',
925 '&phi;' => '&#966;',
926 '&chi;' => '&#967;',
927 '&psi;' => '&#968;',
928 '&omega;' => '&#969;',
929 '&thetasym;' => '&#977;',
930 '&upsih;' => '&#978;',
931 '&piv;' => '&#982;',
932 // General Punctuation
933 '&bull;' => '&#8226;',
934 '&hellip;' => '&#8230;',
935 '&prime;' => '&#8242;',
936 '&Prime;' => '&#8243;',
937 '&oline;' => '&#8254;',
938 '&frasl;' => '&#8260;',
939 // Letterlike Symbols
940 '&weierp;' => '&#8472;',
941 '&image;' => '&#8465;',
942 '&real;' => '&#8476;',
943 '&trade;' => '&#8482;',
944 '&alefsym;' => '&#8501;',
945 // Arrows
946 '&larr;' => '&#8592;',
947 '&uarr;' => '&#8593;',
948 '&rarr;' => '&#8594;',
949 '&darr;' => '&#8595;',
950 '&harr;' => '&#8596;',
951 '&crarr;' => '&#8629;',
952 '&lArr;' => '&#8656;',
953 '&uArr;' => '&#8657;',
954 '&rArr;' => '&#8658;',
955 '&dArr;' => '&#8659;',
956 '&hArr;' => '&#8660;',
957 // Mathematical Operators
958 '&forall;' => '&#8704;',
959 '&part;' => '&#8706;',
960 '&exist;' => '&#8707;',
961 '&empty;' => '&#8709;',
962 '&nabla;' => '&#8711;',
963 '&isin;' => '&#8712;',
964 '&notin;' => '&#8713;',
965 '&ni;' => '&#8715;',
966 '&prod;' => '&#8719;',
967 '&sum;' => '&#8721;',
968 '&minus;' => '&#8722;',
969 '&lowast;' => '&#8727;',
970 '&radic;' => '&#8730;',
971 '&prop;' => '&#8733;',
972 '&infin;' => '&#8734;',
973 '&ang;' => '&#8736;',
974 '&and;' => '&#8743;',
975 '&or;' => '&#8744;',
976 '&cap;' => '&#8745;',
977 '&cup;' => '&#8746;',
978 '&int;' => '&#8747;',
979 '&there4;' => '&#8756;',
980 '&sim;' => '&#8764;',
981 '&cong;' => '&#8773;',
982 '&asymp;' => '&#8776;',
983 '&ne;' => '&#8800;',
984 '&equiv;' => '&#8801;',
985 '&le;' => '&#8804;',
986 '&ge;' => '&#8805;',
987 '&sub;' => '&#8834;',
988 '&sup;' => '&#8835;',
989 '&nsub;' => '&#8836;',
990 '&sube;' => '&#8838;',
991 '&supe;' => '&#8839;',
992 '&oplus;' => '&#8853;',
993 '&otimes;' => '&#8855;',
994 '&perp;' => '&#8869;',
995 '&sdot;' => '&#8901;',
996 // Miscellaneous Technical
997 '&lceil;' => '&#8968;',
998 '&rceil;' => '&#8969;',
999 '&lfloor;' => '&#8970;',
1000 '&rfloor;' => '&#8971;',
1001 '&lang;' => '&#9001;',
1002 '&rang;' => '&#9002;',
1003 // Geometric Shapes
1004 '&loz;' => '&#9674;',
1005 // Miscellaneous Symbols
1006 '&spades;' => '&#9824;',
1007 '&clubs;' => '&#9827;',
1008 '&hearts;' => '&#9829;',
1009 '&diams;' => '&#9830;');
1010
1011 $str = str_replace(array_keys($entities), array_values($entities), $str);
1012
1013 return $str;
1014 }
1015
1016 /* ------------------------------ main --------------------------- */
1017
1018 global $squirrelmail_language, $languages, $use_gettext;
1019
1020 if (! sqgetGlobalVar('squirrelmail_language',$squirrelmail_language,SQ_COOKIE)) {
1021 $squirrelmail_language = '';
1022 }
1023
1024 /**
1025 * Array specifies the available translations.
1026 *
1027 * Structure of array:
1028 * $languages['language']['variable'] = 'value'
1029 *
1030 * Possible 'variable' names:
1031 * NAME - Translation name in English
1032 * CHARSET - Encoding used by translation
1033 * ALIAS - used when 'language' is only short name and 'value' should provide long language name
1034 * ALTNAME - Native translation name. Any 8bit symbols must be html encoded.
1035 * LOCALE - Full locale name (in xx_XX.charset format). It can use array with more than one locale name since 1.4.5 and 1.5.1
1036 * DIR - Text direction. Used to define Right-to-Left languages. Possible values 'rtl' or 'ltr'. If undefined - defaults to 'ltr'
1037 * XTRA_CODE - translation uses special functions. See doc/i18n.txt
1038 *
1039 * Each 'language' definition requires NAME+CHARSET or ALIAS variables.
1040 *
1041 * @name $languages
1042 * @global array $languages
1043 */
1044 $languages['en_US']['NAME'] = 'English';
1045 $languages['en_US']['CHARSET'] = 'iso-8859-1';
1046 $languages['en_US']['LOCALE'] = 'en_US.ISO8859-1';
1047 $languages['en']['ALIAS'] = 'en_US';
1048
1049 /**
1050 * Automatic translation loading from setup.php files.
1051 * Solution for bug. 1240889.
1052 * setup.php file can contain $languages array entries and XTRA_CODE functions.
1053 */
1054 if (is_dir(SM_PATH . 'locale') &&
1055 is_readable(SM_PATH . 'locale')) {
1056 $localedir = dir(SM_PATH . 'locale');
1057 while($lang_dir=$localedir->read()) {
1058 // remove trailing slash, if present
1059 if (substr($lang_dir,-1)=='/') {
1060 $lang_dir = substr($lang_dir,0,-1);
1061 }
1062 if ($lang_dir != '..' && $lang_dir != '.' && $lang_dir != 'CVS' &&
1063 $lang_dir != '.svn' && is_dir(SM_PATH.'locale/'.$lang_dir) &&
1064 file_exists(SM_PATH.'locale/'.$lang_dir.'/setup.php')) {
1065 include_once(SM_PATH.'locale/'.$lang_dir.'/setup.php');
1066 }
1067 }
1068 $localedir->close();
1069 }
1070
1071 /* Detect whether gettext is installed. */
1072 $gettext_flags = 0;
1073 if (function_exists('_')) {
1074 $gettext_flags += 1;
1075 }
1076 if (function_exists('bindtextdomain')) {
1077 $gettext_flags += 2;
1078 }
1079 if (function_exists('textdomain')) {
1080 $gettext_flags += 4;
1081 }
1082 if (function_exists('ngettext')) {
1083 $gettext_flags += 8;
1084 }
1085
1086 /* If gettext is fully loaded, cool */
1087 if ($gettext_flags == 15) {
1088 $use_gettext = true;
1089 }
1090
1091 /* If ngettext support is missing, load it */
1092 elseif ($gettext_flags == 7) {
1093 $use_gettext = true;
1094 // load internal ngettext functions
1095 include_once(SM_PATH . 'class/l10n.class.php');
1096 include_once(SM_PATH . 'functions/ngettext.php');
1097 }
1098
1099 /* If we can fake gettext, try that */
1100 elseif ($gettext_flags == 0) {
1101 $use_gettext = true;
1102 include_once(SM_PATH . 'functions/gettext.php');
1103 } else {
1104 /* Uh-ho. A weird install */
1105 if (! $gettext_flags & 1) {
1106 /**
1107 * Function is used as replacement in broken installs
1108 * @ignore
1109 */
1110 function _($str) {
1111 return $str;
1112 }
1113 }
1114 if (! $gettext_flags & 2) {
1115 /**
1116 * Function is used as replacement in broken installs
1117 * @ignore
1118 */
1119 function bindtextdomain() {
1120 return;
1121 }
1122 }
1123 if (! $gettext_flags & 4) {
1124 /**
1125 * Function is used as replacemet in broken installs
1126 * @ignore
1127 */
1128 function textdomain() {
1129 return;
1130 }
1131 }
1132 if (! $gettext_flags & 8) {
1133 /**
1134 * Function is used as replacemet in broken installs
1135 * @ignore
1136 */
1137 function ngettext($str,$str2,$number) {
1138 if ($number>1) {
1139 return $str2;
1140 } else {
1141 return $str;
1142 }
1143 }
1144 }
1145 if (! function_exists('dgettext')) {
1146 /**
1147 * Replacement for broken setups.
1148 * @ignore
1149 */
1150 function dgettext($domain,$str) {
1151 return $str;
1152 }
1153 }
1154 if (! function_exists('dngettext')) {
1155 /**
1156 * Replacement for broken setups
1157 * @ignore
1158 */
1159 function dngettext($domain,$str1,$strn,$number) {
1160 return ($number==1 ? $str1 : $strn);
1161 }
1162 }
1163 }