?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
* Fix notice & blank text on nogomatch
[mediawiki.git] / includes / Parser.php
blob7cb0ad12816a60fdc1d2fd77cfb9ff2b91c58324
1 <?php
2
3 /**
4 * File for Parser and related classes
5 *
6 * @package MediaWiki
7 * @version $Id$
8 */
9
10 /**
11 * Variable substitution O(N^2) attack
12 *
13 * Without countermeasures, it would be possible to attack the parser by saving
14 * a page filled with a large number of inclusions of large pages. The size of
15 * the generated page would be proportional to the square of the input size.
16 * Hence, we limit the number of inclusions of any given page, thus bringing any
17 * attack back to O(N).
18 */
19 define( 'MAX_INCLUDE_REPEAT', 100 );
20 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
21
22 # Allowed values for $mOutputType
23 define( 'OT_HTML', 1 );
24 define( 'OT_WIKI', 2 );
25 define( 'OT_MSG' , 3 );
26
27 # string parameter for extractTags which will cause it
28 # to strip HTML comments in addition to regular
29 # <XML>-style tags. This should not be anything we
30 # may want to use in wikisyntax
31 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
32
33 # prefix for escaping, used in two functions at least
34 define( 'UNIQ_PREFIX', 'NaodW29');
35
36 # Constants needed for external link processing
37 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
38 define( 'HTTP_PROTOCOLS', 'http|https' );
39 # Everything except bracket, space, or control characters
40 define( 'EXT_LINK_URL_CLASS', '[^]\\x00-\\x20\\x7F]' );
41 define( 'INVERSE_EXT_LINK_URL_CLASS', '[\]\\x00-\\x20\\x7F]' );
42 # Including space
43 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
44 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
45 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
46 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS.'):'.EXT_LINK_URL_CLASS.'+) *('.EXT_LINK_TEXT_CLASS.'*?)\]/S' );
47 define( 'EXT_IMAGE_REGEX',
48 '/^('.HTTP_PROTOCOLS.':)'. # Protocol
49 '('.EXT_LINK_URL_CLASS.'+)\\/'. # Hostname and path
50 '('.EXT_IMAGE_FNAME_CLASS.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS.')$/S' # Filename
51 );
52
53 /**
54 * PHP Parser
55 *
56 * Processes wiki markup
57 *
58 * <pre>
59 * There are three main entry points into the Parser class:
60 * parse()
61 * produces HTML output
62 * preSaveTransform().
63 * produces altered wiki markup.
64 * transformMsg()
65 * performs brace substitution on MediaWiki messages
66 *
67 * Globals used:
68 * objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
69 *
70 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
71 *
72 * settings:
73 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
74 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
75 * $wgLocaltimezone
76 *
77 * * only within ParserOptions
78 * </pre>
79 *
80 * @package MediaWiki
81 */
82 class Parser
83 {
84 /**#@+
85 * @access private
86 */
87 # Persistent:
88 var $mTagHooks;
89
90 # Cleared with clearState():
91 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
92 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
93
94 # Temporary:
95 var $mOptions, $mTitle, $mOutputType,
96 $mTemplates, // cache of already loaded templates, avoids
97 // multiple SQL queries for the same string
98 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
99 // in this path. Used for loop detection.
100
101 /**#@-*/
102
103 /**
104 * Constructor
105 *
106 * @access public
107 */
108 function Parser() {
109 $this->mTemplates = array();
110 $this->mTemplatePath = array();
111 $this->mTagHooks = array();
112 $this->clearState();
113 }
114
115 /**
116 * Clear Parser state
117 *
118 * @access private
119 */
120 function clearState() {
121 $this->mOutput = new ParserOutput;
122 $this->mAutonumber = 0;
123 $this->mLastSection = "";
124 $this->mDTopen = false;
125 $this->mVariables = false;
126 $this->mIncludeCount = array();
127 $this->mStripState = array();
128 $this->mArgStack = array();
129 $this->mInPre = false;
130 }
131
132 /**
133 * First pass--just handle <nowiki> sections, pass the rest off
134 * to internalParse() which does all the real work.
135 *
136 * @access private
137 * @return ParserOutput a ParserOutput
138 */
139 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
140 global $wgUseTidy;
141 $fname = 'Parser::parse';
142 wfProfileIn( $fname );
143
144 if ( $clearState ) {
145 $this->clearState();
146 }
147
148 $this->mOptions = $options;
149 $this->mTitle =& $title;
150 $this->mOutputType = OT_HTML;
151
152 $stripState = NULL;
153 $text = $this->strip( $text, $this->mStripState );
154 $text = $this->internalParse( $text, $linestart );
155 $text = $this->unstrip( $text, $this->mStripState );
156 # Clean up special characters, only run once, next-to-last before doBlockLevels
157 if(!$wgUseTidy) {
158 $fixtags = array(
159 # french spaces, last one Guillemet-left
160 # only if there is something before the space
161 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1&nbsp;\\2',
162 # french spaces, Guillemet-right
163 "/(\\302\\253) /i"=>"\\1&nbsp;",
164 '/<hr *>/i' => '<hr />',
165 '/<br *>/i' => '<br />',
166 '/<center *>/i' => '<div class="center">',
167 '/<\\/center *>/i' => '</div>',
168 # Clean up spare ampersands; note that we probably ought to be
169 # more careful about named entities.
170 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&amp;'
171 );
172 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
173 } else {
174 $fixtags = array(
175 # french spaces, last one Guillemet-left
176 '/ (\\?|:|;|!|\\302\\273)/i' => '&nbsp;\\1',
177 # french spaces, Guillemet-right
178 '/(\\302\\253) /i' => '\\1&nbsp;',
179 '/<center *>/i' => '<div class="center">',
180 '/<\\/center *>/i' => '</div>'
181 );
182 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
183 }
184 # only once and last
185 $text = $this->doBlockLevels( $text, $linestart );
186 $text = $this->unstripNoWiki( $text, $this->mStripState );
187 $this->mOutput->setText( $text );
188 wfProfileOut( $fname );
189 return $this->mOutput;
190 }
191
192 /**
193 * Get a random string
194 *
195 * @access private
196 * @static
197 */
198 function getRandomString() {
199 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
200 }
201
202 /**
203 * Replaces all occurrences of <$tag>content</$tag> in the text
204 * with a random marker and returns the new text. the output parameter
205 * $content will be an associative array filled with data on the form
206 * $unique_marker => content.
207 *
208 * If $content is already set, the additional entries will be appended
209 * If $tag is set to STRIP_COMMENTS, the function will extract
210 * <!-- HTML comments -->
211 *
212 * @access private
213 * @static
214 */
215 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
216 $rnd = $uniq_prefix . '-' . $tag . Parser::getRandomString();
217 if ( !$content ) {
218 $content = array( );
219 }
220 $n = 1;
221 $stripped = '';
222
223 while ( '' != $text ) {
224 if($tag==STRIP_COMMENTS) {
225 $p = preg_split( '/<!--/i', $text, 2 );
226 } else {
227 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
228 }
229 $stripped .= $p[0];
230 if ( ( count( $p ) < 2 ) || ( '' == $p[1] ) ) {
231 $text = '';
232 } else {
233 if($tag==STRIP_COMMENTS) {
234 $q = preg_split( '/-->/i', $p[1], 2 );
235 } else {
236 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
237 }
238 $marker = $rnd . sprintf('%08X', $n++);
239 $content[$marker] = $q[0];
240 $stripped .= $marker;
241 $text = $q[1];
242 }
243 }
244 return $stripped;
245 }
246
247 /**
248 * Strips and renders nowiki, pre, math, hiero
249 * If $render is set, performs necessary rendering operations on plugins
250 * Returns the text, and fills an array with data needed in unstrip()
251 * If the $state is already a valid strip state, it adds to the state
252 *
253 * @param bool $stripcomments when set, HTML comments <!-- like this -->
254 * will be stripped in addition to other tags. This is important
255 * for section editing, where these comments cause confusion when
256 * counting the sections in the wikisource
257 *
258 * @access private
259 */
260 function strip( $text, &$state, $stripcomments = false ) {
261 $render = ($this->mOutputType == OT_HTML);
262 $html_content = array();
263 $nowiki_content = array();
264 $math_content = array();
265 $pre_content = array();
266 $comment_content = array();
267 $ext_content = array();
268
269 # Replace any instances of the placeholders
270 $uniq_prefix = UNIQ_PREFIX;
271 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
272
273 # html
274 global $wgRawHtml, $wgWhitelistEdit;
275 if( $wgRawHtml && $wgWhitelistEdit ) {
276 $text = Parser::extractTags('html', $text, $html_content, $uniq_prefix);
277 foreach( $html_content as $marker => $content ) {
278 if ($render ) {
279 # Raw and unchecked for validity.
280 $html_content[$marker] = $content;
281 } else {
282 $html_content[$marker] = '<html>'.$content.'</html>';
283 }
284 }
285 }
286
287 # nowiki
288 $text = Parser::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
289 foreach( $nowiki_content as $marker => $content ) {
290 if( $render ){
291 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
292 } else {
293 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
294 }
295 }
296
297 # math
298 $text = Parser::extractTags('math', $text, $math_content, $uniq_prefix);
299 foreach( $math_content as $marker => $content ){
300 if( $render ) {
301 if( $this->mOptions->getUseTeX() ) {
302 $math_content[$marker] = renderMath( $content );
303 } else {
304 $math_content[$marker] = '&lt;math&gt;'.$content.'&lt;math&gt;';
305 }
306 } else {
307 $math_content[$marker] = '<math>'.$content.'</math>';
308 }
309 }
310
311 # pre
312 $text = Parser::extractTags('pre', $text, $pre_content, $uniq_prefix);
313 foreach( $pre_content as $marker => $content ){
314 if( $render ){
315 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
316 } else {
317 $pre_content[$marker] = '<pre>'.$content.'</pre>';
318 }
319 }
320
321 # Comments
322 if($stripcomments) {
323 $text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
324 foreach( $comment_content as $marker => $content ){
325 $comment_content[$marker] = '<!--'.$content.'-->';
326 }
327 }
328
329 # Extensions
330 foreach ( $this->mTagHooks as $tag => $callback ) {
331 $ext_contents[$tag] = array();
332 $text = Parser::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
333 foreach( $ext_content[$tag] as $marker => $content ) {
334 if ( $render ) {
335 $ext_content[$tag][$marker] = $callback( $content );
336 } else {
337 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
338 }
339 }
340 }
341
342 # Merge state with the pre-existing state, if there is one
343 if ( $state ) {
344 $state['html'] = $state['html'] + $html_content;
345 $state['nowiki'] = $state['nowiki'] + $nowiki_content;
346 $state['math'] = $state['math'] + $math_content;
347 $state['pre'] = $state['pre'] + $pre_content;
348 $state['comment'] = $state['comment'] + $comment_content;
349
350 foreach( $ext_content as $tag => $array ) {
351 if ( array_key_exists( $tag, $state ) ) {
352 $state[$tag] = $state[$tag] + $array;
353 }
354 }
355 } else {
356 $state = array(
357 'html' => $html_content,
358 'nowiki' => $nowiki_content,
359 'math' => $math_content,
360 'pre' => $pre_content,
361 'comment' => $comment_content,
362 ) + $ext_content;
363 }
364 return $text;
365 }
366
367 /**
368 * restores pre, math, and heiro removed by strip()
369 *
370 * always call unstripNoWiki() after this one
371 * @access private
372 */
373 function unstrip( $text, &$state ) {
374 # Must expand in reverse order, otherwise nested tags will be corrupted
375 $contentDict = end( $state );
376 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
377 if( key($state) != 'nowiki' && key($state) != 'html') {
378 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
379 $text = str_replace( key( $contentDict ), $content, $text );
380 }
381 }
382 }
383
384 return $text;
385 }
386
387 /**
388 * always call this after unstrip() to preserve the order
389 *
390 * @access private
391 */
392 function unstripNoWiki( $text, &$state ) {
393 # Must expand in reverse order, otherwise nested tags will be corrupted
394 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
395 $text = str_replace( key( $state['nowiki'] ), $content, $text );
396 }
397
398 global $wgRawHtml;
399 if ($wgRawHtml) {
400 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
401 $text = str_replace( key( $state['html'] ), $content, $text );
402 }
403 }
404
405 return $text;
406 }
407
408 /**
409 * Add an item to the strip state
410 * Returns the unique tag which must be inserted into the stripped text
411 * The tag will be replaced with the original text in unstrip()
412 *
413 * @access private
414 */
415 function insertStripItem( $text, &$state ) {
416 $rnd = UNIQ_PREFIX . '-item' . Parser::getRandomString();
417 if ( !$state ) {
418 $state = array(
419 'html' => array(),
420 'nowiki' => array(),
421 'math' => array(),
422 'pre' => array()
423 );
424 }
425 $state['item'][$rnd] = $text;
426 return $rnd;
427 }
428
429 /**
430 * Return allowed HTML attributes
431 *
432 * @access private
433 */
434 function getHTMLattrs () {
435 $htmlattrs = array( # Allowed attributes--no scripting, etc.
436 'title', 'align', 'lang', 'dir', 'width', 'height',
437 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
438 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
439 /* FONT */ 'type', 'start', 'value', 'compact',
440 /* For various lists, mostly deprecated but safe */
441 'summary', 'width', 'border', 'frame', 'rules',
442 'cellspacing', 'cellpadding', 'valign', 'char',
443 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
444 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
445 'id', 'class', 'name', 'style' /* For CSS */
446 );
447 return $htmlattrs ;
448 }
449
450 /**
451 * Remove non approved attributes and javascript in css
452 *
453 * @access private
454 */
455 function fixTagAttributes ( $t ) {
456 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
457 $htmlattrs = $this->getHTMLattrs() ;
458
459 # Strip non-approved attributes from the tag
460 $t = preg_replace(
461 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
462 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
463 $t);
464
465 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
466
467 # Strip javascript "expression" from stylesheets. Brute force approach:
468 # If anythin offensive is found, all attributes of the HTML tag are dropped
469
470 if( preg_match(
471 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
472 wfMungeToUtf8( $t ) ) )
473 {
474 $t='';
475 }
476
477 return trim ( $t ) ;
478 }
479
480 /**
481 * interface with html tidy, used if $wgUseTidy = true
482 *
483 * @access public
484 * @static
485 */
486 function tidy ( $text ) {
487 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
488 global $wgInputEncoding, $wgOutputEncoding;
489 $fname = 'Parser::tidy';
490 wfProfileIn( $fname );
491
492 $cleansource = '';
493 $opts = '';
494 switch(strtoupper($wgOutputEncoding)) {
495 case 'ISO-8859-1':
496 $opts .= ($wgInputEncoding == $wgOutputEncoding)? ' -latin1':' -raw';
497 break;
498 case 'UTF-8':
499 $opts .= ($wgInputEncoding == $wgOutputEncoding)? ' -utf8':' -raw';
500 break;
501 default:
502 $opts .= ' -raw';
503 }
504
505 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
506 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
507 '<head><title>test</title></head><body>'.$text.'</body></html>';
508 $descriptorspec = array(
509 0 => array('pipe', 'r'),
510 1 => array('pipe', 'w'),
511 2 => array('file', '/dev/null', 'a')
512 );
513 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
514 if (is_resource($process)) {
515 fwrite($pipes[0], $wrappedtext);
516 fclose($pipes[0]);
517 while (!feof($pipes[1])) {
518 $cleansource .= fgets($pipes[1], 1024);
519 }
520 fclose($pipes[1]);
521 $return_value = proc_close($process);
522 }
523
524 wfProfileOut( $fname );
525
526 if( $cleansource == '' && $text != '') {
527 wfDebug( "Tidy error detected!\n" );
528 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
529 } else {
530 return $cleansource;
531 }
532 }
533
534 /**
535 * parse the wiki syntax used to render tables
536 *
537 * @access private
538 */
539 function doTableStuff ( $t ) {
540 $fname = 'Parser::doTableStuff';
541 wfProfileIn( $fname );
542
543 $t = explode ( "\n" , $t ) ;
544 $td = array () ; # Is currently a td tag open?
545 $ltd = array () ; # Was it TD or TH?
546 $tr = array () ; # Is currently a tr tag open?
547 $ltr = array () ; # tr attributes
548 $indent_level = 0; # indent level of the table
549 foreach ( $t AS $k => $x )
550 {
551 $x = trim ( $x ) ;
552 $fc = substr ( $x , 0 , 1 ) ;
553 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
554 $indent_level = strlen( $matches[1] );
555 $t[$k] = "\n" .
556 str_repeat( '<dl><dd>', $indent_level ) .
557 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
558 array_push ( $td , false ) ;
559 array_push ( $ltd , '' ) ;
560 array_push ( $tr , false ) ;
561 array_push ( $ltr , '' ) ;
562 }
563 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
564 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
565 $z = "</table>\n" ;
566 $l = array_pop ( $ltd ) ;
567 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
568 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
569 array_pop ( $ltr ) ;
570 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
571 }
572 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
573 $x = substr ( $x , 1 ) ;
574 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
575 $z = '' ;
576 $l = array_pop ( $ltd ) ;
577 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
578 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
579 array_pop ( $ltr ) ;
580 $t[$k] = $z ;
581 array_push ( $tr , false ) ;
582 array_push ( $td , false ) ;
583 array_push ( $ltd , '' ) ;
584 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
585 }
586 else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) { # Caption
587 # $x is a table row
588 if ( '|+' == substr ( $x , 0 , 2 ) ) {
589 $fc = '+' ;
590 $x = substr ( $x , 1 ) ;
591 }
592 $after = substr ( $x , 1 ) ;
593 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
594 $after = explode ( '||' , $after ) ;
595 $t[$k] = '' ;
596
597 # Loop through each table cell
598 foreach ( $after AS $theline )
599 {
600 $z = '' ;
601 if ( $fc != '+' )
602 {
603 $tra = array_pop ( $ltr ) ;
604 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
605 array_push ( $tr , true ) ;
606 array_push ( $ltr , '' ) ;
607 }
608
609 $l = array_pop ( $ltd ) ;
610 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
611 if ( $fc == '|' ) $l = 'td' ;
612 else if ( $fc == '!' ) $l = 'th' ;
613 else if ( $fc == '+' ) $l = 'caption' ;
614 else $l = '' ;
615 array_push ( $ltd , $l ) ;
616
617 # Cell parameters
618 $y = explode ( '|' , $theline , 2 ) ;
619 # Note that a '|' inside an invalid link should not
620 # be mistaken as delimiting cell parameters
621 if ( strpos( $y[0], '[[' ) !== false ) {
622 $y = array ($theline);
623 }
624 if ( count ( $y ) == 1 )
625 $y = "{$z}<{$l}>{$y[0]}" ;
626 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
627 $t[$k] .= $y ;
628 array_push ( $td , true ) ;
629 }
630 }
631 }
632
633 # Closing open td, tr && table
634 while ( count ( $td ) > 0 )
635 {
636 if ( array_pop ( $td ) ) $t[] = '</td>' ;
637 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
638 $t[] = '</table>' ;
639 }
640
641 $t = implode ( "\n" , $t ) ;
642 # $t = $this->removeHTMLtags( $t );
643 wfProfileOut( $fname );
644 return $t ;
645 }
646
647 /**
648 * Helper function for parse() that transforms wiki markup into
649 * HTML. Only called for $mOutputType == OT_HTML.
650 *
651 * @access private
652 */
653 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
654 global $wgContLang;
655
656 $fname = 'Parser::internalParse';
657 wfProfileIn( $fname );
658
659 $text = $this->removeHTMLtags( $text );
660 $text = $this->replaceVariables( $text, $args );
661
662 $text = $wgContLang->convert($text);
663
664 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
665
666 $text = $this->doHeadings( $text );
667 if($this->mOptions->getUseDynamicDates()) {
668 global $wgDateFormatter;
669 $text = $wgDateFormatter->reformat( $this->mOptions->getDateFormat(), $text );
670 }
671 $text = $this->doAllQuotes( $text );
672 $text = $this->replaceInternalLinks ( $text );
673 # Another call to replace links and images inside captions of images
674 $text = $this->replaceInternalLinks ( $text );
675 $text = $this->replaceExternalLinks( $text );
676 $text = $this->doMagicLinks( $text );
677 $text = $this->doTableStuff( $text );
678 $text = $this->formatHeadings( $text, $isMain );
679 $sk =& $this->mOptions->getSkin();
680 $text = $sk->transformContent( $text );
681
682 wfProfileOut( $fname );
683 return $text;
684 }
685
686 /**
687 * Replace special strings like "ISBN xxx" and "RFC xxx" with
688 * magic external links.
689 *
690 * @access private
691 */
692 function &doMagicLinks( &$text ) {
693 global $wgUseGeoMode;
694 $text = $this->magicISBN( $text );
695 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
696 $text = $this->magicGEO( $text );
697 }
698 $text = $this->magicRFC( $text );
699 return $text;
700 }
701
702 /**
703 * Parse ^^ tokens and return html
704 *
705 * @access private
706 */
707 function doExponent ( $text ) {
708 $fname = 'Parser::doExponent';
709 wfProfileIn( $fname);
710 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
711 wfProfileOut( $fname);
712 return $text;
713 }
714
715 /**
716 * Parse headers and return html
717 *
718 * @access private
719 */
720 function doHeadings( $text ) {
721 $fname = 'Parser::doHeadings';
722 wfProfileIn( $fname );
723 for ( $i = 6; $i >= 1; --$i ) {
724 $h = substr( '======', 0, $i );
725 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
726 "<h{$i}>\\1</h{$i}>\\2", $text );
727 }
728 wfProfileOut( $fname );
729 return $text;
730 }
731
732 /**
733 * Replace single quotes with HTML markup
734 * @access private
735 * @return string the altered text
736 */
737 function doAllQuotes( $text ) {
738 $fname = 'Parser::doAllQuotes';
739 wfProfileIn( $fname );
740 $outtext = '';
741 $lines = explode( "\n", $text );
742 foreach ( $lines as $line ) {
743 $outtext .= $this->doQuotes ( $line ) . "\n";
744 }
745 $outtext = substr($outtext, 0,-1);
746 wfProfileOut( $fname );
747 return $outtext;
748 }
749
750 /**
751 * Helper function for doAllQuotes()
752 * @access private
753 */
754 function doQuotes( $text ) {
755 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE);
756 if (count ($arr) == 1)
757 return $text;
758 else
759 {
760 # First, do some preliminary work. This may shift some apostrophes from
761 # being mark-up to being text. It also counts the number of occurrences
762 # of bold and italics mark-ups.
763 $i = 0;
764 $numbold = 0;
765 $numitalics = 0;
766 foreach ($arr as $r)
767 {
768 if (($i % 2) == 1)
769 {
770 # If there are ever four apostrophes, assume the first is supposed to
771 # be text, and the remaining three constitute mark-up for bold text.
772 if (strlen ($arr[$i]) == 4)
773 {
774 $arr[$i-1] .= "'";
775 $arr[$i] = "'''";
776 }
777 # If there are more than 5 apostrophes in a row, assume they're all
778 # text except for the last 5.
779 else if (strlen ($arr[$i]) > 5)
780 {
781 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
782 $arr[$i] = "'''''";
783 }
784 # Count the number of occurrences of bold and italics mark-ups.
785 # We are not counting sequences of five apostrophes.
786 if (strlen ($arr[$i]) == 2) $numitalics++; else
787 if (strlen ($arr[$i]) == 3) $numbold++; else
788 if (strlen ($arr[$i]) == 5) { $numitalics++; $numbold++; }
789 }
790 $i++;
791 }
792
793 # If there is an odd number of both bold and italics, it is likely
794 # that one of the bold ones was meant to be an apostrophe followed
795 # by italics. Which one we cannot know for certain, but it is more
796 # likely to be one that has a single-letter word before it.
797 if (($numbold % 2 == 1) && ($numitalics % 2 == 1))
798 {
799 $i = 0;
800 $firstsingleletterword = -1;
801 $firstmultiletterword = -1;
802 $firstspace = -1;
803 foreach ($arr as $r)
804 {
805 if (($i % 2 == 1) and (strlen ($r) == 3))
806 {
807 $x1 = substr ($arr[$i-1], -1);
808 $x2 = substr ($arr[$i-1], -2, 1);
809 if ($x1 == ' ') {
810 if ($firstspace == -1) $firstspace = $i;
811 } else if ($x2 == ' ') {
812 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
813 } else {
814 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
815 }
816 }
817 $i++;
818 }
819
820 # If there is a single-letter word, use it!
821 if ($firstsingleletterword > -1)
822 {
823 $arr [ $firstsingleletterword ] = "''";
824 $arr [ $firstsingleletterword-1 ] .= "'";
825 }
826 # If not, but there's a multi-letter word, use that one.
827 else if ($firstmultiletterword > -1)
828 {
829 $arr [ $firstmultiletterword ] = "''";
830 $arr [ $firstmultiletterword-1 ] .= "'";
831 }
832 # ... otherwise use the first one that has neither.
833 # (notice that it is possible for all three to be -1 if, for example,
834 # there is only one pentuple-apostrophe in the line)
835 else if ($firstspace > -1)
836 {
837 $arr [ $firstspace ] = "''";
838 $arr [ $firstspace-1 ] .= "'";
839 }
840 }
841
842 # Now let's actually convert our apostrophic mush to HTML!
843 $output = '';
844 $buffer = '';
845 $state = '';
846 $i = 0;
847 foreach ($arr as $r)
848 {
849 if (($i % 2) == 0)
850 {
851 if ($state == 'both')
852 $buffer .= $r;
853 else
854 $output .= $r;
855 }
856 else
857 {
858 if (strlen ($r) == 2)
859 {
860 if ($state == 'i')
861 { $output .= '</i>'; $state = ''; }
862 else if ($state == 'bi')
863 { $output .= '</i>'; $state = 'b'; }
864 else if ($state == 'ib')
865 { $output .= '</b></i><b>'; $state = 'b'; }
866 else if ($state == 'both')
867 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
868 else # $state can be 'b' or ''
869 { $output .= '<i>'; $state .= 'i'; }
870 }
871 else if (strlen ($r) == 3)
872 {
873 if ($state == 'b')
874 { $output .= '</b>'; $state = ''; }
875 else if ($state == 'bi')
876 { $output .= '</i></b><i>'; $state = 'i'; }
877 else if ($state == 'ib')
878 { $output .= '</b>'; $state = 'i'; }
879 else if ($state == 'both')
880 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
881 else # $state can be 'i' or ''
882 { $output .= '<b>'; $state .= 'b'; }
883 }
884 else if (strlen ($r) == 5)
885 {
886 if ($state == 'b')
887 { $output .= '</b><i>'; $state = 'i'; }
888 else if ($state == 'i')
889 { $output .= '</i><b>'; $state = 'b'; }
890 else if ($state == 'bi')
891 { $output .= '</i></b>'; $state = ''; }
892 else if ($state == 'ib')
893 { $output .= '</b></i>'; $state = ''; }
894 else if ($state == 'both')
895 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
896 else # ($state == '')
897 { $buffer = ''; $state = 'both'; }
898 }
899 }
900 $i++;
901 }
902 # Now close all remaining tags. Notice that the order is important.
903 if ($state == 'b' || $state == 'ib')
904 $output .= '</b>';
905 if ($state == 'i' || $state == 'bi' || $state == 'ib')
906 $output .= '</i>';
907 if ($state == 'bi')
908 $output .= '</b>';
909 if ($state == 'both')
910 $output .= '<b><i>'.$buffer.'</i></b>';
911 return $output;
912 }
913 }
914
915 /**
916 * Replace external links
917 *
918 * Note: we have to do external links before the internal ones,
919 * and otherwise take great care in the order of things here, so
920 * that we don't end up interpreting some URLs twice.
921 *
922 * @access private
923 */
924 function replaceExternalLinks( $text ) {
925 $fname = 'Parser::replaceExternalLinks';
926 wfProfileIn( $fname );
927
928 $sk =& $this->mOptions->getSkin();
929 $linktrail = wfMsgForContent('linktrail');
930 $bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
931
932 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
933
934 $i = 0;
935 while ( $i<count( $bits ) ) {
936 $url = $bits[$i++];
937 $protocol = $bits[$i++];
938 $text = $bits[$i++];
939 $trail = $bits[$i++];
940
941 # If the link text is an image URL, replace it with an <img> tag
942 # This happened by accident in the original parser, but some people used it extensively
943 $img = $this->maybeMakeImageLink( $text );
944 if ( $img !== false ) {
945 $text = $img;
946 }
947
948 $dtrail = '';
949
950 # No link text, e.g. [http://domain.tld/some.link]
951 if ( $text == '' ) {
952 # Autonumber if allowed
953 if ( strpos( HTTP_PROTOCOLS, $protocol ) !== false ) {
954 $text = '[' . ++$this->mAutonumber . ']';
955 } else {
956 # Otherwise just use the URL
957 $text = htmlspecialchars( $url );
958 }
959 } else {
960 # Have link text, e.g. [http://domain.tld/some.link text]s
961 # Check for trail
962 if ( preg_match( $linktrail, $trail, $m2 ) ) {
963 $dtrail = $m2[1];
964 $trail = $m2[2];
965 }
966 }
967
968 $encUrl = htmlspecialchars( $url );
969 # Bit in parentheses showing the URL for the printable version
970 if( $url == $text || preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
971 $paren = '';
972 } else {
973 # Expand the URL for printable version
974 if ( ! $sk->suppressUrlExpansion() ) {
975 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
976 } else {
977 $paren = '';
978 }
979 }
980
981 # Process the trail (i.e. everything after this link up until start of the next link),
982 # replacing any non-bracketed links
983 $trail = $this->replaceFreeExternalLinks( $trail );
984
985 $la = $sk->getExternalLinkAttributes( $url, $text );
986
987 # Use the encoded URL
988 # This means that users can paste URLs directly into the text
989 # Funny characters like &ouml; aren't valid in URLs anyway
990 # This was changed in August 2004
991 $s .= "<a href=\"{$url}\"{$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
992 }
993
994 wfProfileOut( $fname );
995 return $s;
996 }
997
998 /**
999 * Replace anything that looks like a URL with a link
1000 * @access private
1001 */
1002 function replaceFreeExternalLinks( $text ) {
1003 $bits = preg_split( '/((?:'.URL_PROTOCOLS.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1004 $s = array_shift( $bits );
1005 $i = 0;
1006
1007 $sk =& $this->mOptions->getSkin();
1008
1009 while ( $i < count( $bits ) ){
1010 $protocol = $bits[$i++];
1011 $remainder = $bits[$i++];
1012
1013 if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
1014 # Found some characters after the protocol that look promising
1015 $url = $protocol . $m[1];
1016 $trail = $m[2];
1017
1018 # Move trailing punctuation to $trail
1019 $sep = ',;\.:!?';
1020 # If there is no left bracket, then consider right brackets fair game too
1021 if ( strpos( $url, '(' ) === false ) {
1022 $sep .= ')';
1023 }
1024
1025 $numSepChars = strspn( strrev( $url ), $sep );
1026 if ( $numSepChars ) {
1027 $trail = substr( $url, -$numSepChars ) . $trail;
1028 $url = substr( $url, 0, -$numSepChars );
1029 }
1030
1031 # Replace &amp; from obsolete syntax with &
1032 $url = str_replace( '&amp;', '&', $url );
1033
1034 # Is this an external image?
1035 $text = $this->maybeMakeImageLink( $url );
1036 if ( $text === false ) {
1037 # Not an image, make a link
1038 $text = $sk->makeExternalLink( $url, $url );
1039 }
1040 $s .= $text . $trail;
1041 } else {
1042 $s .= $protocol . $remainder;
1043 }
1044 }
1045 return $s;
1046 }
1047
1048 /**
1049 * make an image if it's allowed
1050 * @access private
1051 */
1052 function maybeMakeImageLink( $url ) {
1053 $sk =& $this->mOptions->getSkin();
1054 $text = false;
1055 if ( $this->mOptions->getAllowExternalImages() ) {
1056 if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
1057 # Image found
1058 $text = $sk->makeImage( htmlspecialchars( $url ) );
1059 }
1060 }
1061 return $text;
1062 }
1063
1064 /**
1065 * Process [[ ]] wikilinks
1066 *
1067 * @access private
1068 */
1069 function replaceInternalLinks( $s ) {
1070 global $wgLang, $wgContLang, $wgLinkCache;
1071 static $fname = 'Parser::replaceInternalLinks' ;
1072 wfProfileIn( $fname );
1073
1074 wfProfileIn( $fname.'-setup' );
1075 static $tc = FALSE;
1076 # the % is needed to support urlencoded titles as well
1077 if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
1078 $sk =& $this->mOptions->getSkin();
1079
1080 $redirect = MagicWord::get ( MAG_REDIRECT ) ;
1081
1082 $a = explode( '[[', ' ' . $s );
1083 $s = array_shift( $a );
1084 $s = substr( $s, 1 );
1085
1086 # Match a link having the form [[namespace:link|alternate]]trail
1087 static $e1 = FALSE;
1088 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1089 # Match the end of a line for a word that's not followed by whitespace,
1090 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1091 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1092
1093 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1094 # Special and Media are pseudo-namespaces; no pages actually exist in them
1095
1096 $nottalk = !Namespace::isTalk( $this->mTitle->getNamespace() );
1097
1098 if ( $useLinkPrefixExtension ) {
1099 if ( preg_match( $e2, $s, $m ) ) {
1100 $first_prefix = $m[2];
1101 $s = $m[1];
1102 } else {
1103 $first_prefix = false;
1104 }
1105 } else {
1106 $prefix = '';
1107 }
1108
1109 wfProfileOut( $fname.'-setup' );
1110
1111 # start procedeeding each line
1112 foreach ( $a as $line ) {
1113 wfProfileIn( $fname.'-prefixhandling' );
1114 if ( $useLinkPrefixExtension ) {
1115 if ( preg_match( $e2, $s, $m ) ) {
1116 $prefix = $m[2];
1117 $s = $m[1];
1118 } else {
1119 $prefix='';
1120 }
1121 # first link
1122 if($first_prefix) {
1123 $prefix = $first_prefix;
1124 $first_prefix = false;
1125 }
1126 }
1127 wfProfileOut( $fname.'-prefixhandling' );
1128
1129 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1130 $text = $m[2];
1131 # fix up urlencoded title texts
1132 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1133 $trail = $m[3];
1134 } else { # Invalid form; output directly
1135 $s .= $prefix . '[[' . $line ;
1136 continue;
1137 }
1138
1139 # Don't allow internal links to pages containing
1140 # PROTO: where PROTO is a valid URL protocol; these
1141 # should be external links.
1142 if (preg_match('/((?:'.URL_PROTOCOLS.'):)/', $m[1])) {
1143 $s .= $prefix . '[[' . $line ;
1144 continue;
1145 }
1146
1147 # Make subpage if necessary
1148 $link = $this->maybeDoSubpageLink( $m[1], $text );
1149
1150 $noforce = (substr($m[1], 0, 1) != ':');
1151 if (!$noforce) {
1152 # Strip off leading ':'
1153 $link = substr($link, 1);
1154 }
1155
1156 $wasblank = ( '' == $text );
1157 if( $wasblank ) $text = $link;
1158
1159 $nt = Title::newFromText( $link );
1160 if( !$nt ) {
1161 $s .= $prefix . '[[' . $line;
1162 continue;
1163 }
1164
1165 $ns = $nt->getNamespace();
1166 $iw = $nt->getInterWiki();
1167
1168 # Link not escaped by : , create the various objects
1169 if( $noforce ) {
1170
1171 # Interwikis
1172 if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1173 array_push( $this->mOutput->mLanguageLinks, $nt->getFullText() );
1174 $tmp = $prefix . $trail ;
1175 $s .= (trim($tmp) == '')? '': $tmp;
1176 continue;
1177 }
1178
1179 if ( $ns == NS_IMAGE ) {
1180 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1181 $wgLinkCache->addImageLinkObj( $nt );
1182 continue;
1183 }
1184
1185 if ( $ns == NS_CATEGORY ) {
1186 $t = $nt->getText() ;
1187 $nnt = Title::newFromText ( Namespace::getCanonicalName(NS_CATEGORY).':'.$t ) ;
1188
1189 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1190 $pPLC=$sk->postParseLinkColour();
1191 $sk->postParseLinkColour( false );
1192 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1193 $sk->postParseLinkColour( $pPLC );
1194 $wgLinkCache->resume();
1195
1196 if ( $wasblank ) {
1197 if ( $this->mTitle->getNamespace() == NS_CATEGORY ) {
1198 $sortkey = $this->mTitle->getText();
1199 } else {
1200 $sortkey = $this->mTitle->getPrefixedText();
1201 }
1202 } else {
1203 $sortkey = $text;
1204 }
1205 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1206 $this->mOutput->mCategoryLinks[] = $t ;
1207 $s .= $prefix . $trail ;
1208 continue;
1209 }
1210 }
1211
1212 if( ( $nt->getPrefixedText() === $this->mTitle->getPrefixedText() ) &&
1213 ( strpos( $link, '#' ) === FALSE ) ) {
1214 # Self-links are handled specially; generally de-link and change to bold.
1215 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1216 continue;
1217 }
1218
1219 if( $ns == NS_MEDIA ) {
1220 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1221 $wgLinkCache->addImageLinkObj( $nt );
1222 continue;
1223 } elseif( $ns == NS_SPECIAL ) {
1224 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1225 continue;
1226 }
1227 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1228 }
1229 wfProfileOut( $fname );
1230 return $s;
1231 }
1232
1233 /**
1234 * Handle link to subpage if necessary
1235 * @param $target string the source of the link
1236 * @param &$text the link text, modified as necessary
1237 * @return string the full name of the link
1238 * @access private
1239 */
1240 function maybeDoSubpageLink($target, &$text) {
1241 # Valid link forms:
1242 # Foobar -- normal
1243 # :Foobar -- override special treatment of prefix (images, language links)
1244 # /Foobar -- convert to CurrentPage/Foobar
1245 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1246 global $wgNamespacesWithSubpages;
1247
1248 $fname = 'Parser::maybeDoSubpageLink';
1249 wfProfileIn( $fname );
1250 # Look at the first character
1251 if( $target{0} == '/' ) {
1252 # / at end means we don't want the slash to be shown
1253 if(substr($target,-1,1)=='/') {
1254 $target=substr($target,1,-1);
1255 $noslash=$target;
1256 } else {
1257 $noslash=substr($target,1);
1258 }
1259
1260 # Some namespaces don't allow subpages
1261 if(!empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()])) {
1262 # subpages allowed here
1263 $ret = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
1264 if( '' === $text ) {
1265 $text = $target;
1266 } # this might be changed for ugliness reasons
1267 } else {
1268 # no subpage allowed, use standard link
1269 $ret = $target;
1270 }
1271 } else {
1272 # no subpage
1273 $ret = $target;
1274 }
1275
1276 wfProfileOut( $fname );
1277 return $ret;
1278 }
1279
1280 /**#@+
1281 * Used by doBlockLevels()
1282 * @access private
1283 */
1284 /* private */ function closeParagraph() {
1285 $result = '';
1286 if ( '' != $this->mLastSection ) {
1287 $result = '</' . $this->mLastSection . ">\n";
1288 }
1289 $this->mInPre = false;
1290 $this->mLastSection = '';
1291 return $result;
1292 }
1293 # getCommon() returns the length of the longest common substring
1294 # of both arguments, starting at the beginning of both.
1295 #
1296 /* private */ function getCommon( $st1, $st2 ) {
1297 $fl = strlen( $st1 );
1298 $shorter = strlen( $st2 );
1299 if ( $fl < $shorter ) { $shorter = $fl; }
1300
1301 for ( $i = 0; $i < $shorter; ++$i ) {
1302 if ( $st1{$i} != $st2{$i} ) { break; }
1303 }
1304 return $i;
1305 }
1306 # These next three functions open, continue, and close the list
1307 # element appropriate to the prefix character passed into them.
1308 #
1309 /* private */ function openList( $char ) {
1310 $result = $this->closeParagraph();
1311
1312 if ( '*' == $char ) { $result .= '<ul><li>'; }
1313 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1314 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1315 else if ( ';' == $char ) {
1316 $result .= '<dl><dt>';
1317 $this->mDTopen = true;
1318 }
1319 else { $result = '<!-- ERR 1 -->'; }
1320
1321 return $result;
1322 }
1323
1324 /* private */ function nextItem( $char ) {
1325 if ( '*' == $char || '#' == $char ) { return '</li><li>'; }
1326 else if ( ':' == $char || ';' == $char ) {
1327 $close = '</dd>';
1328 if ( $this->mDTopen ) { $close = '</dt>'; }
1329 if ( ';' == $char ) {
1330 $this->mDTopen = true;
1331 return $close . '<dt>';
1332 } else {
1333 $this->mDTopen = false;
1334 return $close . '<dd>';
1335 }
1336 }
1337 return '<!-- ERR 2 -->';
1338 }
1339
1340 /* private */ function closeList( $char ) {
1341 if ( '*' == $char ) { $text = '</li></ul>'; }
1342 else if ( '#' == $char ) { $text = '</li></ol>'; }
1343 else if ( ':' == $char ) {
1344 if ( $this->mDTopen ) {
1345 $this->mDTopen = false;
1346 $text = '</dt></dl>';
1347 } else {
1348 $text = '</dd></dl>';
1349 }
1350 }
1351 else { return '<!-- ERR 3 -->'; }
1352 return $text."\n";
1353 }
1354 /**#@-*/
1355
1356 /**
1357 * Make lists from lines starting with ':', '*', '#', etc.
1358 *
1359 * @access private
1360 * @return string the lists rendered as HTML
1361 */
1362 function doBlockLevels( $text, $linestart ) {
1363 $fname = 'Parser::doBlockLevels';
1364 wfProfileIn( $fname );
1365
1366 # Parsing through the text line by line. The main thing
1367 # happening here is handling of block-level elements p, pre,
1368 # and making lists from lines starting with * # : etc.
1369 #
1370 $textLines = explode( "\n", $text );
1371
1372 $lastPrefix = $output = $lastLine = '';
1373 $this->mDTopen = $inBlockElem = false;
1374 $prefixLength = 0;
1375 $paragraphStack = false;
1376
1377 if ( !$linestart ) {
1378 $output .= array_shift( $textLines );
1379 }
1380 foreach ( $textLines as $oLine ) {
1381 $lastPrefixLength = strlen( $lastPrefix );
1382 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1383 $preOpenMatch = preg_match('/<pre/i', $oLine );
1384 if ( !$this->mInPre ) {
1385 # Multiple prefixes may abut each other for nested lists.
1386 $prefixLength = strspn( $oLine, '*#:;' );
1387 $pref = substr( $oLine, 0, $prefixLength );
1388
1389 # eh?
1390 $pref2 = str_replace( ';', ':', $pref );
1391 $t = substr( $oLine, $prefixLength );
1392 $this->mInPre = !empty($preOpenMatch);
1393 } else {
1394 # Don't interpret any other prefixes in preformatted text
1395 $prefixLength = 0;
1396 $pref = $pref2 = '';
1397 $t = $oLine;
1398 }
1399
1400 # List generation
1401 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1402 # Same as the last item, so no need to deal with nesting or opening stuff
1403 $output .= $this->nextItem( substr( $pref, -1 ) );
1404 $paragraphStack = false;
1405
1406 if ( substr( $pref, -1 ) == ';') {
1407 # The one nasty exception: definition lists work like this:
1408 # ; title : definition text
1409 # So we check for : in the remainder text to split up the
1410 # title and definition, without b0rking links.
1411 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1412 $t = $t2;
1413 $output .= $term . $this->nextItem( ':' );
1414 }
1415 }
1416 } elseif( $prefixLength || $lastPrefixLength ) {
1417 # Either open or close a level...
1418 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1419 $paragraphStack = false;
1420
1421 while( $commonPrefixLength < $lastPrefixLength ) {
1422 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1423 --$lastPrefixLength;
1424 }
1425 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1426 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1427 }
1428 while ( $prefixLength > $commonPrefixLength ) {
1429 $char = substr( $pref, $commonPrefixLength, 1 );
1430 $output .= $this->openList( $char );
1431
1432 if ( ';' == $char ) {
1433 # FIXME: This is dupe of code above
1434 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1435 $t = $t2;
1436 $output .= $term . $this->nextItem( ':' );
1437 }
1438 }
1439 ++$commonPrefixLength;
1440 }
1441 $lastPrefix = $pref2;
1442 }
1443 if( 0 == $prefixLength ) {
1444 # No prefix (not in list)--go to paragraph mode
1445 $uniq_prefix = UNIQ_PREFIX;
1446 // XXX: use a stack for nestable elements like span, table and div
1447 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1448 $closematch = preg_match(
1449 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1450 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1451 if ( $openmatch or $closematch ) {
1452 $paragraphStack = false;
1453 $output .= $this->closeParagraph();
1454 if($preOpenMatch and !$preCloseMatch) {
1455 $this->mInPre = true;
1456 }
1457 if ( $closematch ) {
1458 $inBlockElem = false;
1459 } else {
1460 $inBlockElem = true;
1461 }
1462 } else if ( !$inBlockElem && !$this->mInPre ) {
1463 if ( ' ' == $t{0} and ( $this->mLastSection == 'pre' or trim($t) != '' ) ) {
1464 // pre
1465 if ($this->mLastSection != 'pre') {
1466 $paragraphStack = false;
1467 $output .= $this->closeParagraph().'<pre>';
1468 $this->mLastSection = 'pre';
1469 }
1470 $t = substr( $t, 1 );
1471 } else {
1472 // paragraph
1473 if ( '' == trim($t) ) {
1474 if ( $paragraphStack ) {
1475 $output .= $paragraphStack.'<br />';
1476 $paragraphStack = false;
1477 $this->mLastSection = 'p';
1478 } else {
1479 if ($this->mLastSection != 'p' ) {
1480 $output .= $this->closeParagraph();
1481 $this->mLastSection = '';
1482 $paragraphStack = '<p>';
1483 } else {
1484 $paragraphStack = '</p><p>';
1485 }
1486 }
1487 } else {
1488 if ( $paragraphStack ) {
1489 $output .= $paragraphStack;
1490 $paragraphStack = false;
1491 $this->mLastSection = 'p';
1492 } else if ($this->mLastSection != 'p') {
1493 $output .= $this->closeParagraph().'<p>';
1494 $this->mLastSection = 'p';
1495 }
1496 }
1497 }
1498 }
1499 }
1500 if ($paragraphStack === false) {
1501 $output .= $t."\n";
1502 }
1503 }
1504 while ( $prefixLength ) {
1505 $output .= $this->closeList( $pref2{$prefixLength-1} );
1506 --$prefixLength;
1507 }
1508 if ( '' != $this->mLastSection ) {
1509 $output .= '</' . $this->mLastSection . '>';
1510 $this->mLastSection = '';
1511 }
1512
1513 wfProfileOut( $fname );
1514 return $output;
1515 }
1516
1517 /**
1518 * Split up a string on ':', ignoring any occurences inside
1519 * <a>..</a> or <span>...</span>
1520 * @param $str string the string to split
1521 * @param &$before string set to everything before the ':'
1522 * @param &$after string set to everything after the ':'
1523 * return string the position of the ':', or false if none found
1524 */
1525 function findColonNoLinks($str, &$before, &$after) {
1526 # I wonder if we should make this count all tags, not just <a>
1527 # and <span>. That would prevent us from matching a ':' that
1528 # comes in the middle of italics other such formatting....
1529 # -- Wil
1530 $fname = 'Parser::findColonNoLinks';
1531 wfProfileIn( $fname );
1532 $pos = 0;
1533 do {
1534 $colon = strpos($str, ':', $pos);
1535
1536 if ($colon !== false) {
1537 $before = substr($str, 0, $colon);
1538 $after = substr($str, $colon + 1);
1539
1540 # Skip any ':' within <a> or <span> pairs
1541 $a = substr_count($before, '<a');
1542 $s = substr_count($before, '<span');
1543 $ca = substr_count($before, '</a>');
1544 $cs = substr_count($before, '</span>');
1545
1546 if ($a <= $ca and $s <= $cs) {
1547 # Tags are balanced before ':'; ok
1548 break;
1549 }
1550 $pos = $colon + 1;
1551 }
1552 } while ($colon !== false);
1553 wfProfileOut( $fname );
1554 return $colon;
1555 }
1556
1557 /**
1558 * Return value of a magic variable (like PAGENAME)
1559 *
1560 * @access private
1561 */
1562 function getVariableValue( $index ) {
1563 global $wgContLang, $wgSitename, $wgServer;
1564
1565 switch ( $index ) {
1566 case MAG_CURRENTMONTH:
1567 return $wgContLang->formatNum( date( 'm' ) );
1568 case MAG_CURRENTMONTHNAME:
1569 return $wgContLang->getMonthName( date('n') );
1570 case MAG_CURRENTMONTHNAMEGEN:
1571 return $wgContLang->getMonthNameGen( date('n') );
1572 case MAG_CURRENTDAY:
1573 return $wgContLang->formatNum( date('j') );
1574 case MAG_PAGENAME:
1575 return $this->mTitle->getText();
1576 case MAG_PAGENAMEE:
1577 return $this->mTitle->getPartialURL();
1578 case MAG_NAMESPACE:
1579 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1580 return $wgContLang->getNsText($this->mTitle->getNamespace()); # Patch by Dori
1581 case MAG_CURRENTDAYNAME:
1582 return $wgContLang->getWeekdayName( date('w')+1 );
1583 case MAG_CURRENTYEAR:
1584 return $wgContLang->formatNum( date( 'Y' ) );
1585 case MAG_CURRENTTIME:
1586 return $wgContLang->time( wfTimestampNow(), false );
1587 case MAG_NUMBEROFARTICLES:
1588 return $wgContLang->formatNum( wfNumberOfArticles() );
1589 case MAG_SITENAME:
1590 return $wgSitename;
1591 case MAG_SERVER:
1592 return $wgServer;
1593 default:
1594 return NULL;
1595 }
1596 }
1597
1598 /**
1599 * initialise the magic variables (like CURRENTMONTHNAME)
1600 *
1601 * @access private
1602 */
1603 function initialiseVariables() {
1604 $fname = 'Parser::initialiseVariables';
1605 wfProfileIn( $fname );
1606 global $wgVariableIDs;
1607 $this->mVariables = array();
1608 foreach ( $wgVariableIDs as $id ) {
1609 $mw =& MagicWord::get( $id );
1610 $mw->addToArray( $this->mVariables, $this->getVariableValue( $id ) );
1611 }
1612 wfProfileOut( $fname );
1613 }
1614
1615 /**
1616 * Replace magic variables, templates, and template arguments
1617 * with the appropriate text. Templates are substituted recursively,
1618 * taking care to avoid infinite loops.
1619 *
1620 * Note that the substitution depends on value of $mOutputType:
1621 * OT_WIKI: only {{subst:}} templates
1622 * OT_MSG: only magic variables
1623 * OT_HTML: all templates and magic variables
1624 *
1625 * @param string $tex The text to transform
1626 * @param array $args Key-value pairs representing template parameters to substitute
1627 * @access private
1628 */
1629 function replaceVariables( $text, $args = array() ) {
1630 global $wgLang, $wgScript, $wgArticlePath;
1631
1632 # Prevent too big inclusions
1633 if(strlen($text)> MAX_INCLUDE_SIZE)
1634 return $text;
1635
1636 $fname = 'Parser::replaceVariables';
1637 wfProfileIn( $fname );
1638
1639 $titleChars = Title::legalChars();
1640
1641 # This function is called recursively. To keep track of arguments we need a stack:
1642 array_push( $this->mArgStack, $args );
1643
1644 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1645 $GLOBALS['wgCurParser'] =& $this;
1646
1647 # Variable substitution
1648 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1649
1650 if ( $this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI ) {
1651 # Argument substitution
1652 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1653 }
1654 # Template substitution
1655 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1656 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1657
1658 array_pop( $this->mArgStack );
1659
1660 wfProfileOut( $fname );
1661 return $text;
1662 }
1663
1664 /**
1665 * Replace magic variables
1666 * @access private
1667 */
1668 function variableSubstitution( $matches ) {
1669 if ( !$this->mVariables ) {
1670 $this->initialiseVariables();
1671 }
1672 $skip = false;
1673 if ( $this->mOutputType == OT_WIKI ) {
1674 # Do only magic variables prefixed by SUBST
1675 $mwSubst =& MagicWord::get( MAG_SUBST );
1676 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1677 $skip = true;
1678 # Note that if we don't substitute the variable below,
1679 # we don't remove the {{subst:}} magic word, in case
1680 # it is a template rather than a magic variable.
1681 }
1682 if ( !$skip && array_key_exists( $matches[1], $this->mVariables ) ) {
1683 $text = $this->mVariables[$matches[1]];
1684 $this->mOutput->mContainsOldMagic = true;
1685 } else {
1686 $text = $matches[0];
1687 }
1688 return $text;
1689 }
1690
1691 # Split template arguments
1692 function getTemplateArgs( $argsString ) {
1693 if ( $argsString === '' ) {
1694 return array();
1695 }
1696
1697 $args = explode( '|', substr( $argsString, 1 ) );
1698
1699 # If any of the arguments contains a '[[' but no ']]', it needs to be
1700 # merged with the next arg because the '|' character between belongs
1701 # to the link syntax and not the template parameter syntax.
1702 $argc = count($args);
1703 $i = 0;
1704 for ( $i = 0; $i < $argc-1; $i++ ) {
1705 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1706 $args[$i] .= '|'.$args[$i+1];
1707 array_splice($args, $i+1, 1);
1708 $i--;
1709 $argc--;
1710 }
1711 }
1712
1713 return $args;
1714 }
1715
1716 /**
1717 * Return the text of a template, after recursively
1718 * replacing any variables or templates within the template.
1719 *
1720 * @param array $matches The parts of the template
1721 * $matches[1]: the title, i.e. the part before the |
1722 * $matches[2]: the parameters (including a leading |), if any
1723 * @return string the text of the template
1724 * @access private
1725 */
1726 function braceSubstitution( $matches ) {
1727 global $wgLinkCache, $wgContLang;
1728 $fname = 'Parser::braceSubstitution';
1729 $found = false;
1730 $nowiki = false;
1731 $noparse = false;
1732
1733 $title = NULL;
1734
1735 # Need to know if the template comes at the start of a line,
1736 # to treat the beginning of the template like the beginning
1737 # of a line for tables and block-level elements.
1738 $linestart = $matches[1];
1739
1740 # $part1 is the bit before the first |, and must contain only title characters
1741 # $args is a list of arguments, starting from index 0, not including $part1
1742
1743 $part1 = $matches[2];
1744 # If the third subpattern matched anything, it will start with |
1745
1746 $args = $this->getTemplateArgs($matches[3]);
1747 $argc = count( $args );
1748
1749 # Don't parse {{{}}} because that's only for template arguments
1750 if ( $linestart === '{' ) {
1751 $text = $matches[0];
1752 $found = true;
1753 $noparse = true;
1754 }
1755
1756 # SUBST
1757 if ( !$found ) {
1758 $mwSubst =& MagicWord::get( MAG_SUBST );
1759 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType == OT_WIKI) ) {
1760 # One of two possibilities is true:
1761 # 1) Found SUBST but not in the PST phase
1762 # 2) Didn't find SUBST and in the PST phase
1763 # In either case, return without further processing
1764 $text = $matches[0];
1765 $found = true;
1766 $noparse = true;
1767 }
1768 }
1769
1770 # MSG, MSGNW and INT
1771 if ( !$found ) {
1772 # Check for MSGNW:
1773 $mwMsgnw =& MagicWord::get( MAG_MSGNW );
1774 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1775 $nowiki = true;
1776 } else {
1777 # Remove obsolete MSG:
1778 $mwMsg =& MagicWord::get( MAG_MSG );
1779 $mwMsg->matchStartAndRemove( $part1 );
1780 }
1781
1782 # Check if it is an internal message
1783 $mwInt =& MagicWord::get( MAG_INT );
1784 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1785 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1786 $text = $linestart . wfMsgReal( $part1, $args, true );
1787 $found = true;
1788 }
1789 }
1790 }
1791
1792 # NS
1793 if ( !$found ) {
1794 # Check for NS: (namespace expansion)
1795 $mwNs = MagicWord::get( MAG_NS );
1796 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1797 if ( intval( $part1 ) ) {
1798 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1799 $found = true;
1800 } else {
1801 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1802 if ( !is_null( $index ) ) {
1803 $text = $linestart . $wgContLang->getNsText( $index );
1804 $found = true;
1805 }
1806 }
1807 }
1808 }
1809
1810 # LOCALURL and LOCALURLE
1811 if ( !$found ) {
1812 $mwLocal = MagicWord::get( MAG_LOCALURL );
1813 $mwLocalE = MagicWord::get( MAG_LOCALURLE );
1814
1815 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1816 $func = 'getLocalURL';
1817 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1818 $func = 'escapeLocalURL';
1819 } else {
1820 $func = '';
1821 }
1822
1823 if ( $func !== '' ) {
1824 $title = Title::newFromText( $part1 );
1825 if ( !is_null( $title ) ) {
1826 if ( $argc > 0 ) {
1827 $text = $linestart . $title->$func( $args[0] );
1828 } else {
1829 $text = $linestart . $title->$func();
1830 }
1831 $found = true;
1832 }
1833 }
1834 }
1835
1836 # GRAMMAR
1837 if ( !$found && $argc == 1 ) {
1838 $mwGrammar =& MagicWord::get( MAG_GRAMMAR );
1839 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1840 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1841 $found = true;
1842 }
1843 }
1844
1845 # Template table test
1846
1847 # Did we encounter this template already? If yes, it is in the cache
1848 # and we need to check for loops.
1849 if ( !$found && isset( $this->mTemplates[$part1] ) ) {
1850 # set $text to cached message.
1851 $text = $linestart . $this->mTemplates[$part1];
1852 $found = true;
1853
1854 # Infinite loop test
1855 if ( isset( $this->mTemplatePath[$part1] ) ) {
1856 $noparse = true;
1857 $found = true;
1858 $text .= '<!-- WARNING: template loop detected -->';
1859 }
1860 }
1861
1862 # Load from database
1863 $itcamefromthedatabase = false;
1864 if ( !$found ) {
1865 $ns = NS_TEMPLATE;
1866 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
1867 if ($subpage !== '') {
1868 $ns = $this->mTitle->getNamespace();
1869 }
1870 $title = Title::newFromText( $part1, $ns );
1871 if ( !is_null( $title ) && !$title->isExternal() ) {
1872 # Check for excessive inclusion
1873 $dbk = $title->getPrefixedDBkey();
1874 if ( $this->incrementIncludeCount( $dbk ) ) {
1875 # This should never be reached.
1876 $article = new Article( $title );
1877 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1878 if ( $articleContent !== false ) {
1879 $found = true;
1880 $text = $linestart . $articleContent;
1881 $itcamefromthedatabase = true;
1882 }
1883 }
1884
1885 # If the title is valid but undisplayable, make a link to it
1886 if ( $this->mOutputType == OT_HTML && !$found ) {
1887 $text = $linestart . '[['.$title->getPrefixedText().']]';
1888 $found = true;
1889 }
1890
1891 # Template cache array insertion
1892 $this->mTemplates[$part1] = $text;
1893 }
1894 }
1895
1896 # Recursive parsing, escaping and link table handling
1897 # Only for HTML output
1898 if ( $nowiki && $found && $this->mOutputType == OT_HTML ) {
1899 $text = wfEscapeWikiText( $text );
1900 } elseif ( ($this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI) && $found && !$noparse) {
1901 # Clean up argument array
1902 $assocArgs = array();
1903 $index = 1;
1904 foreach( $args as $arg ) {
1905 $eqpos = strpos( $arg, '=' );
1906 if ( $eqpos === false ) {
1907 $assocArgs[$index++] = $arg;
1908 } else {
1909 $name = trim( substr( $arg, 0, $eqpos ) );
1910 $value = trim( substr( $arg, $eqpos+1 ) );
1911 if ( $value === false ) {
1912 $value = '';
1913 }
1914 if ( $name !== false ) {
1915 $assocArgs[$name] = $value;
1916 }
1917 }
1918 }
1919
1920 # Add a new element to the templace recursion path
1921 $this->mTemplatePath[$part1] = 1;
1922
1923 $text = $this->strip( $text, $this->mStripState );
1924 $text = $this->removeHTMLtags( $text );
1925 $text = $this->replaceVariables( $text, $assocArgs );
1926
1927 # Resume the link cache and register the inclusion as a link
1928 if ( $this->mOutputType == OT_HTML && !is_null( $title ) ) {
1929 $wgLinkCache->addLinkObj( $title );
1930 }
1931
1932 # If the template begins with a table or block-level
1933 # element, it should be treated as beginning a new line.
1934 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
1935 $text = "\n" . $text;
1936 }
1937 }
1938
1939 # Empties the template path
1940 $this->mTemplatePath = array();
1941 if ( !$found ) {
1942 return $matches[0];
1943 } else {
1944 # replace ==section headers==
1945 # XXX this needs to go away once we have a better parser.
1946 if ( $this->mOutputType != OT_WIKI && $itcamefromthedatabase ) {
1947 if( !is_null( $title ) )
1948 $encodedname = base64_encode($title->getPrefixedDBkey());
1949 else
1950 $encodedname = base64_encode("");
1951 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
1952 PREG_SPLIT_DELIM_CAPTURE);
1953 $text = '';
1954 $nsec = 0;
1955 for( $i = 0; $i < count($m); $i += 2 ) {
1956 $text .= $m[$i];
1957 if (!isset($m[$i + 1]) || $m[$i + 1] == "") continue;
1958 $hl = $m[$i + 1];
1959 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
1960 $text .= $hl;
1961 continue;
1962 }
1963 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
1964 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
1965 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
1966
1967 $nsec++;
1968 }
1969 }
1970 }
1971
1972 # Empties the template path
1973 $this->mTemplatePath = array();
1974 if ( !$found ) {
1975 return $matches[0];
1976 } else {
1977 return $text;
1978 }
1979 }
1980
1981 /**
1982 * Triple brace replacement -- used for template arguments
1983 * @access private
1984 */
1985 function argSubstitution( $matches ) {
1986 $arg = trim( $matches[1] );
1987 $text = $matches[0];
1988 $inputArgs = end( $this->mArgStack );
1989
1990 if ( array_key_exists( $arg, $inputArgs ) ) {
1991 $text = $inputArgs[$arg];
1992 }
1993
1994 return $text;
1995 }
1996
1997 /**
1998 * Returns true if the function is allowed to include this entity
1999 * @access private
2000 */
2001 function incrementIncludeCount( $dbk ) {
2002 if ( !array_key_exists( $dbk, $this->mIncludeCount ) ) {
2003 $this->mIncludeCount[$dbk] = 0;
2004 }
2005 if ( ++$this->mIncludeCount[$dbk] <= MAX_INCLUDE_REPEAT ) {
2006 return true;
2007 } else {
2008 return false;
2009 }
2010 }
2011
2012
2013 /**
2014 * Cleans up HTML, removes dangerous tags and attributes, and
2015 * removes HTML comments
2016 * @access private
2017 */
2018 function removeHTMLtags( $text ) {
2019 global $wgUseTidy, $wgUserHtml;
2020 $fname = 'Parser::removeHTMLtags';
2021 wfProfileIn( $fname );
2022
2023 if( $wgUserHtml ) {
2024 $htmlpairs = array( # Tags that must be closed
2025 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2026 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2027 'strike', 'strong', 'tt', 'var', 'div', 'center',
2028 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2029 'ruby', 'rt' , 'rb' , 'rp', 'p'
2030 );
2031 $htmlsingle = array(
2032 'br', 'hr', 'li', 'dt', 'dd'
2033 );
2034 $htmlnest = array( # Tags that can be nested--??
2035 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2036 'dl', 'font', 'big', 'small', 'sub', 'sup'
2037 );
2038 $tabletags = array( # Can only appear inside table
2039 'td', 'th', 'tr'
2040 );
2041 } else {
2042 $htmlpairs = array();
2043 $htmlsingle = array();
2044 $htmlnest = array();
2045 $tabletags = array();
2046 }
2047
2048 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2049 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2050
2051 $htmlattrs = $this->getHTMLattrs () ;
2052
2053 # Remove HTML comments
2054 $text = $this->removeHTMLcomments( $text );
2055
2056 $bits = explode( '<', $text );
2057 $text = array_shift( $bits );
2058 if(!$wgUseTidy) {
2059 $tagstack = array(); $tablestack = array();
2060 foreach ( $bits as $x ) {
2061 $prev = error_reporting( E_ALL & ~( E_NOTICE | E_WARNING ) );
2062 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2063 $x, $regs );
2064 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2065 error_reporting( $prev );
2066
2067 $badtag = 0 ;
2068 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2069 # Check our stack
2070 if ( $slash ) {
2071 # Closing a tag...
2072 if ( ! in_array( $t, $htmlsingle ) &&
2073 ( $ot = @array_pop( $tagstack ) ) != $t ) {
2074 @array_push( $tagstack, $ot );
2075 $badtag = 1;
2076 } else {
2077 if ( $t == 'table' ) {
2078 $tagstack = array_pop( $tablestack );
2079 }
2080 $newparams = '';
2081 }
2082 } else {
2083 # Keep track for later
2084 if ( in_array( $t, $tabletags ) &&
2085 ! in_array( 'table', $tagstack ) ) {
2086 $badtag = 1;
2087 } else if ( in_array( $t, $tagstack ) &&
2088 ! in_array ( $t , $htmlnest ) ) {
2089 $badtag = 1 ;
2090 } else if ( ! in_array( $t, $htmlsingle ) ) {
2091 if ( $t == 'table' ) {
2092 array_push( $tablestack, $tagstack );
2093 $tagstack = array();
2094 }
2095 array_push( $tagstack, $t );
2096 }
2097 # Strip non-approved attributes from the tag
2098 $newparams = $this->fixTagAttributes($params);
2099
2100 }
2101 if ( ! $badtag ) {
2102 $rest = str_replace( '>', '&gt;', $rest );
2103 $text .= "<$slash$t $newparams$brace$rest";
2104 continue;
2105 }
2106 }
2107 $text .= '&lt;' . str_replace( '>', '&gt;', $x);
2108 }
2109 # Close off any remaining tags
2110 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2111 $text .= "</$t>\n";
2112 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2113 }
2114 } else {
2115 # this might be possible using tidy itself
2116 foreach ( $bits as $x ) {
2117 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2118 $x, $regs );
2119 @list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2120 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2121 $newparams = $this->fixTagAttributes($params);
2122 $rest = str_replace( '>', '&gt;', $rest );
2123 $text .= "<$slash$t $newparams$brace$rest";
2124 } else {
2125 $text .= '&lt;' . str_replace( '>', '&gt;', $x);
2126 }
2127 }
2128 }
2129 wfProfileOut( $fname );
2130 return $text;
2131 }
2132
2133 /**
2134 * Remove '<!--', '-->', and everything between.
2135 * To avoid leaving blank lines, when a comment is both preceded
2136 * and followed by a newline (ignoring spaces), trim leading and
2137 * trailing spaces and one of the newlines.
2138 *
2139 * @access private
2140 */
2141 function removeHTMLcomments( $text ) {
2142 $fname='Parser::removeHTMLcomments';
2143 wfProfileIn( $fname );
2144 while (($start = strpos($text, '<!--')) !== false) {
2145 $end = strpos($text, '-->', $start + 4);
2146 if ($end === false) {
2147 # Unterminated comment; bail out
2148 break;
2149 }
2150
2151 $end += 3;
2152
2153 # Trim space and newline if the comment is both
2154 # preceded and followed by a newline
2155 $spaceStart = max($start - 1, 0);
2156 $spaceLen = $end - $spaceStart;
2157 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2158 $spaceStart--;
2159 $spaceLen++;
2160 }
2161 while (substr($text, $spaceStart + $spaceLen, 1) === ' ')
2162 $spaceLen++;
2163 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart + $spaceLen, 1) === "\n") {
2164 # Remove the comment, leading and trailing
2165 # spaces, and leave only one newline.
2166 $text = substr_replace($text, "\n", $spaceStart, $spaceLen + 1);
2167 }
2168 else {
2169 # Remove just the comment.
2170 $text = substr_replace($text, '', $start, $end - $start);
2171 }
2172 }
2173 wfProfileOut( $fname );
2174 return $text;
2175 }
2176
2177 /**
2178 * This function accomplishes several tasks:
2179 * 1) Auto-number headings if that option is enabled
2180 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2181 * 3) Add a Table of contents on the top for users who have enabled the option
2182 * 4) Auto-anchor headings
2183 *
2184 * It loops through all headlines, collects the necessary data, then splits up the
2185 * string and re-inserts the newly formatted headlines.
2186 * @access private
2187 */
2188 /* private */ function formatHeadings( $text, $isMain=true ) {
2189 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2190
2191 $doNumberHeadings = $this->mOptions->getNumberHeadings();
2192 $doShowToc = $this->mOptions->getShowToc();
2193 $forceTocHere = false;
2194 if( !$this->mTitle->userCanEdit() ) {
2195 $showEditLink = 0;
2196 $rightClickHack = 0;
2197 } else {
2198 $showEditLink = $this->mOptions->getEditSection();
2199 $rightClickHack = $this->mOptions->getEditSectionOnRightClick();
2200 }
2201
2202 # Inhibit editsection links if requested in the page
2203 $esw =& MagicWord::get( MAG_NOEDITSECTION );
2204 if( $esw->matchAndRemove( $text ) ) {
2205 $showEditLink = 0;
2206 }
2207 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2208 # do not add TOC
2209 $mw =& MagicWord::get( MAG_NOTOC );
2210 if( $mw->matchAndRemove( $text ) ) {
2211 $doShowToc = 0;
2212 }
2213
2214 # never add the TOC to the Main Page. This is an entry page that should not
2215 # be more than 1-2 screens large anyway
2216 if( $this->mTitle->getPrefixedText() == wfMsg('mainpage') ) {
2217 $doShowToc = 0;
2218 }
2219
2220 # Get all headlines for numbering them and adding funky stuff like [edit]
2221 # links - this is for later, but we need the number of headlines right now
2222 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2223
2224 # if there are fewer than 4 headlines in the article, do not show TOC
2225 if( $numMatches < 4 ) {
2226 $doShowToc = 0;
2227 }
2228
2229 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2230 # override above conditions and always show TOC at that place
2231 $mw =& MagicWord::get( MAG_TOC );
2232 if ($mw->match( $text ) ) {
2233 $doShowToc = 1;
2234 $forceTocHere = true;
2235 } else {
2236 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2237 # override above conditions and always show TOC above first header
2238 $mw =& MagicWord::get( MAG_FORCETOC );
2239 if ($mw->matchAndRemove( $text ) ) {
2240 $doShowToc = 1;
2241 }
2242 }
2243
2244
2245
2246 # We need this to perform operations on the HTML
2247 $sk =& $this->mOptions->getSkin();
2248
2249 # headline counter
2250 $headlineCount = 0;
2251 $sectionCount = 0; # headlineCount excluding template sections
2252
2253 # Ugh .. the TOC should have neat indentation levels which can be
2254 # passed to the skin functions. These are determined here
2255 $toclevel = 0;
2256 $toc = '';
2257 $full = '';
2258 $head = array();
2259 $sublevelCount = array();
2260 $level = 0;
2261 $prevlevel = 0;
2262 foreach( $matches[3] as $headline ) {
2263 $istemplate = 0;
2264 $templatetitle = "";
2265 $templatesection = 0;
2266
2267 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2268 $istemplate = 1;
2269 $templatetitle = base64_decode($mat[1]);
2270 $templatesection = 1 + (int)base64_decode($mat[2]);
2271 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2272 }
2273
2274 $numbering = '';
2275 if( $level ) {
2276 $prevlevel = $level;
2277 }
2278 $level = $matches[1][$headlineCount];
2279 if( ( $doNumberHeadings || $doShowToc ) && $prevlevel && $level > $prevlevel ) {
2280 # reset when we enter a new level
2281 $sublevelCount[$level] = 0;
2282 $toc .= $sk->tocIndent( $level - $prevlevel );
2283 $toclevel += $level - $prevlevel;
2284 }
2285 if( ( $doNumberHeadings || $doShowToc ) && $level < $prevlevel ) {
2286 # reset when we step back a level
2287 $sublevelCount[$level+1]=0;
2288 $toc .= $sk->tocUnindent( $prevlevel - $level );
2289 $toclevel -= $prevlevel - $level;
2290 }
2291 # count number of headlines for each level
2292 @$sublevelCount[$level]++;
2293 if( $doNumberHeadings || $doShowToc ) {
2294 $dot = 0;
2295 for( $i = 1; $i <= $level; $i++ ) {
2296 if( !empty( $sublevelCount[$i] ) ) {
2297 if( $dot ) {
2298 $numbering .= '.';
2299 }
2300 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2301 $dot = 1;
2302 }
2303 }
2304 }
2305
2306 # The canonized header is a version of the header text safe to use for links
2307 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2308 $canonized_headline = $this->unstrip( $headline, $this->mStripState );
2309 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState );
2310
2311 # Remove link placeholders by the link text.
2312 # <!--LINK number-->
2313 # turns into
2314 # link text with suffix
2315 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2316 "\$wgLinkHolders['texts'][\$1]",
2317 $canonized_headline );
2318
2319 # strip out HTML
2320 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2321 $tocline = trim( $canonized_headline );
2322 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT, $wgInputEncoding ) );
2323 $replacearray = array(
2324 '%3A' => ':',
2325 '%' => '.'
2326 );
2327 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2328 $refer[$headlineCount] = $canonized_headline;
2329
2330 # count how many in assoc. array so we can track dupes in anchors
2331 @$refers[$canonized_headline]++;
2332 $refcount[$headlineCount]=$refers[$canonized_headline];
2333
2334 # Prepend the number to the heading text
2335
2336 if( $doNumberHeadings || $doShowToc ) {
2337 $tocline = $numbering . ' ' . $tocline;
2338
2339 # Don't number the heading if it is the only one (looks silly)
2340 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2341 # the two are different if the line contains a link
2342 $headline=$numbering . ' ' . $headline;
2343 }
2344 }
2345
2346 # Create the anchor for linking from the TOC to the section
2347 $anchor = $canonized_headline;
2348 if($refcount[$headlineCount] > 1 ) {
2349 $anchor .= '_' . $refcount[$headlineCount];
2350 }
2351 if( $doShowToc && ( !isset($wgMaxTocLevel) || $toclevel<$wgMaxTocLevel ) ) {
2352 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2353 }
2354 if( $showEditLink && ( !$istemplate || $templatetitle !== "" ) ) {
2355 if ( empty( $head[$headlineCount] ) ) {
2356 $head[$headlineCount] = '';
2357 }
2358 if( $istemplate )
2359 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2360 else
2361 $head[$headlineCount] .= $sk->editSectionLink($sectionCount+1);
2362 }
2363
2364 # Add the edit section span
2365 if( $rightClickHack ) {
2366 if( $istemplate )
2367 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2368 else
2369 $headline = $sk->editSectionScript($sectionCount+1,$headline);
2370 }
2371
2372 # give headline the correct <h#> tag
2373 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2374
2375 $headlineCount++;
2376 if( !$istemplate )
2377 $sectionCount++;
2378 }
2379
2380 if( $doShowToc ) {
2381 $toclines = $headlineCount;
2382 $toc .= $sk->tocUnindent( $toclevel );
2383 $toc = $sk->tocTable( $toc );
2384 }
2385
2386 # split up and insert constructed headlines
2387
2388 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2389 $i = 0;
2390
2391 foreach( $blocks as $block ) {
2392 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2393 # This is the [edit] link that appears for the top block of text when
2394 # section editing is enabled
2395
2396 # Disabled because it broke block formatting
2397 # For example, a bullet point in the top line
2398 # $full .= $sk->editSectionLink(0);
2399 }
2400 $full .= $block;
2401 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2402 # Top anchor now in skin
2403 $full = $full.$toc;
2404 }
2405
2406 if( !empty( $head[$i] ) ) {
2407 $full .= $head[$i];
2408 }
2409 $i++;
2410 }
2411 if($forceTocHere) {
2412 $mw =& MagicWord::get( MAG_TOC );
2413 return $mw->replace( $toc, $full );
2414 } else {
2415 return $full;
2416 }
2417 }
2418
2419 /**
2420 * Return an HTML link for the "ISBN 123456" text
2421 * @access private
2422 */
2423 function magicISBN( $text ) {
2424 global $wgLang;
2425 $fname = 'Parser::magicISBN';
2426 wfProfileIn( $fname );
2427
2428 $a = split( 'ISBN ', ' '.$text );
2429 if ( count ( $a ) < 2 ) {
2430 wfProfileOut( $fname );
2431 return $text;
2432 }
2433 $text = substr( array_shift( $a ), 1);
2434 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2435
2436 foreach ( $a as $x ) {
2437 $isbn = $blank = '' ;
2438 while ( ' ' == $x{0} ) {
2439 $blank .= ' ';
2440 $x = substr( $x, 1 );
2441 }
2442 if ( $x == '' ) { # blank isbn
2443 $text .= "ISBN $blank";
2444 continue;
2445 }
2446 while ( strstr( $valid, $x{0} ) != false ) {
2447 $isbn .= $x{0};
2448 $x = substr( $x, 1 );
2449 }
2450 $num = str_replace( '-', '', $isbn );
2451 $num = str_replace( ' ', '', $num );
2452
2453 if ( '' == $num ) {
2454 $text .= "ISBN $blank$x";
2455 } else {
2456 $titleObj = Title::makeTitle( NS_SPECIAL, 'Booksources' );
2457 $text .= '<a href="' .
2458 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2459 "\" class=\"internal\">ISBN $isbn</a>";
2460 $text .= $x;
2461 }
2462 }
2463 wfProfileOut( $fname );
2464 return $text;
2465 }
2466
2467 /**
2468 * Return an HTML link for the "GEO ..." text
2469 * @access private
2470 */
2471 function magicGEO( $text ) {
2472 global $wgLang, $wgUseGeoMode;
2473 $fname = 'Parser::magicGEO';
2474 wfProfileIn( $fname );
2475
2476 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2477 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2478 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['N']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2479 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['N']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2480 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['S']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2481 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['S']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2482
2483 $a = split( 'GEO ', ' '.$text );
2484 if ( count ( $a ) < 2 ) {
2485 wfProfileOut( $fname );
2486 return $text;
2487 }
2488 $text = substr( array_shift( $a ), 1);
2489 $valid = '0123456789.+-:';
2490
2491 foreach ( $a as $x ) {
2492 $geo = $blank = '' ;
2493 while ( ' ' == $x{0} ) {
2494 $blank .= ' ';
2495 $x = substr( $x, 1 );
2496 }
2497 while ( strstr( $valid, $x{0} ) != false ) {
2498 $geo .= $x{0};
2499 $x = substr( $x, 1 );
2500 }
2501 $num = str_replace( '+', '', $geo );
2502 $num = str_replace( ' ', '', $num );
2503
2504 if ( '' == $num || count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2505 $text .= "GEO $blank$x";
2506 } else {
2507 $titleObj = Title::makeTitle( NS_SPECIAL, 'Geo' );
2508 $text .= '<a href="' .
2509 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2510 "\" class=\"internal\">GEO $geo</a>";
2511 $text .= $x;
2512 }
2513 }
2514 wfProfileOut( $fname );
2515 return $text;
2516 }
2517
2518 /**
2519 * Return an HTML link for the "RFC 1234" text
2520 * @access private
2521 * @param string $text text to be processed
2522 */
2523 function magicRFC( $text ) {
2524 global $wgLang;
2525
2526 $valid = '0123456789';
2527 $internal = false;
2528
2529 $a = split( 'RFC ', ' '.$text );
2530 if ( count ( $a ) < 2 ) return $text;
2531 $text = substr( array_shift( $a ), 1);
2532
2533 /* Check if RFC keyword is preceed by [[.
2534 * This test is made here cause of the array_shift above
2535 * that prevent the test to be done in the foreach.
2536 */
2537 if(substr($text, -2) == '[[') { $internal = true; }
2538
2539 foreach ( $a as $x ) {
2540 /* token might be empty if we have RFC RFC 1234 */
2541 if($x=='') {
2542 $text.='RFC ';
2543 continue;
2544 }
2545
2546 $rfc = $blank = '' ;
2547
2548 /** remove and save whitespaces in $blank */
2549 while ( $x{0} == ' ' ) {
2550 $blank .= ' ';
2551 $x = substr( $x, 1 );
2552 }
2553
2554 /** remove and save the rfc number in $rfc */
2555 while ( strstr( $valid, $x{0} ) != false ) {
2556 $rfc .= $x{0};
2557 $x = substr( $x, 1 );
2558 }
2559
2560 if ( $rfc == '') {
2561 /* call back stripped spaces*/
2562 $text .= "RFC $blank$x";
2563 } elseif( $internal) {
2564 /* normal link */
2565 $text .= "RFC $rfc$x";
2566 } else {
2567 /* build the external link*/
2568 $url = wfmsg( 'rfcurl' );
2569 $url = str_replace( '$1', $rfc, $url);
2570 $sk =& $this->mOptions->getSkin();
2571 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2572 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2573 }
2574
2575 /* Check if the next RFC keyword is preceed by [[ */
2576 $internal = (substr($x,-2) == '[[');
2577 }
2578 return $text;
2579 }
2580
2581 /**
2582 * Transform wiki markup when saving a page by doing \r\n -> \n
2583 * conversion, substitting signatures, {{subst:}} templates, etc.
2584 *
2585 * @param string $text the text to transform
2586 * @param Title &$title the Title object for the current article
2587 * @param User &$user the User object describing the current user
2588 * @param ParserOptions $options parsing options
2589 * @param bool $clearState whether to clear the parser state first
2590 * @return string the altered wiki markup
2591 * @access public
2592 */
2593 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2594 $this->mOptions = $options;
2595 $this->mTitle =& $title;
2596 $this->mOutputType = OT_WIKI;
2597
2598 if ( $clearState ) {
2599 $this->clearState();
2600 }
2601
2602 $stripState = false;
2603 $pairs = array(
2604 "\r\n" => "\n",
2605 );
2606 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2607 // now with regexes
2608 /*
2609 $pairs = array(
2610 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2611 "/<br *?>/i" => "<br />",
2612 );
2613 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2614 */
2615 $text = $this->strip( $text, $stripState, false );
2616 $text = $this->pstPass2( $text, $user );
2617 $text = $this->unstrip( $text, $stripState );
2618 $text = $this->unstripNoWiki( $text, $stripState );
2619 return $text;
2620 }
2621
2622 /**
2623 * Pre-save transform helper function
2624 * @access private
2625 */
2626 function pstPass2( $text, &$user ) {
2627 global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
2628
2629 # Variable replacement
2630 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2631 $text = $this->replaceVariables( $text );
2632
2633 # Signatures
2634 #
2635 $n = $user->getName();
2636 $k = $user->getOption( 'nickname' );
2637 if ( '' == $k ) { $k = $n; }
2638 if(isset($wgLocaltimezone)) {
2639 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2640 }
2641 /* Note: this is an ugly timezone hack for the European wikis */
2642 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2643 ' (' . date( 'T' ) . ')';
2644 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2645
2646 $text = preg_replace( '/~~~~~/', $d, $text );
2647 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER ) . ":$n|$k]] $d", $text );
2648 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER ) . ":$n|$k]]", $text );
2649
2650 # Context links: [[|name]] and [[name (context)|]]
2651 #
2652 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2653 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2654 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2655 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2656
2657 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2658 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2659 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2660 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2661 $context = '';
2662 $t = $this->mTitle->getText();
2663 if ( preg_match( $conpat, $t, $m ) ) {
2664 $context = $m[2];
2665 }
2666 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2667 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2668 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2669
2670 if ( '' == $context ) {
2671 $text = preg_replace( $p2, '[[\\1]]', $text );
2672 } else {
2673 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2674 }
2675
2676 # Trim trailing whitespace
2677 # MAG_END (__END__) tag allows for trailing
2678 # whitespace to be deliberately included
2679 $text = rtrim( $text );
2680 $mw =& MagicWord::get( MAG_END );
2681 $mw->matchAndRemove( $text );
2682
2683 return $text;
2684 }
2685
2686 /**
2687 * Set up some variables which are usually set up in parse()
2688 * so that an external function can call some class members with confidence
2689 * @access public
2690 */
2691 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2692 $this->mTitle =& $title;
2693 $this->mOptions = $options;
2694 $this->mOutputType = $outputType;
2695 if ( $clearState ) {
2696 $this->clearState();
2697 }
2698 }
2699
2700 /**
2701 * Transform a MediaWiki message by replacing magic variables.
2702 *
2703 * @param string $text the text to transform
2704 * @param ParserOptions $options options
2705 * @return string the text with variables substituted
2706 * @access public
2707 */
2708 function transformMsg( $text, $options ) {
2709 global $wgTitle;
2710 static $executing = false;
2711
2712 # Guard against infinite recursion
2713 if ( $executing ) {
2714 return $text;
2715 }
2716 $executing = true;
2717
2718 $this->mTitle = $wgTitle;
2719 $this->mOptions = $options;
2720 $this->mOutputType = OT_MSG;
2721 $this->clearState();
2722 $text = $this->replaceVariables( $text );
2723
2724 $executing = false;
2725 return $text;
2726 }
2727
2728 /**
2729 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2730 * Callback will be called with the text within
2731 * Transform and return the text within
2732 * @access public
2733 */
2734 function setHook( $tag, $callback ) {
2735 $oldVal = @$this->mTagHooks[$tag];
2736 $this->mTagHooks[$tag] = $callback;
2737 return $oldVal;
2738 }
2739 }
2740
2741 /**
2742 * @todo document
2743 * @package MediaWiki
2744 */
2745 class ParserOutput
2746 {
2747 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2748 var $mCacheTime; # Used in ParserCache
2749
2750 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2751 $containsOldMagic = false )
2752 {
2753 $this->mText = $text;
2754 $this->mLanguageLinks = $languageLinks;
2755 $this->mCategoryLinks = $categoryLinks;
2756 $this->mContainsOldMagic = $containsOldMagic;
2757 $this->mCacheTime = '';
2758 }
2759
2760 function getText() { return $this->mText; }
2761 function getLanguageLinks() { return $this->mLanguageLinks; }
2762 function getCategoryLinks() { return $this->mCategoryLinks; }
2763 function getCacheTime() { return $this->mCacheTime; }
2764 function containsOldMagic() { return $this->mContainsOldMagic; }
2765 function setText( $text ) { return wfSetVar( $this->mText, $text ); }
2766 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks, $ll ); }
2767 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks, $cl ); }
2768 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
2769 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime, $t ); }
2770
2771 function merge( $other ) {
2772 $this->mLanguageLinks = array_merge( $this->mLanguageLinks, $other->mLanguageLinks );
2773 $this->mCategoryLinks = array_merge( $this->mCategoryLinks, $this->mLanguageLinks );
2774 $this->mContainsOldMagic = $this->mContainsOldMagic || $other->mContainsOldMagic;
2775 }
2776
2777 }
2778
2779 /**
2780 * Set options of the Parser
2781 * @todo document
2782 * @package MediaWiki
2783 */
2784 class ParserOptions
2785 {
2786 # All variables are private
2787 var $mUseTeX; # Use texvc to expand <math> tags
2788 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2789 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2790 var $mAllowExternalImages; # Allow external images inline
2791 var $mSkin; # Reference to the preferred skin
2792 var $mDateFormat; # Date format index
2793 var $mEditSection; # Create "edit section" links
2794 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2795 var $mNumberHeadings; # Automatically number headings
2796 var $mShowToc; # Show table of contents
2797
2798 function getUseTeX() { return $this->mUseTeX; }
2799 function getUseDynamicDates() { return $this->mUseDynamicDates; }
2800 function getInterwikiMagic() { return $this->mInterwikiMagic; }
2801 function getAllowExternalImages() { return $this->mAllowExternalImages; }
2802 function getSkin() { return $this->mSkin; }
2803 function getDateFormat() { return $this->mDateFormat; }
2804 function getEditSection() { return $this->mEditSection; }
2805 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick; }
2806 function getNumberHeadings() { return $this->mNumberHeadings; }
2807 function getShowToc() { return $this->mShowToc; }
2808
2809 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX, $x ); }
2810 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates, $x ); }
2811 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic, $x ); }
2812 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages, $x ); }
2813 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat, $x ); }
2814 function setEditSection( $x ) { return wfSetVar( $this->mEditSection, $x ); }
2815 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick, $x ); }
2816 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings, $x ); }
2817 function setShowToc( $x ) { return wfSetVar( $this->mShowToc, $x ); }
2818
2819 function setSkin( &$x ) { $this->mSkin =& $x; }
2820
2821 # Get parser options
2822 /* static */ function newFromUser( &$user ) {
2823 $popts = new ParserOptions;
2824 $popts->initialiseFromUser( $user );
2825 return $popts;
2826 }
2827
2828 # Get user options
2829 function initialiseFromUser( &$userInput ) {
2830 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2831
2832 $fname = 'ParserOptions::initialiseFromUser';
2833 wfProfileIn( $fname );
2834 if ( !$userInput ) {
2835 $user = new User;
2836 $user->setLoaded( true );
2837 } else {
2838 $user =& $userInput;
2839 }
2840
2841 $this->mUseTeX = $wgUseTeX;
2842 $this->mUseDynamicDates = $wgUseDynamicDates;
2843 $this->mInterwikiMagic = $wgInterwikiMagic;
2844 $this->mAllowExternalImages = $wgAllowExternalImages;
2845 wfProfileIn( $fname.'-skin' );
2846 $this->mSkin =& $user->getSkin();
2847 wfProfileOut( $fname.'-skin' );
2848 $this->mDateFormat = $user->getOption( 'date' );
2849 $this->mEditSection = $user->getOption( 'editsection' );
2850 $this->mEditSectionOnRightClick = $user->getOption( 'editsectiononrightclick' );
2851 $this->mNumberHeadings = $user->getOption( 'numberheadings' );
2852 $this->mShowToc = $user->getOption( 'showtoc' );
2853 wfProfileOut( $fname );
2854 }
2855
2856
2857 }
2858
2859 # Regex callbacks, used in Parser::replaceVariables
2860 function wfBraceSubstitution( $matches ) {
2861 global $wgCurParser;
2862 return $wgCurParser->braceSubstitution( $matches );
2863 }
2864
2865 function wfArgSubstitution( $matches ) {
2866 global $wgCurParser;
2867 return $wgCurParser->argSubstitution( $matches );
2868 }
2869
2870 function wfVariableSubstitution( $matches ) {
2871 global $wgCurParser;
2872 return $wgCurParser->variableSubstitution( $matches );
2873 }
2874
2875 /**
2876 * Return the total number of articles
2877 */
2878 function wfNumberOfArticles() {
2879 global $wgNumberOfArticles;
2880
2881 wfLoadSiteStats();
2882 return $wgNumberOfArticles;
2883 }
2884
2885 /**
2886 * Get various statistics from the database
2887 * @private
2888 */
2889 function wfLoadSiteStats() {
2890 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2891 $fname = 'wfLoadSiteStats';
2892
2893 if ( -1 != $wgNumberOfArticles ) return;
2894 $dbr =& wfGetDB( DB_SLAVE );
2895 $s = $dbr->getArray( 'site_stats',
2896 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2897 array( 'ss_row_id' => 1 ), $fname
2898 );
2899
2900 if ( $s === false ) {
2901 return;
2902 } else {
2903 $wgTotalViews = $s->ss_total_views;
2904 $wgTotalEdits = $s->ss_total_edits;
2905 $wgNumberOfArticles = $s->ss_good_articles;
2906 }
2907 }
2908
2909 function wfEscapeHTMLTagsOnly( $in ) {
2910 return str_replace(
2911 array( '"', '>', '<' ),
2912 array( '&quot;', '&gt;', '&lt;' ),
2913 $in );
2914 }
2915
2916 ?>
MediaWiki Core