2 rcs_id('$Id: InlineParser.php,v 1.62 2005-02-05 15:35:37 rurban Exp $');
3 /* Copyright (C) 2002 Geoffrey T. Dairiki <dairiki@dairiki.org>
4 * Copyright (C) 2004 Reini Urban
6 * This file is part of PhpWiki.
8 * PhpWiki is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation; either version 2 of the License, or
11 * (at your option) any later version.
13 * PhpWiki is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with PhpWiki; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 * This is the code which deals with the inline part of the (new-style)
27 * @author Geoffrey T. Dairiki
33 * This is the character used in wiki markup to escape characters with
36 define('ESCAPE_CHAR', '~');
38 require_once(dirname(__FILE__).'/HtmlElement.php');
39 require_once('lib/CachedMarkup.php');
40 require_once(dirname(__FILE__).'/stdlib.php');
43 function WikiEscape($text) {
44 return str_replace('#', ESCAPE_CHAR . '#', $text);
47 function UnWikiEscape($text) {
48 return preg_replace('/' . ESCAPE_CHAR . '(.)/', '\1', $text);
52 * Return type from RegexpSet::match and RegexpSet::nextMatch.
56 class RegexpSet_match {
58 * The text leading up the the next match.
66 * The text following the matched text.
70 * Index of the regular expression which matched.
76 * A set of regular expressions.
78 * This class is probably only useful for InlineTransformer.
84 * @param array $regexps A list of regular expressions. The
85 * regular expressions should not include any sub-pattern groups
86 * "(...)". (Anonymous groups, like "(?:...)", as well as
87 * look-ahead and look-behind assertions are okay.)
89 function RegexpSet ($regexps) {
91 $this->_regexps = array_unique($regexps);
92 if (!defined('_INLINE_OPTIMIZATION')) define('_INLINE_OPTIMIZATION',0);
96 * Search text for the next matching regexp from the Regexp Set.
98 * @param string $text The text to search.
100 * @return RegexpSet_match A RegexpSet_match object, or false if no match.
102 function match ($text) {
103 return $this->_match($text, $this->_regexps, '*?');
107 * Search for next matching regexp.
109 * Here, 'next' has two meanings:
111 * Match the next regexp(s) in the set, at the same position as the last match.
113 * If that fails, match the whole RegexpSet, starting after the position of the
116 * @param string $text Text to search.
118 * @param RegexpSet_match $prevMatch A RegexpSet_match object.
119 * $prevMatch should be a match object obtained by a previous
120 * match upon the same value of $text.
122 * @return RegexpSet_match A RegexpSet_match object, or false if no match.
124 function nextMatch ($text, $prevMatch) {
125 // Try to find match at same position.
126 $pos = strlen($prevMatch->prematch);
127 $regexps = array_slice($this->_regexps, $prevMatch->regexp_ind + 1);
129 $repeat = sprintf('{%d}', $pos);
130 if ( ($match = $this->_match($text, $regexps, $repeat)) ) {
131 $match->regexp_ind += $prevMatch->regexp_ind + 1;
137 // Failed. Look for match after current position.
138 $repeat = sprintf('{%d,}?', $pos + 1);
139 return $this->_match($text, $this->_regexps, $repeat);
142 // Syntax: http://www.pcre.org/pcre.txt
143 // x - EXTENDED, ignore whitespace
147 function _match ($text, $regexps, $repeat) {
148 // If one of the regexps is an empty string, php will crash here:
149 // sf.net: Fatal error: Allowed memory size of 8388608 bytes exhausted
150 // (tried to allocate 634 bytes)
151 if (_INLINE_OPTIMIZATION) { // disabled, wrong
152 // So we try to minize memory usage, by looping explicitly,
153 // and storing only those regexp which actually match.
154 // There may be more than one, so we have to find the longest,
155 // and match inside until the shortest is empty.
156 $matched = array(); $matched_ind = array();
157 for ($i=0; $i<count($regexps); $i++) {
158 if (!trim($regexps[$i])) {
159 trigger_error("empty regexp $i", E_USER_WARNING);
162 $pat= "/ ( . $repeat ) ( " . $regexps[$i] . " ) /x";
163 if (preg_match($pat, $text, $_m)) {
164 $m = $_m; // FIXME: prematch, postmatch is wrong
165 $matched[] = $regexps[$i];
170 // To overcome ANCHORED:
171 // We could sort by longest match and iterate over these.
172 if (empty($matched)) return false;
174 $match = new RegexpSet_match;
176 // Optimization: if the matches are only "$" and another, then omit "$"
177 if (! _INLINE_OPTIMIZATION or count($matched) > 2) {
178 assert(!empty($repeat));
179 assert(!empty($regexps));
180 for ($i=0; $i<count($regexps); $i++) {
181 if (!trim($regexps[$i])) {
182 trigger_error("empty regexp $i", E_USER_WARNING);
183 $regexps[$i] = '\Wxxxx\w\W\w\W\w\W\w\W\w\W\w'; // some placeholder
186 // We could do much better, if we would know the matching markup for the
187 // longest regexp match:
188 $hugepat= "/ ( . $repeat ) ( (" . join(')|(', $regexps) . ") ) /Asx";
189 // Proposed premature optimization 1:
190 //$hugepat= "/ ( . $repeat ) ( (" . join(')|(', array_values($matched)) . ") ) /Asx";
192 '/ ( . *? ) ( ($)|(~(?: [[:alnum:]]+ | .))|(&#\d{2,5};)|(\#? \[ .*? [^]\s] .*? \])|((?<![[:alnum:]]) (?:http|https|mailto|ftp|news|nntp|ssh|gopher) : [^\s<>"']+ (?<![ ,.?; \] \) ]))|((?<! [[:alnum:]])(?:AbbeNormal|AcadWiki|Acronym|Advogato|AIWiki|ALife|Annotation|AnnotationWiki|AwarenessWiki|BcWireless|BenefitsWiki|BridgesWiki|bsdWiki|C2find|Cache|Category|CLiki|CmWiki|CreationMatters|DejaNews|DeWikiPedia|Dict|Dictionary|DiveIntoOsx|DocBook|DolphinWiki|DseWiki|EfnetCeeWiki|EfnetCppWiki|EfnetPythonWiki|EfnetXmlWiki|EljWiki|EmacsWiki|FinalEmpire|Foldoc|FoxWiki|FreeBSDman|FreeNetworks|FreshMeat|Google|GoogleGroups|GreenCheese|HammondWiki|Haribeau|IAWiki|MRQE|IMDB|ISBN|JargonFile|JiniWiki|JspWiki|KmWiki|KnowHow|LanifexWiki|LegoWiki|LinuxWiki|LugKR|MathSongsWiki|MbTest|MeatBall|MetaWiki|MetaWikiPedia|MoinMoin|MuWeb|NetVillage|OpenWiki|OrgPatterns|PangalacticOrg|PersonalTelco|php-function|php-lookup|PhpWiki|PhpWikiCvs|PhpWikiDemo|Pikie|PolitizenWiki|PPR|PurlNet|PythonInfo|PythonWiki|PyWiki|RFC|SeaPig|SeattleWireless|SenseisLibrary|Shakti|SourceForge|Squeak|StrikiWiki|SVGWiki|Tavi|Thesaurus|Thinki|TmNet|TMwiki|TWiki|TwistedWiki|Unreal|UseMod|VisualWorks|WebDevWikiNL|WebSeitzWiki|Why|Wiki|WikiPedia|WikiWorld|YpsiEyeball|ZWiki|Upload): \S+ (?<![ ,.?;! \] \) " \' ]))|( (?<![[:alnum:]])(?:[[:upper:]][[:lower:]]+){2,}(?![[:alnum:]]))|((?: (?<! %) %%% (?! %) | <(?:br|BR)> ))|(''|__)|((?:(?<= \s|^|[-"'\/:]) (?: _ (?! _)|\* (?! \*)|= (?! =))|(?<= =) (?: _ (?! _)|\* (?! \*)) (?! =)|(?<= _) (?: \* (?! \*)|= (?! =)) (?! _)|(?<= \*) (?: _ (?! _)|= (?! =)) (?! \*)|(?<= { ) (?: _ (?! _)|\* (?! \*)|= (?! =)) (?! } )|(?<= < ) (?: _ (?! _)|\* (?! \*)|= (?! =)) (?! > )|(?<= \( ) (?: _ (?! _)|\* (?! \*)|= (?! =)) (?! \) ))(?= \S))|(<(?: b|big|i|small|tt|em|strong|cite|code|dfn|kbd|samp|var|sup|sub )>)|(<(?: abbr|acronym )(?: \stitle=[^>]*)?>)|(<\?plugin(?:-form)?\s[^\n]+?\?>) ) /Asx'
194 if (! preg_match($hugepat, $text, $m)) {
197 // Proposed premature optimization 1:
198 //$match->regexp_ind = $matched_ind[count($m) - 4];
199 $match->regexp_ind = count($m) - 4;
201 $match->regexp_ind = $regexp_ind;
204 $match->postmatch = substr($text, strlen($m[0]));
205 $match->prematch = $m[1];
206 $match->match = $m[2];
211 var_dump($regexps); var_dump($matched); var_dump($matched_inc);
212 PrintXML(HTML::dl(HTML::dt("input"),
213 HTML::dd(HTML::pre($text)),
215 HTML::dd(HTML::pre($match->regexp_ind, ":", $regexps[$match->regexp_ind])),
216 HTML::dt("prematch"),
217 HTML::dd(HTML::pre($match->prematch)),
219 HTML::dd(HTML::pre($match->match)),
220 HTML::dt("postmatch"),
221 HTML::dd(HTML::pre($match->postmatch))
232 * A simple markup rule (i.e. terminal token).
234 * These are defined by a regexp.
236 * When a match is found for the regexp, the matching text is replaced.
237 * The replacement content is obtained by calling the SimpleMarkup::markup method.
245 * @return string Regexp which matches this token.
247 function getMatchRegexp () {
248 return $this->_match_regexp;
251 /** Markup matching text.
253 * @param string $match The text which matched the regexp
254 * (obtained from getMatchRegexp).
256 * @return mixed The expansion of the matched text.
258 function markup ($match /*, $body */) {
259 trigger_error("pure virtual", E_USER_ERROR);
264 * A balanced markup rule.
266 * These are defined by a start regexp, and an end regexp.
272 /** Get the starting regexp for this rule.
274 * @return string The starting regexp.
276 function getStartRegexp () {
277 return $this->_start_regexp;
280 /** Get the ending regexp for this rule.
282 * @param string $match The text which matched the starting regexp.
284 * @return string The ending regexp.
286 function getEndRegexp ($match) {
287 return $this->_end_regexp;
290 /** Get expansion for matching input.
292 * @param string $match The text which matched the starting regexp.
294 * @param mixed $body Transformed text found between the starting
295 * and ending regexps.
297 * @return mixed The expansion of the matched text.
299 function markup ($match, $body) {
300 trigger_error("pure virtual", E_USER_ERROR);
304 class Markup_escape extends SimpleMarkup
306 function getMatchRegexp () {
307 return ESCAPE_CHAR . '(?: [[:alnum:]]+ | .)';
310 function markup ($match) {
311 assert(strlen($match) >= 2);
312 return substr($match, 1);
317 * [image.jpg size=50% border=5], [image.jpg size=50x30]
318 * Support for the following attributes: see stdlib.php:LinkImage()
319 * size=<precent>%, size=<width>x<height>
320 * border=n, align=\w+, hspace=n, vspace=n
322 function isImageLink($link) {
323 if (!$link) return false;
324 assert(defined('INLINE_IMAGES'));
325 return preg_match("/\\.(" . INLINE_IMAGES . ")$/i", $link)
326 or preg_match("/\\.(" . INLINE_IMAGES . ")\s+(size|border|align|hspace|vspace)=/i", $link);
329 function LinkBracketLink($bracketlink) {
331 // $bracketlink will start and end with brackets; in between will
332 // be either a page name, a URL or both separated by a pipe.
334 // strip brackets and leading space
335 // FIXME: \n inside [] will lead to errors
336 preg_match('/(\#?) \[\s* (?: (.*?) \s* (?<!' . ESCAPE_CHAR . ')(\|) )? \s* (.+?) \s*\]/x',
337 $bracketlink, $matches);
338 if (count($matches) < 4) {
339 trigger_error(_("Invalid [] syntax ignored").": ".$bracketlink, E_USER_NOTICE);
340 return new Cached_Link;
342 list (, $hash, $label, $bar, $rawlink) = $matches;
344 $label = UnWikiEscape($label);
346 * Check if the user has typed a explicit URL. This solves the
347 * problem where the URLs have a ~ character, which would be stripped away.
348 * "[http:/server/~name/]" will work as expected
349 * "http:/server/~name/" will NOT work as expected, will remove the ~
351 if (strstr($rawlink, "http://") or strstr($rawlink, "https://")) {
353 // Mozilla Browser URI Obfuscation Weakness 2004-06-14
354 // http://www.securityfocus.com/bid/10532/
355 // goodurl+"%2F%20%20%20."+badurl
356 if (preg_match("/%2F(%20)+\./i", $rawlink)) {
357 $rawlink = preg_replace("/%2F(%20)+\./i","%2F.",$rawlink);
360 $link = UnWikiEscape($rawlink);
363 // if label looks like a url to an image, we want an image link.
364 if (isImageLink($label)) {
366 $intermap = getInterwikiMap();
367 if (preg_match("/^" . $intermap->getRegexp() . ":/", $label)) {
368 $imgurl = $intermap->link($label);
369 $imgurl = $imgurl->getAttr('href');
370 } elseif (! preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $imgurl)) {
371 // local theme linkname like 'images/next.gif'.
373 $imgurl = $WikiTheme->getImageURL($imgurl);
375 $label = LinkImage($imgurl, $link);
379 // It's an anchor, not a link...
380 $id = MangleXmlIdentifier($link);
381 return HTML::a(array('name' => $id, 'id' => $id),
382 $bar ? $label : $link);
385 if (preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $link)) {
386 // if it's an image, embed it; otherwise, it's a regular link
387 if (isImageLink($link))
388 return LinkImage($link, $label);
390 return new Cached_ExternalLink($link, $label);
392 elseif (preg_match("/^phpwiki:/", $link))
393 return new Cached_PhpwikiURL($link, $label);
395 * Inline images in Interwiki urls's:
396 * [File:my_image.gif] inlines the image,
397 * File:my_image.gif shows a plain inter-wiki link,
398 * [what a pic|File:my_image.gif] shows a named inter-wiki link to the gif
399 * [File:my_image.gif|what a pic] shows a inlimed image linked to the page "what a pic"
401 elseif (strstr($link,':') and
402 ($intermap = getInterwikiMap()) and
403 preg_match("/^" . $intermap->getRegexp() . ":/", $link)) {
404 if (empty($label) && isImageLink($link)) {
405 // if without label => inlined image [File:xx.gif]
406 $imgurl = $intermap->link($link);
407 return LinkImage($imgurl->getAttr('href'), $label);
409 return new Cached_InterwikiLink($link, $label);
411 // Split anchor off end of pagename.
412 if (preg_match('/\A(.*)(?<!'.ESCAPE_CHAR.')#(.*?)\Z/', $rawlink, $m)) {
413 list(,$rawlink,$anchor) = $m;
414 $pagename = UnWikiEscape($rawlink);
415 $anchor = UnWikiEscape($anchor);
423 return new Cached_WikiLink($pagename, $label, $anchor);
427 class Markup_bracketlink extends SimpleMarkup
429 var $_match_regexp = "\\#? \\[ .*? [^]\\s] .*? \\]";
431 function markup ($match) {
432 $link = LinkBracketLink($match);
433 assert($link->isInlineElement());
438 class Markup_url extends SimpleMarkup
440 function getMatchRegexp () {
441 return "(?<![[:alnum:]]) (?:" . ALLOWED_PROTOCOLS . ") : [^\s<>\"']+ (?<![ ,.?; \] \) ])";
444 function markup ($match) {
445 return new Cached_ExternalLink(UnWikiEscape($match));
450 class Markup_interwiki extends SimpleMarkup
452 function getMatchRegexp () {
454 $map = getInterwikiMap();
455 return "(?<! [[:alnum:]])" . $map->getRegexp(). ": \S+ (?<![ ,.?;! \] \) \" \' ])";
458 function markup ($match) {
459 //$map = getInterwikiMap();
460 return new Cached_InterwikiLink(UnWikiEscape($match));
464 class Markup_wikiword extends SimpleMarkup
466 function getMatchRegexp () {
467 global $WikiNameRegexp;
468 if (!trim($WikiNameRegexp)) return " " . WIKI_NAME_REGEXP;
469 return " $WikiNameRegexp";
472 function markup ($match) {
473 if (!$match) return false;
474 if ($this->_isWikiUserPage($match))
475 return new Cached_UserLink($match); //$this->_UserLink($match);
477 return new Cached_WikiLink($match);
480 // FIXME: there's probably a more useful place to put these two functions
481 function _isWikiUserPage ($page) {
483 $dbi = $request->getDbh();
484 $page_handle = $dbi->getPage($page);
485 if ($page_handle and $page_handle->get('pref'))
491 function _UserLink($PageName) {
492 $link = HTML::a(array('href' => $PageName));
493 $link->pushContent(PossiblyGlueIconToText('wikiuser', $PageName));
494 $link->setAttr('class', 'wikiuser');
499 class Markup_linebreak extends SimpleMarkup
501 //var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> | <(?:br|BR) \/> )";
502 var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> )";
504 function markup ($match) {
509 class Markup_old_emphasis extends BalancedMarkup
511 var $_start_regexp = "''|__";
513 function getEndRegexp ($match) {
517 function markup ($match, $body) {
518 $tag = $match == "''" ? 'em' : 'strong';
519 return new HtmlElement($tag, $body);
523 class Markup_nestled_emphasis extends BalancedMarkup
525 function getStartRegexp() {
526 static $start_regexp = false;
528 if (!$start_regexp) {
529 // The three possible delimiters
530 // (none of which can be followed by itself.)
535 $any = "(?: ${i}|${b}|${tt})"; // any of the three.
537 // Any of [_*=] is okay if preceded by space or one of [-"'/:]
538 $start[] = "(?<= \\s|^|[-\"'\\/:]) ${any}";
540 // _ or * is okay after = as long as not immediately followed by =
541 $start[] = "(?<= =) (?: ${i}|${b}) (?! =)";
543 $start[] = "(?<= _) (?: ${b}|${tt}) (?! _)";
544 $start[] = "(?<= \\*) (?: ${i}|${tt}) (?! \\*)";
547 // any delimiter okay after an opening brace ( [{<(] )
548 // as long as it's not immediately followed by the matching closing
550 $start[] = "(?<= { ) ${any} (?! } )";
551 $start[] = "(?<= < ) ${any} (?! > )";
552 $start[] = "(?<= \\( ) ${any} (?! \\) )";
554 $start = "(?:" . join('|', $start) . ")";
556 // Any of the above must be immediately followed by non-whitespace.
557 $start_regexp = $start . "(?= \S)";
560 return $start_regexp;
563 function getEndRegexp ($match) {
564 $chr = preg_quote($match);
565 return "(?<= \S | ^ ) (?<! $chr) $chr (?! $chr) (?= \s | [-)}>\"'\\/:.,;!? _*=] | $)";
568 function markup ($match, $body) {
570 case '*': return new HtmlElement('b', $body);
571 case '=': return new HtmlElement('tt', $body);
572 case '_': return new HtmlElement('i', $body);
577 class Markup_html_emphasis extends BalancedMarkup
580 "<(?: b|big|i|small|tt|em|strong|cite|code|dfn|kbd|samp|var|sup|sub )>";
582 function getEndRegexp ($match) {
583 return "<\\/" . substr($match, 1);
586 function markup ($match, $body) {
587 $tag = substr($match, 1, -1);
588 return new HtmlElement($tag, $body);
592 class Markup_html_abbr extends BalancedMarkup
594 //rurban: abbr|acronym need an optional title tag.
596 var $_start_regexp = "<(?: abbr|acronym )(?: \stitle=[^>]*)?>";
598 function getEndRegexp ($match) {
599 if (substr($match,1,4) == 'abbr')
603 return "<\\/" . $tag . '>';
606 function markup ($match, $body) {
607 if (substr($match,1,4) == 'abbr')
611 $rest = substr($match,1+strlen($tag),-1);
613 list($key,$val) = explode("=",$rest);
614 $args = array($key => $val);
615 } else $args = array();
616 return new HtmlElement($tag, $args, $body);
620 // Special version for single-line plugins formatting,
621 // like: '<small>< ?plugin PopularNearby ? ></small>'
622 class Markup_plugin extends SimpleMarkup
624 var $_match_regexp = '<\?plugin(?:-form)?\s[^\n]+?\?>';
626 function markup ($match) {
627 //$xml = new Cached_PluginInvocation($match);
628 //$xml->setTightness(true,true);
629 return new Cached_PluginInvocation($match);
634 // TODO: "..." => "…" browser specific display (not cached?)
635 // TODO: "--" => "&emdash;" browser specific display (not cached?)
637 // FIXME: escape '&' somehow.
638 class Markup_isonumchars extends SimpleMarkup {
639 // var $_match_regexp = '<\?plugin(?:-form)?\s[^\n]+?\? >';
640 // no hexnums yet, like ¤ <=> ¤
641 var $_match_regexp = '\&\#\d{2,5};';
643 function markup ($match) {
648 // FIXME: escape '&' somehow.
649 class Markup_isohexchars extends SimpleMarkup {
650 var $_match_regexp = '\&\#x[0-9a-fA-F]{2,4};';
652 function markup ($match) {
657 // FIXME: Do away with magic phpwiki forms. (Maybe phpwiki: links too?)
658 // FIXME: Do away with plugin-links. They seem not to be used.
662 class InlineTransformer
664 var $_regexps = array();
665 var $_markup = array();
667 function InlineTransformer ($markup_types = false) {
669 $markup_types = array('escape', /*'isonumchars', 'isohexchars',*/
670 'bracketlink', 'url',
671 'interwiki', 'wikiword', 'linebreak',
672 'old_emphasis', 'nestled_emphasis',
673 'html_emphasis', 'html_abbr', 'plugin');
674 foreach ($markup_types as $mtype) {
675 $class = "Markup_$mtype";
676 $this->_addMarkup(new $class);
680 function _addMarkup ($markup) {
681 if (isa($markup, 'SimpleMarkup'))
682 $regexp = $markup->getMatchRegexp();
684 $regexp = $markup->getStartRegexp();
686 assert(!isset($this->_markup[$regexp]));
687 $this->_regexps[] = $regexp;
688 $this->_markup[] = $markup;
691 function parse (&$text, $end_regexps = array('$')) {
692 $regexps = $this->_regexps;
694 // $end_re takes precedence: "favor reduce over shift"
695 array_unshift($regexps, $end_regexps[0]);
696 //array_push($regexps, $end_regexps[0]);
697 $regexps = new RegexpSet($regexps);
700 $output = new XmlContent;
702 $match = $regexps->match($input);
705 if ($match->regexp_ind == 0) {
706 // No start pattern found before end pattern.
708 if (isset($markup) and is_object($markup) and isa($markup,'Markup_plugin')) {
709 $current =& $output->_content[count($output->_content)-1];
710 $current->setTightness(true,true);
712 $output->pushContent($match->prematch);
713 $text = $match->postmatch;
717 $markup = $this->_markup[$match->regexp_ind - 1];
718 $body = $this->_parse_markup_body($markup, $match->match, $match->postmatch, $end_regexps);
720 // Couldn't match balanced expression.
721 // Ignore and look for next matching start regexp.
722 $match = $regexps->nextMatch($input, $match);
726 // Matched markup. Eat input, push output.
727 // FIXME: combine adjacent strings.
728 if (isa($markup, 'SimpleMarkup'))
729 $current = $markup->markup($match->match);
731 $current = $markup->markup($match->match, $body);
732 $input = $match->postmatch;
733 if (isset($markup) and is_object($markup) and isa($markup,'Markup_plugin')) {
734 $current->setTightness(true,true);
736 $output->pushContent($match->prematch, $current);
738 $match = $regexps->match($input);
741 // No pattern matched, not even the end pattern.
746 function _parse_markup_body ($markup, $match, &$text, $end_regexps) {
747 if (isa($markup, 'SimpleMarkup'))
748 return true; // Done. SimpleMarkup is simple.
750 if (!is_object($markup)) return false; // Some error: Should assert
751 array_unshift($end_regexps, $markup->getEndRegexp($match));
753 // Optimization: if no end pattern in text, we know the
754 // parse will fail. This is an important optimization,
755 // e.g. when text is "*lots *of *start *delims *with
756 // *no *matching *end *delims".
757 $ends_pat = "/(?:" . join(").*(?:", $end_regexps) . ")/xs";
758 if (!preg_match($ends_pat, $text))
760 return $this->parse($text, $end_regexps);
764 class LinkTransformer extends InlineTransformer
766 function LinkTransformer () {
767 $this->InlineTransformer(array('escape', 'bracketlink', 'url',
768 'interwiki', 'wikiword'));
772 function TransformInline($text, $markup = 2.0, $basepage=false) {
776 $trfm = new InlineTransformer;
780 $text = ConvertOldMarkup($text, 'inline');
784 return new CacheableMarkup($trfm->parse($text), $basepage);
786 return $trfm->parse($text);
789 function TransformLinks($text, $markup = 2.0, $basepage = false) {
793 $trfm = new LinkTransformer;
797 $text = ConvertOldMarkup($text, 'links');
801 return new CacheableMarkup($trfm->parse($text), $basepage);
803 return $trfm->parse($text);
806 // (c-file-style: "gnu")
811 // c-hanging-comment-ender-p: nil
812 // indent-tabs-mode: nil