2 rcs_id('$Id: InlineParser.php,v 1.91 2007-06-07 18:56:57 rurban Exp $');
3 /* Copyright (C) 2002 Geoffrey T. Dairiki <dairiki@dairiki.org>
4 * Copyright (C) 2004,2005,2006,2007 Reini Urban
6 * This file is part of PhpWiki.
8 * PhpWiki is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation; either version 2 of the License, or
11 * (at your option) any later version.
13 * PhpWiki is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with PhpWiki; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 * This is the code which deals with the inline part of the (new-style)
27 * @author Geoffrey T. Dairiki, Reini Urban
33 * This is the character used in wiki markup to escape characters with
36 define('ESCAPE_CHAR', '~');
38 require_once(dirname(__FILE__).'/HtmlElement.php');
39 require_once('lib/CachedMarkup.php');
40 require_once(dirname(__FILE__).'/stdlib.php');
43 function WikiEscape($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 // We could do much better, if we would know the matching markup for the
181 // longest regexp match:
182 $hugepat= "/ ( . $repeat ) ( (" . join(')|(', $regexps) . ") ) /Asx";
183 // Proposed premature optimization 1:
184 //$hugepat= "/ ( . $repeat ) ( (" . join(')|(', array_values($matched)) . ") ) /Asx";
185 if (! preg_match($hugepat, $text, $m)) {
188 // Proposed premature optimization 1:
189 //$match->regexp_ind = $matched_ind[count($m) - 4];
190 $match->regexp_ind = count($m) - 4;
192 $match->regexp_ind = $regexp_ind;
195 $match->postmatch = substr($text, strlen($m[0]));
196 $match->prematch = $m[1];
197 $match->match = $m[2];
200 if (DEBUG & _DEBUG_PARSER) {
201 static $_already_dumped = 0;
202 if (!$_already_dumped) {
204 if (_INLINE_OPTIMIZATION)
206 var_dump($matched_inc);
208 $_already_dumped = 1;
209 PrintXML(HTML::dl(HTML::dt("input"),
210 HTML::dd(HTML::pre($text)),
212 HTML::dd(HTML::pre($match->regexp_ind, ":", $regexps[$match->regexp_ind])),
213 HTML::dt("prematch"),
214 HTML::dd(HTML::pre($match->prematch)),
216 HTML::dd(HTML::pre($match->match)),
217 HTML::dt("postmatch"),
218 HTML::dd(HTML::pre($match->postmatch))
228 * A simple markup rule (i.e. terminal token).
230 * These are defined by a regexp.
232 * When a match is found for the regexp, the matching text is replaced.
233 * The replacement content is obtained by calling the SimpleMarkup::markup method.
241 * @return string Regexp which matches this token.
243 function getMatchRegexp () {
244 return $this->_match_regexp;
247 /** Markup matching text.
249 * @param string $match The text which matched the regexp
250 * (obtained from getMatchRegexp).
252 * @return mixed The expansion of the matched text.
254 function markup ($match /*, $body */) {
255 trigger_error("pure virtual", E_USER_ERROR);
260 * A balanced markup rule.
262 * These are defined by a start regexp, and an end regexp.
268 /** Get the starting regexp for this rule.
270 * @return string The starting regexp.
272 function getStartRegexp () {
273 return $this->_start_regexp;
276 /** Get the ending regexp for this rule.
278 * @param string $match The text which matched the starting regexp.
280 * @return string The ending regexp.
282 function getEndRegexp ($match) {
283 return $this->_end_regexp;
286 /** Get expansion for matching input.
288 * @param string $match The text which matched the starting regexp.
290 * @param mixed $body Transformed text found between the starting
291 * and ending regexps.
293 * @return mixed The expansion of the matched text.
295 function markup ($match, $body) {
296 trigger_error("pure virtual", E_USER_ERROR);
300 class Markup_escape extends SimpleMarkup
302 function getMatchRegexp () {
303 return ESCAPE_CHAR . '(?: [[:alnum:]]+ | .)';
306 function markup ($match) {
307 assert(strlen($match) >= 2);
308 return substr($match, 1);
313 * [image.jpg size=50% border=5], [image.jpg size=50x30]
314 * Support for the following attributes: see stdlib.php:LinkImage()
315 * size=<precent>%, size=<width>x<height>
316 * border=n, align=\w+, hspace=n, vspace=n
318 function isImageLink($link) {
319 if (!$link) return false;
320 assert(defined('INLINE_IMAGES'));
321 return preg_match("/\\.(" . INLINE_IMAGES . ")$/i", $link)
322 or preg_match("/\\.(" . INLINE_IMAGES . ")\s+(size|border|align|hspace|vspace)=/i", $link);
325 function LinkBracketLink($bracketlink) {
327 // $bracketlink will start and end with brackets; in between will
328 // be either a page name, a URL or both separated by a pipe.
330 // Strip brackets and leading space
331 // FIXME: \n inside [] will lead to errors
332 preg_match('/(\#?) \[\s* (?: (.*?) \s* (?<!' . ESCAPE_CHAR . ')(\|) )? \s* (.+?) \s*\]/x',
333 $bracketlink, $matches);
334 if (count($matches) < 4) {
335 trigger_error(_("Invalid [] syntax ignored").": ".$bracketlink, E_USER_WARNING);
336 return new Cached_Link;
338 list (, $hash, $label, $bar, $rawlink) = $matches;
340 $label = UnWikiEscape($label);
342 * Check if the user has typed a explicit URL. This solves the
343 * problem where the URLs have a ~ character, which would be stripped away.
344 * "[http:/server/~name/]" will work as expected
345 * "http:/server/~name/" will NOT work as expected, will remove the ~
347 if ( string_starts_with ($rawlink, "http://")
348 or string_starts_with ($rawlink, "https://") )
351 // Mozilla Browser URI Obfuscation Weakness 2004-06-14
352 // http://www.securityfocus.com/bid/10532/
353 // goodurl+"%2F%20%20%20."+badurl
354 if (preg_match("/%2F(%20)+\./i", $rawlink)) {
355 $rawlink = preg_replace("/%2F(%20)+\./i","%2F.",$rawlink);
358 $link = UnWikiEscape($rawlink);
360 /* Relatives links by Joel Schaubert.
361 * Recognize [../bla] or [/bla] as relative links, without needing http://
362 * but [ /link ] only if SUBPAGE_SEPERATOR is not "/".
363 * Normally /Page links to the subpage /Page.
365 if (SUBPAGE_SEPARATOR == '/') {
366 if (preg_match('/^\.\.\//', $link)) {
367 return new Cached_ExternalLink($link, $label);
369 } else if (preg_match('/^(\.\.\/|\/)/', $link)) {
370 return new Cached_ExternalLink($link, $label);
373 // if label looks like a url to an image, we want an image link.
374 if (isImageLink($label)) {
376 $intermap = getInterwikiMap();
377 if (preg_match("/^" . $intermap->getRegexp() . ":/", $label)) {
378 $imgurl = $intermap->link($label);
379 $imgurl = $imgurl->getAttr('href');
380 } elseif (! preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $imgurl)) {
381 // local theme linkname like 'images/next.gif'.
383 $imgurl = $WikiTheme->getImageURL($imgurl);
385 $label = LinkImage($imgurl, $link);
389 // It's an anchor, not a link...
390 $id = MangleXmlIdentifier($link);
391 return HTML::a(array('name' => $id, 'id' => $id),
392 $bar ? $label : $link);
395 if (preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $link)) {
396 // if it's an image, embed it; otherwise, it's a regular link
397 if (isImageLink($link) and empty($label)) // patch #1348996 by Robert Litwiniec
398 return LinkImage($link, $label);
400 return new Cached_ExternalLink($link, $label);
402 elseif (substr($link,0,8) == 'phpwiki:')
403 return new Cached_PhpwikiURL($link, $label);
405 /* Semantic relations and attributes.
406 * Relation and attribute names must be word chars only, no space.
407 * Links and Attributes may contain everything. word, nums, units, space, groupsep, numsep, ...
409 elseif (preg_match("/^ (\w+) (:[:=]) (.*) $/x", $link) and !isImageLink($link))
410 return new Cached_SemanticLink($link, $label);
412 /* Do not store the link */
413 elseif (substr($link,0,1) == ':')
414 return new Cached_WikiLink($link, $label);
417 * Inline images in Interwiki urls's:
418 * [File:my_image.gif] inlines the image,
419 * File:my_image.gif shows a plain inter-wiki link,
420 * [what a pic|File:my_image.gif] shows a named inter-wiki link to the gif
421 * [File:my_image.gif|what a pic] shows a inlimed image linked to the page "what a pic"
423 * Note that for simplicity we will accept embedded object tags (non-images)
424 * here also, and seperate them later in LinkImage()
426 elseif (strstr($link,':')
427 and ($intermap = getInterwikiMap())
428 and preg_match("/^" . $intermap->getRegexp() . ":/", $link))
430 // trigger_error("label: $label link: $link", E_USER_WARNING);
431 if (empty($label) and isImageLink($link)) {
432 // if without label => inlined image [File:xx.gif]
433 $imgurl = $intermap->link($link);
434 return LinkImage($imgurl->getAttr('href'), $label);
436 return new Cached_InterwikiLink($link, $label);
438 // Split anchor off end of pagename.
439 if (preg_match('/\A(.*)(?<!'.ESCAPE_CHAR.')#(.*?)\Z/', $rawlink, $m)) {
440 list(,$rawlink,$anchor) = $m;
441 $pagename = UnWikiEscape($rawlink);
442 $anchor = UnWikiEscape($anchor);
450 return new Cached_WikiLink($pagename, $label, $anchor);
454 class Markup_bracketlink extends SimpleMarkup
456 var $_match_regexp = "\\#? \\[ .*? [^]\\s] .*? \\]";
458 function markup ($match) {
459 $link = LinkBracketLink($match);
460 assert($link->isInlineElement());
465 class Markup_spellcheck extends SimpleMarkup
467 function Markup_spellcheck () {
468 $this->suggestions = $GLOBALS['request']->getArg('suggestions');
470 function getMatchRegexp () {
471 if (empty($this->suggestions))
472 return "(?# false )";
473 $words = array_keys($this->suggestions);
474 return "(?<= \W ) (?:" . join('|', $words) . ") (?= \W )";
477 function markup ($match) {
478 if (empty($this->suggestions) or empty($this->suggestions[$match]))
480 return new Cached_SpellCheck(UnWikiEscape($match), $this->suggestions[$match]);
484 class Markup_searchhighlight extends SimpleMarkup
486 function Markup_searchhighlight () {
487 $result = $GLOBALS['request']->_searchhighlight;
488 require_once("lib/TextSearchQuery.php");
489 $query = new TextSearchQuery($result['query']);
490 $this->hilight_re = $query->getHighlightRegexp();
491 $this->engine = $result['engine'];
493 function getMatchRegexp () {
494 return $this->hilight_re;
496 function markup ($match) {
497 return new Cached_SearchHighlight(UnWikiEscape($match), $this->engine);
501 class Markup_url extends SimpleMarkup
503 function getMatchRegexp () {
504 return "(?<![[:alnum:]]) (?:" . ALLOWED_PROTOCOLS . ") : [^\s<>\"']+ (?<![ ,.?; \] \) ])";
507 function markup ($match) {
508 return new Cached_ExternalLink(UnWikiEscape($match));
512 class Markup_interwiki extends SimpleMarkup
514 function getMatchRegexp () {
515 $map = getInterwikiMap();
516 return "(?<! [[:alnum:]])" . $map->getRegexp(). ": [^:=]\S+ (?<![ ,.?;! \] \) \" \' ])";
519 function markup ($match) {
520 return new Cached_InterwikiLink(UnWikiEscape($match));
524 class Markup_semanticlink extends SimpleMarkup
526 var $_match_regexp = "(?:\w+:[:=]\S+)"; // no units seperated by space allowed here
528 function markup ($match) {
529 return new Cached_SemanticLink(UnWikiEscape($match));
533 class Markup_wikiword extends SimpleMarkup
535 function getMatchRegexp () {
536 global $WikiNameRegexp;
537 if (!trim($WikiNameRegexp)) return " " . WIKI_NAME_REGEXP;
538 return " $WikiNameRegexp";
541 function markup ($match) {
542 if (!$match) return false;
543 if ($this->_isWikiUserPage($match))
544 return new Cached_UserLink($match); //$this->_UserLink($match);
546 return new Cached_WikiLink($match);
549 // FIXME: there's probably a more useful place to put these two functions
550 function _isWikiUserPage ($page) {
552 $dbi = $request->getDbh();
553 $page_handle = $dbi->getPage($page);
554 if ($page_handle and $page_handle->get('pref'))
560 function _UserLink($PageName) {
561 $link = HTML::a(array('href' => $PageName));
562 $link->pushContent(PossiblyGlueIconToText('wikiuser', $PageName));
563 $link->setAttr('class', 'wikiuser');
568 class Markup_linebreak extends SimpleMarkup
570 //var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> | <(?:br|BR) \/> )";
571 var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> )";
573 function markup ($match) {
578 class Markup_old_emphasis extends BalancedMarkup
580 var $_start_regexp = "''|__";
582 function getEndRegexp ($match) {
586 function markup ($match, $body) {
587 $tag = $match == "''" ? 'em' : 'strong';
588 return new HtmlElement($tag, $body);
592 class Markup_nestled_emphasis extends BalancedMarkup
594 function getStartRegexp() {
595 static $start_regexp = false;
597 if (!$start_regexp) {
598 // The three possible delimiters
599 // (none of which can be followed by itself.)
604 $any = "(?: ${i}|${b}|${tt})"; // any of the three.
606 // Any of [_*=] is okay if preceded by space or one of [-"'/:]
607 $start[] = "(?<= \\s|^|[-\"'\\/:]) ${any}";
609 // _ or * is okay after = as long as not immediately followed by =
610 $start[] = "(?<= =) (?: ${i}|${b}) (?! =)";
612 $start[] = "(?<= _) (?: ${b}|${tt}) (?! _)";
613 $start[] = "(?<= \\*) (?: ${i}|${tt}) (?! \\*)";
616 // any delimiter okay after an opening brace ( [{<(] )
617 // as long as it's not immediately followed by the matching closing
619 $start[] = "(?<= { ) ${any} (?! } )";
620 $start[] = "(?<= < ) ${any} (?! > )";
621 $start[] = "(?<= \\( ) ${any} (?! \\) )";
623 $start = "(?:" . join('|', $start) . ")";
625 // Any of the above must be immediately followed by non-whitespace.
626 $start_regexp = $start . "(?= \S)";
629 return $start_regexp;
632 function getEndRegexp ($match) {
633 $chr = preg_quote($match);
634 return "(?<= \S | ^ ) (?<! $chr) $chr (?! $chr) (?= \s | [-)}>\"'\\/:.,;!? _*=] | $)";
637 function markup ($match, $body) {
639 case '*': return new HtmlElement('b', $body);
640 case '=': return new HtmlElement('tt', $body);
641 case '_': return new HtmlElement('i', $body);
646 class Markup_html_emphasis extends BalancedMarkup
649 "<(?: b|big|i|small|tt|em|strong|cite|code|dfn|kbd|samp|strike|del|var|sup|sub )>";
651 function getEndRegexp ($match) {
652 return "<\\/" . substr($match, 1);
655 function markup ($match, $body) {
656 $tag = substr($match, 1, -1);
657 return new HtmlElement($tag, $body);
661 class Markup_html_divspan extends BalancedMarkup
664 "<(?: div|span )(?: \s[^>]*)?>";
666 function getEndRegexp ($match) {
667 if (substr($match,1,4) == 'span')
671 return "<\\/" . $tag . '>';
674 function markup ($match, $body) {
675 if (substr($match,1,4) == 'span')
679 $rest = substr($match,1+strlen($tag),-1);
681 list($key,$val) = explode("=",$rest);
682 $args = array($key => $val);
683 } else $args = array();
684 return new HtmlElement($tag, $args, $body);
689 class Markup_html_abbr extends BalancedMarkup
691 //rurban: abbr|acronym need an optional title tag.
693 var $_start_regexp = "<(?: abbr|acronym )(?: \stitle=[^>]*)?>";
695 function getEndRegexp ($match) {
696 if (substr($match,1,4) == 'abbr')
700 return "<\\/" . $tag . '>';
703 function markup ($match, $body) {
704 if (substr($match,1,4) == 'abbr')
708 $rest = substr($match,1+strlen($tag),-1);
710 list($key,$val) = explode("=",$rest);
711 $args = array($key => $val);
712 } else $args = array();
713 return new HtmlElement($tag, $args, $body);
717 /** ENABLE_MARKUP_COLOR
718 * See http://www.pmwiki.org/wiki/PmWiki/WikiStyles and
719 * http://www.flexwiki.com/default.aspx/FlexWiki/FormattingRules.html
721 class Markup_color extends BalancedMarkup {
722 // %color=blue% blue text %% and back to normal
723 var $_start_regexp = "%color=(?: [^%]*)%";
724 var $_end_regexp = "%%";
726 function markup ($match, $body) {
727 $color = strtoupper(substr($match, 7, -1));
728 if (strlen($color) != 7
729 and in_array($color, array('RED', 'BLUE', 'GRAY', 'YELLOW', 'GREEN', 'CYAN', 'BLACK')))
730 { // must be a valid color name
731 return new HtmlElement('font', array('color' => $color), $body);
732 } elseif ((substr($color,0,1) == '#')
733 and (strspn(substr($color,1),'0123456789ABCDEF') == strlen($color)-1)) {
734 return new HtmlElement('font', array('color' => $color), $body);
736 trigger_error(sprintf(_("unknown color %s ignored"), substr($match, 7, -1)), E_USER_WARNING);
742 // Special version for single-line plugins formatting,
743 // like: '<small>< ?plugin PopularNearby ? ></small>'
744 class Markup_plugin extends SimpleMarkup
746 var $_match_regexp = '<\?plugin(?:-form)?\s[^\n]+?\?>';
748 function markup ($match) {
749 //$xml = new Cached_PluginInvocation($match);
750 //$xml->setTightness(true,true);
751 return new Cached_PluginInvocation($match);
755 // Special version for plugins in xml syntax
756 // <name arg=value>body</name> or <name /> => < ? plugin pluginname arg=value body ? >
757 // PLUGIN_MARKUP_MAP = "html:RawHtml dot:GraphViz toc:CreateToc amath:AsciiMath richtable:RichTable include:IncludePage tex:TexToPng"
758 class Markup_xml_plugin extends BalancedMarkup
760 //var $_start_regexp = "<(?: ".join('|',PLUGIN_MARKUP_MAP)." )(?: \s[^>]*)>";
762 function getStartRegexp () {
763 global $PLUGIN_MARKUP_MAP;
764 static $_start_regexp;
765 if ($_start_regexp) return $_start_regexp;
766 if (empty($PLUGIN_MARKUP_MAP))
768 //"<(?: html|dot|toc|amath|richtable|include|tex )(?: \s[^>]*)>"
769 $_start_regexp = "<(?: ".join('|',array_keys($PLUGIN_MARKUP_MAP))." )(?: \s[^>]* | / )>";
770 return $_start_regexp;
772 function getEndRegexp ($match) {
773 return "<\\/" . $match . '>';
775 function markup ($match, $body) {
776 global $PLUGIN_MARKUP_MAP;
777 $name = substr($match,2,-2);
779 if (preg_match('/^(\S+)\|(.*)$/', $name, $_m)) {
781 $vars = $_m[2]; //str_replace(' ', '&', $_m[2]);
783 if (!isset($PLUGIN_MARKUP_MAP[$name])) {
784 trigger_error("No plugin for $name $vars defined.", E_USER_WARNING);
787 $plugin = $PLUGIN_MARKUP_MAP[$name];
788 return new Cached_PluginInvocation("<"."?plugin $plugin $vars $body ?".">");
792 /** ENABLE_MARKUP_TEMPLATE
793 * Template syntax similar to mediawiki
795 * => < ? plugin Template page=template ? >
796 * {{template|var=value|...}}
797 * => < ? plugin Template page=template var=value ... ? >
799 class Markup_template_plugin extends SimpleMarkup
801 // patch #1732793: allow \n, mult. {{ }} in one line, and single letters
802 var $_match_regexp = '\{\{.*?\}\}';
804 function markup ($match) {
805 $page = substr(str_replace("\n", "", $match),2,-2); $vars = '';
806 if (preg_match('/^(\S+)\|(.*)$/', $page, $_m)) {
808 $vars = '"' . preg_replace('/\|/', '" "', $_m[2]) . '"';
809 $vars = preg_replace('/"(\S+)=([^"]*)"/', '\\1="\\2"', $vars);
812 $s = '<'.'?plugin Template page="'.$page.'" '.$vars.' ?'.'>';
814 $s = '<'.'?plugin Template page="' . $page . '" ?'.'>';
815 return new Cached_PluginInvocation($s);
819 // "..." => "…" browser specific display (not cached?)
820 // Support some HTML::Entities: (C) for copy, --- for mdash, -- for ndash
821 // TODO: "--" => "&emdash;" browser specific display (not cached?)
823 class Markup_html_entities extends SimpleMarkup {
824 //var $_match_regexp = '(: \.\.\.|\-\-|\-\-\-|\(C\) )';
826 function Markup_html_entities() {
827 $this->_entities = array('...' => '…',
831 '©' => '©',
832 '™' => '™',
834 $this->_match_regexp =
836 join('|', array_map('preg_quote', array_keys($this->_entities))) .
840 function markup ($match) {
841 return HTML::Raw($this->_entities[$match]);
845 class Markup_isonumchars extends SimpleMarkup {
846 var $_match_regexp = '\&\#\d{2,5};';
848 function markup ($match) {
849 return HTML::Raw($match);
853 class Markup_isohexchars extends SimpleMarkup {
854 // hexnums, like ¤ <=> ¤
855 var $_match_regexp = '\&\#x[0-9a-fA-F]{2,4};';
857 function markup ($match) {
858 return HTML::Raw($match);
862 // FIXME: Do away with magic phpwiki forms. (Maybe phpwiki: links too?)
863 // FIXME: Do away with plugin-links. They seem not to be used.
866 class InlineTransformer
868 var $_regexps = array();
869 var $_markup = array();
871 function InlineTransformer ($markup_types = false) {
873 // We need to extend the inline parsers by certain actions, like SearchHighlight,
874 // SpellCheck and maybe CreateToc.
875 if (!$markup_types) {
876 $non_default = false;
877 $markup_types = array
878 ('escape', 'bracketlink', 'url',
879 'interwiki', 'semanticlink', 'wikiword', 'linebreak',
880 'old_emphasis', 'nestled_emphasis',
881 'html_emphasis', 'html_abbr', 'plugin',
882 'isonumchars', 'isohexchars', /*'html_entities'*/
884 if (DISABLE_MARKUP_WIKIWORD)
885 $markup_types = array_remove($markup_types, 'wikiword');
887 $action = $request->getArg('action');
888 if ($action == 'SpellCheck' and $request->getArg('suggestions'))
889 { // insert it after url
890 array_splice($markup_types, 2, 1, array('url','spellcheck'));
892 if (isset($request->_searchhighlight))
893 { // insert it after url
894 array_splice($markup_types, 2, 1, array('url','searchhighlight'));
895 //$request->setArg('searchhighlight', false);
900 foreach ($markup_types as $mtype) {
901 $class = "Markup_$mtype";
902 $this->_addMarkup(new $class);
904 if (ENABLE_MARKUP_DIVSPAN and !$non_default)
905 $this->_addMarkup(new Markup_html_divspan);
906 if (ENABLE_MARKUP_COLOR and !$non_default)
907 $this->_addMarkup(new Markup_color);
908 if (ENABLE_MARKUP_TEMPLATE and !$non_default)
909 $this->_addMarkup(new Markup_template_plugin);
910 // This does not work yet
911 if (0 and PLUGIN_MARKUP_MAP and !$non_default)
912 $this->_addMarkup(new Markup_xml_plugin);
915 function _addMarkup ($markup) {
916 if (isa($markup, 'SimpleMarkup'))
917 $regexp = $markup->getMatchRegexp();
919 $regexp = $markup->getStartRegexp();
921 assert( !isset($this->_markup[$regexp]) );
922 assert( strlen(trim($regexp)) > 0 );
923 $this->_regexps[] = $regexp;
924 $this->_markup[] = $markup;
927 function parse (&$text, $end_regexps = array('$')) {
928 $regexps = $this->_regexps;
930 // $end_re takes precedence: "favor reduce over shift"
931 array_unshift($regexps, $end_regexps[0]);
932 //array_push($regexps, $end_regexps[0]);
933 $regexps = new RegexpSet($regexps);
936 $output = new XmlContent;
938 $match = $regexps->match($input);
941 if ($match->regexp_ind == 0) {
942 // No start pattern found before end pattern.
944 if (isset($markup) and is_object($markup)
945 and isa($markup,'Markup_plugin'))
947 $current =& $output->_content[count($output->_content)-1];
948 $current->setTightness(true,true);
950 $output->pushContent($match->prematch);
951 $text = $match->postmatch;
955 $markup = $this->_markup[$match->regexp_ind - 1];
956 $body = $this->_parse_markup_body($markup, $match->match,
957 $match->postmatch, $end_regexps);
959 // Couldn't match balanced expression.
960 // Ignore and look for next matching start regexp.
961 $match = $regexps->nextMatch($input, $match);
965 // Matched markup. Eat input, push output.
966 // FIXME: combine adjacent strings.
967 if (isa($markup, 'SimpleMarkup'))
968 $current = $markup->markup($match->match);
970 $current = $markup->markup($match->match, $body);
971 $input = $match->postmatch;
972 if (isset($markup) and is_object($markup)
973 and isa($markup,'Markup_plugin'))
975 $current->setTightness(true,true);
977 $output->pushContent($match->prematch, $current);
979 $match = $regexps->match($input);
982 // No pattern matched, not even the end pattern.
987 function _parse_markup_body ($markup, $match, &$text, $end_regexps) {
988 if (isa($markup, 'SimpleMarkup'))
989 return true; // Done. SimpleMarkup is simple.
991 if (!is_object($markup)) return false; // Some error: Should assert
992 array_unshift($end_regexps, $markup->getEndRegexp($match));
994 // Optimization: if no end pattern in text, we know the
995 // parse will fail. This is an important optimization,
996 // e.g. when text is "*lots *of *start *delims *with
997 // *no *matching *end *delims".
998 $ends_pat = "/(?:" . join(").*(?:", $end_regexps) . ")/xs";
999 if (!preg_match($ends_pat, $text))
1001 return $this->parse($text, $end_regexps);
1005 class LinkTransformer extends InlineTransformer
1007 function LinkTransformer () {
1008 $this->InlineTransformer(array('escape', 'bracketlink', 'url',
1009 'semanticlink', 'interwiki', 'wikiword',
1014 class NowikiTransformer extends InlineTransformer
1016 function NowikiTransformer () {
1017 $this->InlineTransformer
1019 'html_emphasis', 'html_abbr', 'plugin',
1020 'isonumchars', 'isohexchars', /*'html_entities',*/
1025 function TransformInline($text, $markup = 2.0, $basepage=false) {
1027 $action = $GLOBALS['request']->getArg('action');
1028 if (empty($trfm) or $action == 'SpellCheck') {
1029 $trfm = new InlineTransformer;
1032 if ($markup < 2.0) {
1033 $text = ConvertOldMarkup($text, 'inline');
1037 return new CacheableMarkup($trfm->parse($text), $basepage);
1039 return $trfm->parse($text);
1042 function TransformLinks($text, $markup = 2.0, $basepage = false) {
1046 $trfm = new LinkTransformer;
1049 if ($markup < 2.0) {
1050 $text = ConvertOldMarkup($text, 'links');
1054 return new CacheableMarkup($trfm->parse($text), $basepage);
1056 return $trfm->parse($text);
1060 * Transform only html markup and entities.
1062 function TransformInlineNowiki($text, $markup = 2.0, $basepage=false) {
1066 $trfm = new NowikiTransformer;
1069 return new CacheableMarkup($trfm->parse($text), $basepage);
1071 return $trfm->parse($text);
1075 // $Log: not supported by cvs2svn $
1076 // Revision 1.90 2007/03/18 17:35:14 rurban
1077 // Fix :DontStoreLink
1079 // Revision 1.89 2007/02/17 14:16:28 rurban
1080 // fix color GREY to GRAY
1082 // Revision 1.88 2007/01/21 13:15:50 rurban
1083 // Support spaces in attributes and relation links
1085 // Revision 1.87 2007/01/20 15:53:51 rurban
1086 // Rewrite of SearchHighlight: through ActionPage and InlineParser
1088 // Revision 1.86 2007/01/20 11:25:07 rurban
1089 // add SpellCheck support
1091 // Revision 1.85 2007/01/07 18:42:49 rurban
1092 // Add support for non-bracket semantic relation parsing. Assert empty regex (interwikimap?) earlier. Change {{Template||}} vars handling to new style. Stricter interwikimap matching not to find semantic links
1094 // Revision 1.84 2007/01/02 13:18:07 rurban
1095 // fix semantic attributes syntax :=, not :-, disable DIVSPAN and PLUGIN_MARKUP_MAP
1097 // Revision 1.83 2006/12/22 00:23:24 rurban
1098 // Fix Bug #1540007 "hardened-php issue, crawlers related"
1099 // Broken str_replace with strings > 200 chars
1101 // Revision 1.82 2006/12/02 19:53:05 rurban
1102 // Simplify DISABLE_MARKUP_WIKIWORD handling by adding the new function
1103 // stdlib: array_remove(). Hopefully PHP will not add this natively sooner
1106 // Revision 1.81 2006/11/19 13:52:52 rurban
1107 // improve debug output: regex only once
1109 // Revision 1.80 2006/10/12 06:32:30 rurban
1110 // Optionally support new tags <div>, <span> with ENABLE_MARKUP_DIVSPAN (in work)
1112 // Revision 1.79 2006/10/08 12:38:11 rurban
1113 // New special interwiki link markup [:LinkTo] without storing the backlink
1115 // Revision 1.78 2006/09/03 09:53:52 rurban
1116 // more colors, case-insensitive color names
1118 // Revision 1.77 2006/08/25 19:02:02 rurban
1119 // patch #1348996 by Robert Litwiniec: fix show image semantics if label is given
1121 // Revision 1.76 2006/08/19 11:02:35 rurban
1122 // add strike and del to html emphasis: Patch #1542894 by Kai Krakow
1124 // Revision 1.75 2006/08/15 13:43:10 rurban
1125 // add Markup_xml_plugin (untested) and fix Markup_template_plugin
1127 // Revision 1.74 2006/07/23 14:03:18 rurban
1128 // add new feature: DISABLE_MARKUP_WIKIWORD
1130 // Revision 1.73 2006/04/15 12:20:36 rurban
1131 // fix relatives links patch by Joel Schaubert for [/
1133 // Revision 1.72 2006/03/07 20:43:29 rurban
1134 // relative external link, if no internal subpage. by joel Schaubert
1136 // Revision 1.71 2005/11/14 22:31:12 rurban
1137 // add SemanticWeb support
1139 // Revision 1.70 2005/10/31 16:45:23 rurban
1140 // added cfg-able markups only for default TextTransformation, not for links and others
1142 // Revision 1.69 2005/09/14 05:57:19 rurban
1143 // make ENABLE_MARKUP_TEMPLATE optional
1145 // Revision 1.68 2005/09/10 21:24:32 rurban
1146 // optionally support {{Template|vars}} syntax
1148 // Revision 1.67 2005/06/06 17:41:20 rurban
1149 // support new ENABLE_MARKUP_COLOR
1151 // Revision 1.66 2005/04/23 11:15:49 rurban
1152 // handle allowed inlined objects within INLINE_IMAGES
1154 // Revision 1.65 2005/03/27 18:24:17 rurban
1158 // (c-file-style: "gnu")
1162 // c-basic-offset: 4
1163 // c-hanging-comment-ender-p: nil
1164 // indent-tabs-mode: nil