2 rcs_id('$Id: InlineParser.php,v 1.88 2007-01-21 13:15:50 rurban Exp $');
3 /* Copyright (C) 2002 Geoffrey T. Dairiki <dairiki@dairiki.org>
4 * Copyright (C) 2004,2005,2006 Reini Urban
6 * This file is part of PhpWiki.
8 * PhpWiki is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation; either version 2 of the License, or
11 * (at your option) any later version.
13 * PhpWiki is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with PhpWiki; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 * This is the code which deals with the inline part of the (new-style)
27 * @author Geoffrey T. Dairiki, Reini Urban
33 * This is the character used in wiki markup to escape characters with
36 define('ESCAPE_CHAR', '~');
38 require_once(dirname(__FILE__).'/HtmlElement.php');
39 require_once('lib/CachedMarkup.php');
40 require_once(dirname(__FILE__).'/stdlib.php');
43 function WikiEscape($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) {
205 var_dump($matched_inc);
207 $_already_dumped = 1;
208 PrintXML(HTML::dl(HTML::dt("input"),
209 HTML::dd(HTML::pre($text)),
211 HTML::dd(HTML::pre($match->regexp_ind, ":", $regexps[$match->regexp_ind])),
212 HTML::dt("prematch"),
213 HTML::dd(HTML::pre($match->prematch)),
215 HTML::dd(HTML::pre($match->match)),
216 HTML::dt("postmatch"),
217 HTML::dd(HTML::pre($match->postmatch))
227 * A simple markup rule (i.e. terminal token).
229 * These are defined by a regexp.
231 * When a match is found for the regexp, the matching text is replaced.
232 * The replacement content is obtained by calling the SimpleMarkup::markup method.
240 * @return string Regexp which matches this token.
242 function getMatchRegexp () {
243 return $this->_match_regexp;
246 /** Markup matching text.
248 * @param string $match The text which matched the regexp
249 * (obtained from getMatchRegexp).
251 * @return mixed The expansion of the matched text.
253 function markup ($match /*, $body */) {
254 trigger_error("pure virtual", E_USER_ERROR);
259 * A balanced markup rule.
261 * These are defined by a start regexp, and an end regexp.
267 /** Get the starting regexp for this rule.
269 * @return string The starting regexp.
271 function getStartRegexp () {
272 return $this->_start_regexp;
275 /** Get the ending regexp for this rule.
277 * @param string $match The text which matched the starting regexp.
279 * @return string The ending regexp.
281 function getEndRegexp ($match) {
282 return $this->_end_regexp;
285 /** Get expansion for matching input.
287 * @param string $match The text which matched the starting regexp.
289 * @param mixed $body Transformed text found between the starting
290 * and ending regexps.
292 * @return mixed The expansion of the matched text.
294 function markup ($match, $body) {
295 trigger_error("pure virtual", E_USER_ERROR);
299 class Markup_escape extends SimpleMarkup
301 function getMatchRegexp () {
302 return ESCAPE_CHAR . '(?: [[:alnum:]]+ | .)';
305 function markup ($match) {
306 assert(strlen($match) >= 2);
307 return substr($match, 1);
312 * [image.jpg size=50% border=5], [image.jpg size=50x30]
313 * Support for the following attributes: see stdlib.php:LinkImage()
314 * size=<precent>%, size=<width>x<height>
315 * border=n, align=\w+, hspace=n, vspace=n
317 function isImageLink($link) {
318 if (!$link) return false;
319 assert(defined('INLINE_IMAGES'));
320 return preg_match("/\\.(" . INLINE_IMAGES . ")$/i", $link)
321 or preg_match("/\\.(" . INLINE_IMAGES . ")\s+(size|border|align|hspace|vspace)=/i", $link);
324 function LinkBracketLink($bracketlink) {
326 // $bracketlink will start and end with brackets; in between will
327 // be either a page name, a URL or both separated by a pipe.
329 // Strip brackets and leading space
330 // FIXME: \n inside [] will lead to errors
331 preg_match('/(\#?) \[\s* (?: (.*?) \s* (?<!' . ESCAPE_CHAR . ')(\|) )? \s* (.+?) \s*\]/x',
332 $bracketlink, $matches);
333 if (count($matches) < 4) {
334 trigger_error(_("Invalid [] syntax ignored").": ".$bracketlink, E_USER_WARNING);
335 return new Cached_Link;
337 list (, $hash, $label, $bar, $rawlink) = $matches;
339 $label = UnWikiEscape($label);
341 * Check if the user has typed a explicit URL. This solves the
342 * problem where the URLs have a ~ character, which would be stripped away.
343 * "[http:/server/~name/]" will work as expected
344 * "http:/server/~name/" will NOT work as expected, will remove the ~
346 if ( string_starts_with ($rawlink, "http://")
347 or string_starts_with ($rawlink, "https://") )
350 // Mozilla Browser URI Obfuscation Weakness 2004-06-14
351 // http://www.securityfocus.com/bid/10532/
352 // goodurl+"%2F%20%20%20."+badurl
353 if (preg_match("/%2F(%20)+\./i", $rawlink)) {
354 $rawlink = preg_replace("/%2F(%20)+\./i","%2F.",$rawlink);
357 $link = UnWikiEscape($rawlink);
359 /* Relatives links by Joel Schaubert.
360 * Recognize [../bla] or [/bla] as relative links, without needing http://
361 * but [ /link ] only if SUBPAGE_SEPERATOR is not "/".
362 * Normally /Page links to the subpage /Page.
364 if (SUBPAGE_SEPARATOR == '/') {
365 if (preg_match('/^\.\.\//', $link)) {
366 return new Cached_ExternalLink($link, $label);
368 } else if (preg_match('/^(\.\.\/|\/)/', $link)) {
369 return new Cached_ExternalLink($link, $label);
372 // if label looks like a url to an image, we want an image link.
373 if (isImageLink($label)) {
375 $intermap = getInterwikiMap();
376 if (preg_match("/^" . $intermap->getRegexp() . ":/", $label)) {
377 $imgurl = $intermap->link($label);
378 $imgurl = $imgurl->getAttr('href');
379 } elseif (! preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $imgurl)) {
380 // local theme linkname like 'images/next.gif'.
382 $imgurl = $WikiTheme->getImageURL($imgurl);
384 $label = LinkImage($imgurl, $link);
388 // It's an anchor, not a link...
389 $id = MangleXmlIdentifier($link);
390 return HTML::a(array('name' => $id, 'id' => $id),
391 $bar ? $label : $link);
394 if (preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $link)) {
395 // if it's an image, embed it; otherwise, it's a regular link
396 if (isImageLink($link) and empty($label)) // patch #1348996 by Robert Litwiniec
397 return LinkImage($link, $label);
399 return new Cached_ExternalLink($link, $label);
401 elseif (substr($link,0,8) == 'phpwiki:')
402 return new Cached_PhpwikiURL($link, $label);
404 /* Semantic relations and attributes.
405 * Relation and attribute names must be word chars only, no space.
406 * Links and Attributes may contain everything. word, nums, units, space, groupsep, numsep, ...
408 elseif (preg_match("/^ (\w+) (:[:=]) (.*) $/x", $link) and !isImageLink($link))
409 return new Cached_SemanticLink($link, $label);
411 /* Do not store the link */
412 elseif (substr($link,0,1) == ':')
413 return new Cached_InterwikiLink($link, $label);
416 * Inline images in Interwiki urls's:
417 * [File:my_image.gif] inlines the image,
418 * File:my_image.gif shows a plain inter-wiki link,
419 * [what a pic|File:my_image.gif] shows a named inter-wiki link to the gif
420 * [File:my_image.gif|what a pic] shows a inlimed image linked to the page "what a pic"
422 * Note that for simplicity we will accept embedded object tags (non-images)
423 * here also, and seperate them later in LinkImage()
425 elseif (strstr($link,':')
426 and ($intermap = getInterwikiMap())
427 and preg_match("/^" . $intermap->getRegexp() . ":/", $link))
429 // trigger_error("label: $label link: $link", E_USER_WARNING);
430 if (empty($label) and isImageLink($link)) {
431 // if without label => inlined image [File:xx.gif]
432 $imgurl = $intermap->link($link);
433 return LinkImage($imgurl->getAttr('href'), $label);
435 return new Cached_InterwikiLink($link, $label);
437 // Split anchor off end of pagename.
438 if (preg_match('/\A(.*)(?<!'.ESCAPE_CHAR.')#(.*?)\Z/', $rawlink, $m)) {
439 list(,$rawlink,$anchor) = $m;
440 $pagename = UnWikiEscape($rawlink);
441 $anchor = UnWikiEscape($anchor);
449 return new Cached_WikiLink($pagename, $label, $anchor);
453 class Markup_bracketlink extends SimpleMarkup
455 var $_match_regexp = "\\#? \\[ .*? [^]\\s] .*? \\]";
457 function markup ($match) {
458 $link = LinkBracketLink($match);
459 assert($link->isInlineElement());
464 class Markup_spellcheck extends SimpleMarkup
466 function Markup_spellcheck () {
467 $this->suggestions = $GLOBALS['request']->getArg('suggestions');
469 function getMatchRegexp () {
470 if (empty($this->suggestions))
471 return "(?# false )";
472 $words = array_keys($this->suggestions);
473 return "(?<= \W ) (?:" . join('|', $words) . ") (?= \W )";
476 function markup ($match) {
477 if (empty($this->suggestions) or empty($this->suggestions[$match]))
479 return new Cached_SpellCheck(UnWikiEscape($match), $this->suggestions[$match]);
483 class Markup_searchhighlight extends SimpleMarkup
485 function Markup_searchhighlight () {
486 $result = $GLOBALS['request']->_searchhighlight;
487 require_once("lib/TextSearchQuery.php");
488 $query = new TextSearchQuery($result['query']);
489 $this->hilight_re = $query->getHighlightRegexp();
490 $this->engine = $result['engine'];
492 function getMatchRegexp () {
493 return $this->hilight_re;
495 function markup ($match) {
496 return new Cached_SearchHighlight(UnWikiEscape($match), $this->engine);
500 class Markup_url extends SimpleMarkup
502 function getMatchRegexp () {
503 return "(?<![[:alnum:]]) (?:" . ALLOWED_PROTOCOLS . ") : [^\s<>\"']+ (?<![ ,.?; \] \) ])";
506 function markup ($match) {
507 return new Cached_ExternalLink(UnWikiEscape($match));
511 class Markup_interwiki extends SimpleMarkup
513 function getMatchRegexp () {
514 $map = getInterwikiMap();
515 return "(?<! [[:alnum:]])" . $map->getRegexp(). ": [^:=]\S+ (?<![ ,.?;! \] \) \" \' ])";
518 function markup ($match) {
519 return new Cached_InterwikiLink(UnWikiEscape($match));
523 class Markup_semanticlink extends SimpleMarkup
525 var $_match_regexp = "(?:\w+:[:=]\S+)"; // no units seperated by space allowed here
527 function markup ($match) {
528 return new Cached_SemanticLink(UnWikiEscape($match));
532 class Markup_wikiword extends SimpleMarkup
534 function getMatchRegexp () {
535 global $WikiNameRegexp;
536 if (!trim($WikiNameRegexp)) return " " . WIKI_NAME_REGEXP;
537 return " $WikiNameRegexp";
540 function markup ($match) {
541 if (!$match) return false;
542 if ($this->_isWikiUserPage($match))
543 return new Cached_UserLink($match); //$this->_UserLink($match);
545 return new Cached_WikiLink($match);
548 // FIXME: there's probably a more useful place to put these two functions
549 function _isWikiUserPage ($page) {
551 $dbi = $request->getDbh();
552 $page_handle = $dbi->getPage($page);
553 if ($page_handle and $page_handle->get('pref'))
559 function _UserLink($PageName) {
560 $link = HTML::a(array('href' => $PageName));
561 $link->pushContent(PossiblyGlueIconToText('wikiuser', $PageName));
562 $link->setAttr('class', 'wikiuser');
567 class Markup_linebreak extends SimpleMarkup
569 //var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> | <(?:br|BR) \/> )";
570 var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> )";
572 function markup ($match) {
577 class Markup_old_emphasis extends BalancedMarkup
579 var $_start_regexp = "''|__";
581 function getEndRegexp ($match) {
585 function markup ($match, $body) {
586 $tag = $match == "''" ? 'em' : 'strong';
587 return new HtmlElement($tag, $body);
591 class Markup_nestled_emphasis extends BalancedMarkup
593 function getStartRegexp() {
594 static $start_regexp = false;
596 if (!$start_regexp) {
597 // The three possible delimiters
598 // (none of which can be followed by itself.)
603 $any = "(?: ${i}|${b}|${tt})"; // any of the three.
605 // Any of [_*=] is okay if preceded by space or one of [-"'/:]
606 $start[] = "(?<= \\s|^|[-\"'\\/:]) ${any}";
608 // _ or * is okay after = as long as not immediately followed by =
609 $start[] = "(?<= =) (?: ${i}|${b}) (?! =)";
611 $start[] = "(?<= _) (?: ${b}|${tt}) (?! _)";
612 $start[] = "(?<= \\*) (?: ${i}|${tt}) (?! \\*)";
615 // any delimiter okay after an opening brace ( [{<(] )
616 // as long as it's not immediately followed by the matching closing
618 $start[] = "(?<= { ) ${any} (?! } )";
619 $start[] = "(?<= < ) ${any} (?! > )";
620 $start[] = "(?<= \\( ) ${any} (?! \\) )";
622 $start = "(?:" . join('|', $start) . ")";
624 // Any of the above must be immediately followed by non-whitespace.
625 $start_regexp = $start . "(?= \S)";
628 return $start_regexp;
631 function getEndRegexp ($match) {
632 $chr = preg_quote($match);
633 return "(?<= \S | ^ ) (?<! $chr) $chr (?! $chr) (?= \s | [-)}>\"'\\/:.,;!? _*=] | $)";
636 function markup ($match, $body) {
638 case '*': return new HtmlElement('b', $body);
639 case '=': return new HtmlElement('tt', $body);
640 case '_': return new HtmlElement('i', $body);
645 class Markup_html_emphasis extends BalancedMarkup
648 "<(?: b|big|i|small|tt|em|strong|cite|code|dfn|kbd|samp|strike|del|var|sup|sub )>";
650 function getEndRegexp ($match) {
651 return "<\\/" . substr($match, 1);
654 function markup ($match, $body) {
655 $tag = substr($match, 1, -1);
656 return new HtmlElement($tag, $body);
660 class Markup_html_divspan extends BalancedMarkup
663 "<(?: div|span )(?: \s[^>]*)?>";
665 function getEndRegexp ($match) {
666 if (substr($match,1,4) == 'span')
670 return "<\\/" . $tag . '>';
673 function markup ($match, $body) {
674 if (substr($match,1,4) == 'span')
678 $rest = substr($match,1+strlen($tag),-1);
680 list($key,$val) = explode("=",$rest);
681 $args = array($key => $val);
682 } else $args = array();
683 return new HtmlElement($tag, $args, $body);
688 class Markup_html_abbr extends BalancedMarkup
690 //rurban: abbr|acronym need an optional title tag.
692 var $_start_regexp = "<(?: abbr|acronym )(?: \stitle=[^>]*)?>";
694 function getEndRegexp ($match) {
695 if (substr($match,1,4) == 'abbr')
699 return "<\\/" . $tag . '>';
702 function markup ($match, $body) {
703 if (substr($match,1,4) == 'abbr')
707 $rest = substr($match,1+strlen($tag),-1);
709 list($key,$val) = explode("=",$rest);
710 $args = array($key => $val);
711 } else $args = array();
712 return new HtmlElement($tag, $args, $body);
716 /** ENABLE_MARKUP_COLOR
717 * See http://www.pmwiki.org/wiki/PmWiki/WikiStyles and
718 * http://www.flexwiki.com/default.aspx/FlexWiki/FormattingRules.html
720 class Markup_color extends BalancedMarkup {
721 // %color=blue% blue text %% and back to normal
722 var $_start_regexp = "%color=(?: [^%]*)%";
723 var $_end_regexp = "%%";
725 function markup ($match, $body) {
726 $color = strtoupper(substr($match, 7, -1));
727 if (strlen($color) != 7
728 and in_array($color, array('RED', 'BLUE', 'GREY', 'YELLOW', 'GREEN', 'CYAN', 'BLACK')))
729 { // must be a valid color name
730 return new HtmlElement('font', array('color' => $color), $body);
731 } elseif ((substr($color,0,1) == '#')
732 and (strspn(substr($color,1),'0123456789ABCDEF') == strlen($color)-1)) {
733 return new HtmlElement('font', array('color' => $color), $body);
735 trigger_error(sprintf(_("unknown color %s ignored"), substr($match, 7, -1)), E_USER_WARNING);
741 // Special version for single-line plugins formatting,
742 // like: '<small>< ?plugin PopularNearby ? ></small>'
743 class Markup_plugin extends SimpleMarkup
745 var $_match_regexp = '<\?plugin(?:-form)?\s[^\n]+?\?>';
747 function markup ($match) {
748 //$xml = new Cached_PluginInvocation($match);
749 //$xml->setTightness(true,true);
750 return new Cached_PluginInvocation($match);
754 // Special version for plugins in xml syntax
755 // <name arg=value>body</name> or <name /> => < ? plugin pluginname arg=value body ? >
756 // PLUGIN_MARKUP_MAP = "html:RawHtml dot:GraphViz toc:CreateToc amath:AsciiMath richtable:RichTable include:IncludePage tex:TexToPng"
757 class Markup_xml_plugin extends BalancedMarkup
759 //var $_start_regexp = "<(?: ".join('|',PLUGIN_MARKUP_MAP)." )(?: \s[^>]*)>";
761 function getStartRegexp () {
762 global $PLUGIN_MARKUP_MAP;
763 static $_start_regexp;
764 if ($_start_regexp) return $_start_regexp;
765 if (empty($PLUGIN_MARKUP_MAP))
767 //"<(?: html|dot|toc|amath|richtable|include|tex )(?: \s[^>]*)>"
768 $_start_regexp = "<(?: ".join('|',array_keys($PLUGIN_MARKUP_MAP))." )(?: \s[^>]* | / )>";
769 return $_start_regexp;
771 function getEndRegexp ($match) {
772 return "<\\/" . $match . '>';
774 function markup ($match, $body) {
775 global $PLUGIN_MARKUP_MAP;
776 $name = substr($match,2,-2);
778 if (preg_match('/^(\S+)\|(.*)$/', $name, $_m)) {
780 $vars = $_m[2]; //str_replace(' ', '&', $_m[2]);
782 if (!isset($PLUGIN_MARKUP_MAP[$name])) {
783 trigger_error("No plugin for $name $vars defined.", E_USER_WARNING);
786 $plugin = $PLUGIN_MARKUP_MAP[$name];
787 return new Cached_PluginInvocation("<"."?plugin $plugin $vars $body ?".">");
791 /** ENABLE_MARKUP_TEMPLATE
792 * Template syntax similar to mediawiki
794 * => < ? plugin Template page=template ? >
795 * {{template|var=value|...}}
796 * => < ? plugin Template page=template var=value ... ? >
798 class Markup_template_plugin extends SimpleMarkup
800 var $_match_regexp = '\{\{\w[^\n]+\}\}';
802 function markup ($match) {
803 $page = substr($match, 2, -2); $vars = '';
804 if (preg_match('/^(\S+)\|(.*)$/', $page, $_m)) {
806 $vars = preg_replace('/\|/', ' ', $_m[2]);
809 $s = '<'.'?plugin Template page="' . $page . '" ' . $vars . ' ?'.'>';
811 $s = '<'.'?plugin Template page="' . $page . '" ?'.'>';
812 return new Cached_PluginInvocation($s);
816 // "..." => "…" browser specific display (not cached?)
817 // Support some HTML::Entities: (C) for copy, --- for mdash, -- for ndash
818 // TODO: "--" => "&emdash;" browser specific display (not cached?)
820 class Markup_html_entities extends SimpleMarkup {
821 //var $_match_regexp = '(: \.\.\.|\-\-|\-\-\-|\(C\) )';
823 function Markup_html_entities() {
824 $this->_entities = array('...' => '…',
828 '©' => '©',
829 '™' => '™',
831 $this->_match_regexp =
833 join('|', array_map('preg_quote', array_keys($this->_entities))) .
837 function markup ($match) {
838 return HTML::Raw($this->_entities[$match]);
842 class Markup_isonumchars extends SimpleMarkup {
843 var $_match_regexp = '\&\#\d{2,5};';
845 function markup ($match) {
846 return HTML::Raw($match);
850 class Markup_isohexchars extends SimpleMarkup {
851 // hexnums, like ¤ <=> ¤
852 var $_match_regexp = '\&\#x[0-9a-fA-F]{2,4};';
854 function markup ($match) {
855 return HTML::Raw($match);
859 // FIXME: Do away with magic phpwiki forms. (Maybe phpwiki: links too?)
860 // FIXME: Do away with plugin-links. They seem not to be used.
863 class InlineTransformer
865 var $_regexps = array();
866 var $_markup = array();
868 function InlineTransformer ($markup_types = false) {
870 // We need to extend the inline parsers by certain actions, like SearchHighlight,
871 // SpellCheck and maybe CreateToc.
872 if (!$markup_types) {
873 $non_default = false;
874 $markup_types = array
875 ('escape', 'bracketlink', 'url',
876 'interwiki', 'semanticlink', 'wikiword', 'linebreak',
877 'old_emphasis', 'nestled_emphasis',
878 'html_emphasis', 'html_abbr', 'plugin',
879 'isonumchars', 'isohexchars', /*'html_entities'*/
881 if (DISABLE_MARKUP_WIKIWORD)
882 $markup_types = array_remove($markup_types, 'wikiword');
884 $action = $request->getArg('action');
885 if ($action == 'SpellCheck' and $request->getArg('suggestions'))
886 { // insert it after url
887 array_splice($markup_types, 2, 1, array('url','spellcheck'));
889 if (isset($request->_searchhighlight))
890 { // insert it after url
891 array_splice($markup_types, 2, 1, array('url','searchhighlight'));
892 //$request->setArg('searchhighlight', false);
897 foreach ($markup_types as $mtype) {
898 $class = "Markup_$mtype";
899 $this->_addMarkup(new $class);
901 if (ENABLE_MARKUP_DIVSPAN and !$non_default)
902 $this->_addMarkup(new Markup_html_divspan);
903 if (ENABLE_MARKUP_COLOR and !$non_default)
904 $this->_addMarkup(new Markup_color);
905 if (ENABLE_MARKUP_TEMPLATE and !$non_default)
906 $this->_addMarkup(new Markup_template_plugin);
907 // This does not work yet
908 if (0 and PLUGIN_MARKUP_MAP and !$non_default)
909 $this->_addMarkup(new Markup_xml_plugin);
912 function _addMarkup ($markup) {
913 if (isa($markup, 'SimpleMarkup'))
914 $regexp = $markup->getMatchRegexp();
916 $regexp = $markup->getStartRegexp();
918 assert( !isset($this->_markup[$regexp]) );
919 assert( strlen(trim($regexp)) > 0 );
920 $this->_regexps[] = $regexp;
921 $this->_markup[] = $markup;
924 function parse (&$text, $end_regexps = array('$')) {
925 $regexps = $this->_regexps;
927 // $end_re takes precedence: "favor reduce over shift"
928 array_unshift($regexps, $end_regexps[0]);
929 //array_push($regexps, $end_regexps[0]);
930 $regexps = new RegexpSet($regexps);
933 $output = new XmlContent;
935 $match = $regexps->match($input);
938 if ($match->regexp_ind == 0) {
939 // No start pattern found before end pattern.
941 if (isset($markup) and is_object($markup)
942 and isa($markup,'Markup_plugin'))
944 $current =& $output->_content[count($output->_content)-1];
945 $current->setTightness(true,true);
947 $output->pushContent($match->prematch);
948 $text = $match->postmatch;
952 $markup = $this->_markup[$match->regexp_ind - 1];
953 $body = $this->_parse_markup_body($markup, $match->match,
954 $match->postmatch, $end_regexps);
956 // Couldn't match balanced expression.
957 // Ignore and look for next matching start regexp.
958 $match = $regexps->nextMatch($input, $match);
962 // Matched markup. Eat input, push output.
963 // FIXME: combine adjacent strings.
964 if (isa($markup, 'SimpleMarkup'))
965 $current = $markup->markup($match->match);
967 $current = $markup->markup($match->match, $body);
968 $input = $match->postmatch;
969 if (isset($markup) and is_object($markup)
970 and isa($markup,'Markup_plugin'))
972 $current->setTightness(true,true);
974 $output->pushContent($match->prematch, $current);
976 $match = $regexps->match($input);
979 // No pattern matched, not even the end pattern.
984 function _parse_markup_body ($markup, $match, &$text, $end_regexps) {
985 if (isa($markup, 'SimpleMarkup'))
986 return true; // Done. SimpleMarkup is simple.
988 if (!is_object($markup)) return false; // Some error: Should assert
989 array_unshift($end_regexps, $markup->getEndRegexp($match));
991 // Optimization: if no end pattern in text, we know the
992 // parse will fail. This is an important optimization,
993 // e.g. when text is "*lots *of *start *delims *with
994 // *no *matching *end *delims".
995 $ends_pat = "/(?:" . join(").*(?:", $end_regexps) . ")/xs";
996 if (!preg_match($ends_pat, $text))
998 return $this->parse($text, $end_regexps);
1002 class LinkTransformer extends InlineTransformer
1004 function LinkTransformer () {
1005 $this->InlineTransformer(array('escape', 'bracketlink', 'url',
1006 'semanticlink', 'interwiki', 'wikiword',
1011 class NowikiTransformer extends InlineTransformer
1013 function NowikiTransformer () {
1014 $this->InlineTransformer
1016 'html_emphasis', 'html_abbr', 'plugin',
1017 'isonumchars', 'isohexchars', /*'html_entities',*/
1022 function TransformInline($text, $markup = 2.0, $basepage=false) {
1024 $action = $GLOBALS['request']->getArg('action');
1025 if (empty($trfm) or $action == 'SpellCheck') {
1026 $trfm = new InlineTransformer;
1029 if ($markup < 2.0) {
1030 $text = ConvertOldMarkup($text, 'inline');
1034 return new CacheableMarkup($trfm->parse($text), $basepage);
1036 return $trfm->parse($text);
1039 function TransformLinks($text, $markup = 2.0, $basepage = false) {
1043 $trfm = new LinkTransformer;
1046 if ($markup < 2.0) {
1047 $text = ConvertOldMarkup($text, 'links');
1051 return new CacheableMarkup($trfm->parse($text), $basepage);
1053 return $trfm->parse($text);
1057 * Transform only html markup and entities.
1059 function TransformInlineNowiki($text, $markup = 2.0, $basepage=false) {
1063 $trfm = new NowikiTransformer;
1066 return new CacheableMarkup($trfm->parse($text), $basepage);
1068 return $trfm->parse($text);
1072 // $Log: not supported by cvs2svn $
1073 // Revision 1.87 2007/01/20 15:53:51 rurban
1074 // Rewrite of SearchHighlight: through ActionPage and InlineParser
1076 // Revision 1.86 2007/01/20 11:25:07 rurban
1077 // add SpellCheck support
1079 // Revision 1.85 2007/01/07 18:42:49 rurban
1080 // 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
1082 // Revision 1.84 2007/01/02 13:18:07 rurban
1083 // fix semantic attributes syntax :=, not :-, disable DIVSPAN and PLUGIN_MARKUP_MAP
1085 // Revision 1.83 2006/12/22 00:23:24 rurban
1086 // Fix Bug #1540007 "hardened-php issue, crawlers related"
1087 // Broken str_replace with strings > 200 chars
1089 // Revision 1.82 2006/12/02 19:53:05 rurban
1090 // Simplify DISABLE_MARKUP_WIKIWORD handling by adding the new function
1091 // stdlib: array_remove(). Hopefully PHP will not add this natively sooner
1094 // Revision 1.81 2006/11/19 13:52:52 rurban
1095 // improve debug output: regex only once
1097 // Revision 1.80 2006/10/12 06:32:30 rurban
1098 // Optionally support new tags <div>, <span> with ENABLE_MARKUP_DIVSPAN (in work)
1100 // Revision 1.79 2006/10/08 12:38:11 rurban
1101 // New special interwiki link markup [:LinkTo] without storing the backlink
1103 // Revision 1.78 2006/09/03 09:53:52 rurban
1104 // more colors, case-insensitive color names
1106 // Revision 1.77 2006/08/25 19:02:02 rurban
1107 // patch #1348996 by Robert Litwiniec: fix show image semantics if label is given
1109 // Revision 1.76 2006/08/19 11:02:35 rurban
1110 // add strike and del to html emphasis: Patch #1542894 by Kai Krakow
1112 // Revision 1.75 2006/08/15 13:43:10 rurban
1113 // add Markup_xml_plugin (untested) and fix Markup_template_plugin
1115 // Revision 1.74 2006/07/23 14:03:18 rurban
1116 // add new feature: DISABLE_MARKUP_WIKIWORD
1118 // Revision 1.73 2006/04/15 12:20:36 rurban
1119 // fix relatives links patch by Joel Schaubert for [/
1121 // Revision 1.72 2006/03/07 20:43:29 rurban
1122 // relative external link, if no internal subpage. by joel Schaubert
1124 // Revision 1.71 2005/11/14 22:31:12 rurban
1125 // add SemanticWeb support
1127 // Revision 1.70 2005/10/31 16:45:23 rurban
1128 // added cfg-able markups only for default TextTransformation, not for links and others
1130 // Revision 1.69 2005/09/14 05:57:19 rurban
1131 // make ENABLE_MARKUP_TEMPLATE optional
1133 // Revision 1.68 2005/09/10 21:24:32 rurban
1134 // optionally support {{Template|vars}} syntax
1136 // Revision 1.67 2005/06/06 17:41:20 rurban
1137 // support new ENABLE_MARKUP_COLOR
1139 // Revision 1.66 2005/04/23 11:15:49 rurban
1140 // handle allowed inlined objects within INLINE_IMAGES
1142 // Revision 1.65 2005/03/27 18:24:17 rurban
1146 // (c-file-style: "gnu")
1150 // c-basic-offset: 4
1151 // c-hanging-comment-ender-p: nil
1152 // indent-tabs-mode: nil