3 /* Copyright (C) 2002 Geoffrey T. Dairiki <dairiki@dairiki.org>
4 * Copyright (C) 2004-2008 Reini Urban
5 * Copyright (C) 2008-2009 Marc-Etienne Vargenau, Alcatel-Lucent
7 * This file is part of PhpWiki.
9 * PhpWiki is free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License as published by
11 * the Free Software Foundation; either version 2 of the License, or
12 * (at your option) any later version.
14 * PhpWiki is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 * GNU General Public License for more details.
19 * You should have received a copy of the GNU General Public License
20 * along with PhpWiki; if not, write to the Free Software
21 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
24 * This is the code which deals with the inline part of the (new-style)
28 * @author Geoffrey T. Dairiki, Reini Urban
34 * This is the character used in wiki markup to escape characters with
37 define('ESCAPE_CHAR', '~');
39 require_once(dirname(__FILE__).'/HtmlElement.php');
40 require_once('lib/CachedMarkup.php');
41 require_once(dirname(__FILE__).'/stdlib.php');
44 function WikiEscape($text) {
45 return str_replace('#', ESCAPE_CHAR . '#', $text);
48 function UnWikiEscape($text) {
49 return preg_replace('/' . ESCAPE_CHAR . '(.)/', '\1', $text);
53 * Return type from RegexpSet::match and RegexpSet::nextMatch.
57 class RegexpSet_match {
59 * The text leading up the the next match.
67 * The text following the matched text.
71 * Index of the regular expression which matched.
77 * A set of regular expressions.
79 * This class is probably only useful for InlineTransformer.
85 * @param array $regexps A list of regular expressions. The
86 * regular expressions should not include any sub-pattern groups
87 * "(...)". (Anonymous groups, like "(?:...)", as well as
88 * look-ahead and look-behind assertions are okay.)
90 function RegexpSet ($regexps) {
92 $this->_regexps = array_unique($regexps);
93 if (!defined('_INLINE_OPTIMIZATION')) define('_INLINE_OPTIMIZATION',0);
97 * Search text for the next matching regexp from the Regexp Set.
99 * @param string $text The text to search.
101 * @return RegexpSet_match A RegexpSet_match object, or false if no match.
103 function match ($text) {
104 return $this->_match($text, $this->_regexps, '*?');
108 * Search for next matching regexp.
110 * Here, 'next' has two meanings:
112 * Match the next regexp(s) in the set, at the same position as the last match.
114 * If that fails, match the whole RegexpSet, starting after the position of the
117 * @param string $text Text to search.
119 * @param RegexpSet_match $prevMatch A RegexpSet_match object.
120 * $prevMatch should be a match object obtained by a previous
121 * match upon the same value of $text.
123 * @return RegexpSet_match A RegexpSet_match object, or false if no match.
125 function nextMatch ($text, $prevMatch) {
126 // Try to find match at same position.
127 $pos = strlen($prevMatch->prematch);
128 $regexps = array_slice($this->_regexps, $prevMatch->regexp_ind + 1);
130 $repeat = sprintf('{%d}', $pos);
131 if ( ($match = $this->_match($text, $regexps, $repeat)) ) {
132 $match->regexp_ind += $prevMatch->regexp_ind + 1;
138 // Failed. Look for match after current position.
139 $repeat = sprintf('{%d,}?', $pos + 1);
140 return $this->_match($text, $this->_regexps, $repeat);
143 // Syntax: http://www.pcre.org/pcre.txt
144 // x - EXTENDED, ignore whitespace
148 function _match ($text, $regexps, $repeat) {
149 // If one of the regexps is an empty string, php will crash here:
150 // sf.net: Fatal error: Allowed memory size of 8388608 bytes exhausted
151 // (tried to allocate 634 bytes)
152 if (_INLINE_OPTIMIZATION) { // disabled, wrong
153 // So we try to minize memory usage, by looping explicitly,
154 // and storing only those regexp which actually match.
155 // There may be more than one, so we have to find the longest,
156 // and match inside until the shortest is empty.
157 $matched = array(); $matched_ind = array();
158 for ($i=0; $i<count($regexps); $i++) {
159 if (!trim($regexps[$i])) {
160 trigger_error("empty regexp $i", E_USER_WARNING);
163 $pat= "/ ( . $repeat ) ( " . $regexps[$i] . " ) /x";
164 if (preg_match($pat, $text, $_m)) {
165 $m = $_m; // FIXME: prematch, postmatch is wrong
166 $matched[] = $regexps[$i];
171 // To overcome ANCHORED:
172 // We could sort by longest match and iterate over these.
173 if (empty($matched)) return false;
175 $match = new RegexpSet_match;
177 // Optimization: if the matches are only "$" and another, then omit "$"
178 if (! _INLINE_OPTIMIZATION or count($matched) > 2) {
179 assert(!empty($repeat));
180 assert(!empty($regexps));
181 // We could do much better, if we would know the matching markup for the
182 // longest regexp match:
183 $hugepat= "/ ( . $repeat ) ( (" . join(')|(', $regexps) . ") ) /Asx";
184 // Proposed premature optimization 1:
185 //$hugepat= "/ ( . $repeat ) ( (" . join(')|(', array_values($matched)) . ") ) /Asx";
186 if (! preg_match($hugepat, $text, $m)) {
189 // Proposed premature optimization 1:
190 //$match->regexp_ind = $matched_ind[count($m) - 4];
191 $match->regexp_ind = count($m) - 4;
193 $match->regexp_ind = $regexp_ind;
196 $match->postmatch = substr($text, strlen($m[0]));
197 $match->prematch = $m[1];
198 $match->match = $m[2];
201 if (DEBUG & _DEBUG_PARSER) {
202 static $_already_dumped = 0;
203 if (!$_already_dumped) {
205 if (_INLINE_OPTIMIZATION)
207 var_dump($matched_inc);
209 $_already_dumped = 1;
210 PrintXML(HTML::dl(HTML::dt("input"),
211 HTML::dd(HTML::pre($text)),
213 HTML::dd(HTML::pre($match->regexp_ind, ":", $regexps[$match->regexp_ind])),
214 HTML::dt("prematch"),
215 HTML::dd(HTML::pre($match->prematch)),
217 HTML::dd(HTML::pre($match->match)),
218 HTML::dt("postmatch"),
219 HTML::dd(HTML::pre($match->postmatch))
229 * A simple markup rule (i.e. terminal token).
231 * These are defined by a regexp.
233 * When a match is found for the regexp, the matching text is replaced.
234 * The replacement content is obtained by calling the SimpleMarkup::markup method.
242 * @return string Regexp which matches this token.
244 function getMatchRegexp () {
245 return $this->_match_regexp;
248 /** Markup matching text.
250 * @param string $match The text which matched the regexp
251 * (obtained from getMatchRegexp).
253 * @return mixed The expansion of the matched text.
255 function markup ($match /*, $body */) {
256 trigger_error("pure virtual", E_USER_ERROR);
261 * A balanced markup rule.
263 * These are defined by a start regexp, and an end regexp.
269 /** Get the starting regexp for this rule.
271 * @return string The starting regexp.
273 function getStartRegexp () {
274 return $this->_start_regexp;
277 /** Get the ending regexp for this rule.
279 * @param string $match The text which matched the starting regexp.
281 * @return string The ending regexp.
283 function getEndRegexp ($match) {
284 return $this->_end_regexp;
287 /** Get expansion for matching input.
289 * @param string $match The text which matched the starting regexp.
291 * @param mixed $body Transformed text found between the starting
292 * and ending regexps.
294 * @return mixed The expansion of the matched text.
296 function markup ($match, $body) {
297 trigger_error("pure virtual", E_USER_ERROR);
301 class Markup_escape extends SimpleMarkup
303 function getMatchRegexp () {
304 return ESCAPE_CHAR . '(?: [[:alnum:]]+ | .)';
307 function markup ($match) {
308 assert(strlen($match) >= 2);
309 return substr($match, 1);
314 * [image.jpg size=50% border=5], [image.jpg size=50x30]
315 * Support for the following attributes: see stdlib.php:LinkImage()
316 * size=<precent>%, size=<width>x<height>
317 * border=n, align=\w+, hspace=n, vspace=n
320 function isImageLink($link) {
321 if (!$link) return false;
322 assert(defined('INLINE_IMAGES'));
323 return preg_match("/\\.(" . INLINE_IMAGES . ")$/i", $link)
324 or preg_match("/\\.(" . INLINE_IMAGES . ")\s+(size|border|align|hspace|vspace|type|data|width|height)=/i", $link);
327 function LinkBracketLink($bracketlink) {
329 // $bracketlink will start and end with brackets; in between will
330 // be either a page name, a URL or both separated by a pipe.
332 $wikicreolesyntax = false;
334 if (string_starts_with($bracketlink, "[[")) {
335 $wikicreolesyntax = true;
336 $bracketlink = str_replace("[[", "[", $bracketlink);
337 $bracketlink = str_replace("]]", "]", $bracketlink);
340 // Strip brackets and leading space
341 // bug#1904088 Some brackets links on 2 lines cause the parser to crash
342 preg_match('/(\#?) \[\s* (?: (.*?) \s* (?<!' . ESCAPE_CHAR . ')(\|) )? \s* (.+?) \s*\]/x',
343 str_replace("\n", " ", $bracketlink), $matches);
344 if (count($matches) < 4) {
345 // "[ personal\ninformation manager | PhpWiki:PersonalWiki ]"
346 trigger_error(_("Invalid [] syntax ignored").": ".$bracketlink, E_USER_WARNING);
347 return new Cached_Link;
349 list (, $hash, $label, $bar, $rawlink) = $matches;
351 if ($wikicreolesyntax and $label) {
357 $label = UnWikiEscape($label);
359 * Check if the user has typed a explicit URL. This solves the
360 * problem where the URLs have a ~ character, which would be stripped away.
361 * "[http:/server/~name/]" will work as expected
362 * "http:/server/~name/" will NOT work as expected, will remove the ~
364 if ( string_starts_with ($rawlink, "http://")
365 or string_starts_with ($rawlink, "https://") )
368 // Mozilla Browser URI Obfuscation Weakness 2004-06-14
369 // http://www.securityfocus.com/bid/10532/
370 // goodurl+"%2F%20%20%20."+badurl
371 if (preg_match("/%2F(%20)+\./i", $rawlink)) {
372 $rawlink = preg_replace("/%2F(%20)+\./i","%2F.",$rawlink);
375 $link = UnWikiEscape($rawlink);
377 /* Relatives links by Joel Schaubert.
378 * Recognize [../bla] or [/bla] as relative links, without needing http://
379 * but [ /link ] only if SUBPAGE_SEPERATOR is not "/".
380 * Normally /Page links to the subpage /Page.
382 if (SUBPAGE_SEPARATOR == '/') {
383 if (preg_match('/^\.\.\//', $link)) {
384 return new Cached_ExternalLink($link, $label);
386 } else if (preg_match('/^(\.\.\/|\/)/', $link)) {
387 return new Cached_ExternalLink($link, $label);
390 // If label looks like a url to an image or object, we want an image link.
391 if (isImageLink($label)) {
393 $intermap = getInterwikiMap();
394 if (preg_match("/^" . $intermap->getRegexp() . ":/", $label)) {
395 $imgurl = $intermap->link($label);
396 $imgurl = $imgurl->getAttr('href');
397 } elseif (! preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $imgurl)) {
398 // local theme linkname like 'images/next.gif'.
400 $imgurl = $WikiTheme->getImageURL($imgurl);
402 // for objects (non-images) the link is taken as alt tag,
403 // which is in return taken as alternative img
404 $label = LinkImage($imgurl, $link);
408 // It's an anchor, not a link...
409 $id = MangleXmlIdentifier($link);
410 return HTML::a(array('name' => $id, 'id' => $id),
411 $bar ? $label : $link);
414 if (preg_match("#^(" . ALLOWED_PROTOCOLS . "):#", $link)) {
415 // if it's an image, embed it; otherwise, it's a regular link
416 if (isImageLink($link) and empty($label)) // patch #1348996 by Robert Litwiniec
417 return LinkImage($link, $label);
419 return new Cached_ExternalLink($link, $label);
421 elseif (substr($link,0,8) == 'phpwiki:')
422 return new Cached_PhpwikiURL($link, $label);
424 /* Semantic relations and attributes.
425 * Relation and attribute names must be word chars only, no space.
426 * Links and Attributes may contain everything. word, nums, units, space, groupsep, numsep, ...
428 elseif (preg_match("/^ (\w+) (:[:=]) (.*) $/x", $link) and !isImageLink($link))
429 return new Cached_SemanticLink($link, $label);
431 /* Do not store the link */
432 elseif (substr($link,0,1) == ':')
433 return new Cached_WikiLink($link, $label);
436 * Inline images in Interwiki urls's:
437 * [File:my_image.gif] inlines the image,
438 * File:my_image.gif shows a plain inter-wiki link,
439 * [what a pic|File:my_image.gif] shows a named inter-wiki link to the gif
440 * [File:my_image.gif|what a pic] shows a inlimed image linked to the page "what a pic"
442 * Note that for simplicity we will accept embedded object tags (non-images)
443 * here also, and seperate them later in LinkImage()
445 elseif (strstr($link,':')
446 and ($intermap = getInterwikiMap())
447 and preg_match("/^" . $intermap->getRegexp() . ":/", $link))
449 // trigger_error("label: $label link: $link", E_USER_WARNING);
450 if (empty($label) and isImageLink($link)) {
451 // if without label => inlined image [File:xx.gif]
452 $imgurl = $intermap->link($link);
453 return LinkImage($imgurl->getAttr('href'), $link);
455 return new Cached_InterwikiLink($link, $label);
457 // Split anchor off end of pagename.
458 if (preg_match('/\A(.*)(?<!'.ESCAPE_CHAR.')#(.*?)\Z/', $rawlink, $m)) {
459 list(,$rawlink,$anchor) = $m;
460 $pagename = UnWikiEscape($rawlink);
461 $anchor = UnWikiEscape($anchor);
469 return new Cached_WikiLink($pagename, $label, $anchor);
473 class Markup_wikicreolebracketlink extends SimpleMarkup
475 var $_match_regexp = "\\#? \\[\\[ .*? [^]\\s] .*? \\]\\]";
477 function markup ($match) {
478 $link = LinkBracketLink($match);
479 assert($link->isInlineElement());
484 class Markup_bracketlink extends SimpleMarkup
486 var $_match_regexp = "\\#? \\[ .*? [^]\\s] .*? \\]";
488 function markup ($match) {
489 $link = LinkBracketLink($match);
490 assert($link->isInlineElement());
495 class Markup_spellcheck extends SimpleMarkup
497 function Markup_spellcheck () {
498 $this->suggestions = $GLOBALS['request']->getArg('suggestions');
500 function getMatchRegexp () {
501 if (empty($this->suggestions))
502 return "(?# false )";
503 $words = array_keys($this->suggestions);
504 return "(?<= \W ) (?:" . join('|', $words) . ") (?= \W )";
507 function markup ($match) {
508 if (empty($this->suggestions) or empty($this->suggestions[$match]))
510 return new Cached_SpellCheck(UnWikiEscape($match), $this->suggestions[$match]);
514 class Markup_searchhighlight extends SimpleMarkup
516 function Markup_searchhighlight () {
517 $result = $GLOBALS['request']->_searchhighlight;
518 require_once("lib/TextSearchQuery.php");
519 $query = new TextSearchQuery($result['query']);
520 $this->hilight_re = $query->getHighlightRegexp();
521 $this->engine = $result['engine'];
523 function getMatchRegexp () {
524 return $this->hilight_re;
526 function markup ($match) {
527 return new Cached_SearchHighlight(UnWikiEscape($match), $this->engine);
531 class Markup_url extends SimpleMarkup
533 function getMatchRegexp () {
534 return "(?<![[:alnum:]]) (?:" . ALLOWED_PROTOCOLS . ") : [^\s<>\"']+ (?<![ ,.?; \] \) ])";
537 function markup ($match) {
538 return new Cached_ExternalLink(UnWikiEscape($match));
542 class Markup_interwiki extends SimpleMarkup
544 function getMatchRegexp () {
545 $map = getInterwikiMap();
546 return "(?<! [[:alnum:]])" . $map->getRegexp(). ": [^:=]\S+ (?<![ ,.?;! \] \) \" \' ])";
549 function markup ($match) {
550 return new Cached_InterwikiLink(UnWikiEscape($match));
554 class Markup_semanticlink extends SimpleMarkup
556 // No units seperated by space allowed here
557 // For :: (relations) only words, no comma,
558 // but for := (attributes) comma and dots are allowed. Units with groupsep.
559 // Ending dots or comma are not part of the link.
560 var $_match_regexp = "(?: \w+:=\S+(?<![\.,]))|(?: \w+::[\w\.]+(?<!\.))";
562 function markup ($match) {
563 return new Cached_SemanticLink(UnWikiEscape($match));
567 class Markup_wikiword extends SimpleMarkup
569 function getMatchRegexp () {
570 global $WikiNameRegexp;
571 if (!trim($WikiNameRegexp)) return " " . WIKI_NAME_REGEXP;
572 return " $WikiNameRegexp";
575 function markup ($match) {
576 if (!$match) return false;
577 if ($this->_isWikiUserPage($match))
578 return new Cached_UserLink($match); //$this->_UserLink($match);
580 return new Cached_WikiLink($match);
583 // FIXME: there's probably a more useful place to put these two functions
584 function _isWikiUserPage ($page) {
586 $dbi = $request->getDbh();
587 $page_handle = $dbi->getPage($page);
588 if ($page_handle and $page_handle->get('pref'))
594 function _UserLink($PageName) {
595 $link = HTML::a(array('href' => $PageName));
596 $link->pushContent(PossiblyGlueIconToText('wikiuser', $PageName));
597 $link->setAttr('class', 'wikiuser');
602 class Markup_linebreak extends SimpleMarkup
604 //var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> | <(?:br|BR) \/> )";
605 var $_match_regexp = "(?: (?<! %) %%% (?! %) | \\\\\\\\ | <(?:br|BR)> )";
607 function markup ($match) {
612 class Markup_wikicreole_italics extends BalancedMarkup
614 var $_start_regexp = "\\/\\/";
616 function getEndRegexp ($match) {
620 function markup ($match, $body) {
622 return new HtmlElement($tag, $body);
626 class Markup_wikicreole_bold extends BalancedMarkup
628 var $_start_regexp = "\\*\\*";
630 function getEndRegexp ($match) {
634 function markup ($match, $body) {
636 return new HtmlElement($tag, $body);
640 class Markup_wikicreole_monospace extends BalancedMarkup
642 var $_start_regexp = "\\#\\#";
644 function getEndRegexp ($match) {
648 function markup ($match, $body) {
650 return new HtmlElement($tag, $body);
654 class Markup_wikicreole_superscript extends BalancedMarkup
656 var $_start_regexp = "\\^\\^";
658 function getEndRegexp ($match) {
662 function markup ($match, $body) {
664 return new HtmlElement($tag, $body);
668 class Markup_wikicreole_subscript extends BalancedMarkup
670 var $_start_regexp = ",,";
672 function getEndRegexp ($match) {
676 function markup ($match, $body) {
678 return new HtmlElement($tag, $body);
682 class Markup_old_emphasis extends BalancedMarkup
684 var $_start_regexp = "''|__";
686 function getEndRegexp ($match) {
690 function markup ($match, $body) {
691 $tag = $match == "''" ? 'em' : 'strong';
692 return new HtmlElement($tag, $body);
696 class Markup_nestled_emphasis extends BalancedMarkup
698 function getStartRegexp() {
699 static $start_regexp = false;
701 if (!$start_regexp) {
702 // The three possible delimiters
703 // (none of which can be followed by itself.)
708 $any = "(?: ${i}|${b}|${tt})"; // any of the three.
710 // Any of [_*=] is okay if preceded by space or one of [-"'/:]
711 $start[] = "(?<= \\s|^|[-\"'\\/:]) ${any}";
713 // _ or * is okay after = as long as not immediately followed by =
714 $start[] = "(?<= =) (?: ${i}|${b}) (?! =)";
716 $start[] = "(?<= _) (?: ${b}|${tt}) (?! _)";
717 $start[] = "(?<= \\*) (?: ${i}|${tt}) (?! \\*)";
720 // any delimiter okay after an opening brace ( [{<(] )
721 // as long as it's not immediately followed by the matching closing
723 $start[] = "(?<= { ) ${any} (?! } )";
724 $start[] = "(?<= < ) ${any} (?! > )";
725 $start[] = "(?<= \\( ) ${any} (?! \\) )";
727 $start = "(?:" . join('|', $start) . ")";
729 // Any of the above must be immediately followed by non-whitespace.
730 $start_regexp = $start . "(?= \S)";
733 return $start_regexp;
736 function getEndRegexp ($match) {
737 $chr = preg_quote($match);
738 return "(?<= \S | ^ ) (?<! $chr) $chr (?! $chr) (?= \s | [-)}>\"'\\/:.,;!? _*=] | $)";
741 function markup ($match, $body) {
743 case '*': return new HtmlElement('b', $body);
744 case '=': return new HtmlElement('tt', $body);
745 case '_': return new HtmlElement('i', $body);
750 class Markup_html_emphasis extends BalancedMarkup
753 "<(?: b|big|i|small|tt|em|strong|cite|code|dfn|kbd|samp|s|strike|del|var|sup|sub )>";
755 function getEndRegexp ($match) {
756 return "<\\/" . substr($match, 1);
759 function markup ($match, $body) {
760 $tag = substr($match, 1, -1);
761 return new HtmlElement($tag, $body);
765 class Markup_html_divspan extends BalancedMarkup
768 "<(?: div|span )(?: \s[^>]*)?>";
770 function getEndRegexp ($match) {
771 if (substr($match,1,4) == 'span')
775 return "<\\/" . $tag . '>';
778 function markup ($match, $body) {
779 if (substr($match,1,4) == 'span')
783 $rest = substr($match,1+strlen($tag),-1);
785 list($key,$val) = explode("=",$rest);
786 $args = array($key => $val);
787 } else $args = array();
788 return new HtmlElement($tag, $args, $body);
793 class Markup_html_abbr extends BalancedMarkup
795 //rurban: abbr|acronym need an optional title tag.
797 // allowed attributes: title and lang
798 var $_start_regexp = "<(?: abbr|acronym )(?: [^>]*)?>";
800 function getEndRegexp ($match) {
801 if (substr($match,1,4) == 'abbr')
805 return "<\\/" . $tag . '>';
808 function markup ($match, $body) {
809 if (substr($match,1,4) == 'abbr')
813 $rest = substr($match,1+strlen($tag),-1);
814 $attrs = parse_attributes($rest);
815 // Remove attributes other than title and lang
816 $allowedargs = array();
817 foreach ($attrs as $key => $value) {
818 if (in_array ($key, array("title", "lang"))) {
819 $allowedargs[$key] = $value;
822 return new HtmlElement($tag, $allowedargs, $body);
826 /** ENABLE_MARKUP_COLOR
827 * See http://www.pmwiki.org/wiki/PmWiki/WikiStyles and
828 * http://www.flexwiki.com/default.aspx/FlexWiki/FormattingRules.html
830 class Markup_color extends BalancedMarkup {
831 // %color=blue% blue text %% and back to normal
832 var $_start_regexp = "%color=(?: [^%]*)%";
833 var $_end_regexp = "%%";
835 function markup ($match, $body) {
836 $color = strtoupper(substr($match, 7, -1));
837 if (strlen($color) != 7
838 and in_array($color, array('RED', 'BLUE', 'GRAY', 'YELLOW', 'GREEN', 'CYAN', 'BLACK')))
839 { // must be a valid color name
840 return new HtmlElement('font', array('color' => $color), $body);
841 } elseif ((substr($color,0,1) == '#')
842 and (strspn(substr($color,1),'0123456789ABCDEF') == strlen($color)-1)) {
843 return new HtmlElement('font', array('color' => $color), $body);
845 trigger_error(sprintf(_("unknown color %s ignored"), substr($match, 7, -1)), E_USER_WARNING);
851 // Special version for single-line plugins formatting,
852 // like: '<small>< ?plugin PopularNearby ? ></small>'
853 class Markup_plugin extends SimpleMarkup
855 var $_match_regexp = '<\?plugin(?:-form)?\s[^\n]+?\?>';
857 function markup ($match) {
858 return new Cached_PluginInvocation($match);
862 // Special version for single-line Wikicreole plugins formatting.
863 class Markup_plugin_wikicreole extends SimpleMarkup
865 var $_match_regexp = '<<[^\n]+?>>';
867 function markup ($match) {
868 $pi = str_replace("<<", "<?plugin ", $match);
869 $pi = str_replace(">>", " ?>", $pi);
870 return new Cached_PluginInvocation($pi);
874 // Special version for plugins in xml syntax
875 // <name arg=value>body</name> or <name /> => < ? plugin pluginname arg=value body ? >
876 // PLUGIN_MARKUP_MAP = "html:RawHtml dot:GraphViz toc:CreateToc amath:AsciiMath richtable:RichTable include:IncludePage tex:TexToPng"
877 class Markup_xml_plugin extends BalancedMarkup
879 //var $_start_regexp = "<(?: ".join('|',PLUGIN_MARKUP_MAP)." )(?: \s[^>]*)>";
881 function getStartRegexp () {
882 global $PLUGIN_MARKUP_MAP;
883 static $_start_regexp;
884 if ($_start_regexp) return $_start_regexp;
885 if (empty($PLUGIN_MARKUP_MAP))
887 //"<(?: html|dot|toc|amath|richtable|include|tex )(?: \s[^>]*)>"
888 $_start_regexp = "<(?: ".join('|',array_keys($PLUGIN_MARKUP_MAP))." )(?: \s[^>]* | / )>";
889 return $_start_regexp;
891 function getEndRegexp ($match) {
892 return "<\\/" . $match . '>';
894 function markup ($match, $body) {
895 global $PLUGIN_MARKUP_MAP;
896 $name = substr($match,2,-2);
898 if (preg_match('/^(\S+)\|(.*)$/', $name, $_m)) {
900 $vars = $_m[2]; //str_replace(' ', '&', $_m[2]);
902 if (!isset($PLUGIN_MARKUP_MAP[$name])) {
903 trigger_error("No plugin for $name $vars defined.", E_USER_WARNING);
906 $plugin = $PLUGIN_MARKUP_MAP[$name];
907 return new Cached_PluginInvocation("<"."?plugin $plugin $vars $body ?".">");
912 * Wikicreole preformatted
916 class Markup_wikicreole_preformatted extends SimpleMarkup
918 var $_match_regexp = '\{\{\{.*?\}\}\}';
920 function markup ($match) {
921 // Remove {{{ and }}}
922 return new HtmlElement('pre', substr($match, 3, -3));
926 /** ENABLE_MARKUP_TEMPLATE
927 * Template syntax similar to Mediawiki
929 * => < ? plugin Template page=template ? >
930 * {{template|var1=value1|var2=value|...}}
931 * => < ? plugin Template page=template var=value ... ? >
933 class Markup_template_plugin extends SimpleMarkup
935 // patch #1732793: allow \n, mult. {{ }} in one line, and single letters
936 var $_match_regexp = '\{\{.*?\}\}';
938 function markup ($match) {
940 $page = substr($match,2,-2);
941 if (strpos($page, "|") === false) {
945 $imagename = substr($page, 0, strpos($page, "|"));
946 $alt = ltrim(strstr($page, "|"), "|");
949 // It's not a Mediawiki template, it's a Wikicreole image
950 if (is_image($imagename)) {
951 return LinkImage(UPLOAD_DATA_PATH . $imagename, $alt);
954 $page = str_replace("\n", "", $page);
957 if (preg_match('/^(\S+?)\|(.*)$/', $page, $_m)) {
959 $vars = '"' . preg_replace('/\|/', '" "', $_m[2]) . '"';
960 $vars = preg_replace('/"(\S+)=([^"]*)"/', '\\1="\\2"', $vars);
963 // page may contain a version number
965 // in that case, output is "page=foo rev=5"
966 if (strstr($page, "?")) {
967 $page = str_replace("?version=", "\" rev=\"", $page);
971 $s = '<'.'?plugin Template page="'.$page.'" '.$vars.' ?'.'>';
973 $s = '<'.'?plugin Template page="' . $page . '" ?'.'>';
974 return new Cached_PluginInvocation($s);
978 /** ENABLE_MARKUP_MEDIAWIKI_TABLE
979 * Table syntax similar to Mediawiki
981 * => <?plugin MediawikiTable
985 class Markup_mediawikitable_plugin extends SimpleMarkup
987 var $_match_regexp = '\{\|.*?\|\}';
989 function markup ($match) {
990 $s = '<'.'?plugin MediawikiTable ' . $match . '?'.'>';
991 return new Cached_PluginInvocation($s);
995 // "..." => "…" browser specific display (not cached?)
996 // Support some HTML::Entities: (C) for copy, --- for mdash, -- for ndash
997 // TODO: "--" => "&emdash;" browser specific display (not cached?)
999 class Markup_html_entities extends SimpleMarkup {
1000 //var $_match_regexp = '(: \.\.\.|\-\-|\-\-\-|\(C\) )';
1002 function Markup_html_entities() {
1003 $this->_entities = array('...' => '…',
1007 '©' => '©',
1008 '™' => '™',
1010 $this->_match_regexp =
1012 join('|', array_map('preg_quote', array_keys($this->_entities))) .
1016 function markup ($match) {
1017 return HTML::Raw($this->_entities[$match]);
1021 class Markup_isonumchars extends SimpleMarkup {
1022 var $_match_regexp = '\&\#\d{2,5};';
1024 function markup ($match) {
1025 return HTML::Raw($match);
1029 class Markup_isohexchars extends SimpleMarkup {
1030 // hexnums, like ¤ <=> ¤
1031 var $_match_regexp = '\&\#x[0-9a-fA-F]{2,4};';
1033 function markup ($match) {
1034 return HTML::Raw($match);
1038 // FIXME: Do away with magic phpwiki forms. (Maybe phpwiki: links too?)
1039 // FIXME: Do away with plugin-links. They seem not to be used.
1042 class InlineTransformer
1044 var $_regexps = array();
1045 var $_markup = array();
1047 function InlineTransformer ($markup_types = false) {
1049 // We need to extend the inline parsers by certain actions, like SearchHighlight,
1050 // SpellCheck and maybe CreateToc.
1051 if (!$markup_types) {
1052 $non_default = false;
1053 $markup_types = array
1054 ('escape', 'wikicreolebracketlink', 'bracketlink', 'url',
1055 'interwiki', 'semanticlink', 'wikiword', 'linebreak',
1056 'wikicreole_italics', 'wikicreole_bold',
1057 'wikicreole_monospace', 'wikicreole_superscript',
1058 'wikicreole_subscript', 'old_emphasis', 'nestled_emphasis',
1059 'html_emphasis', 'html_abbr', 'plugin', 'plugin_wikicreole',
1060 'isonumchars', 'isohexchars', /*'html_entities'*/
1062 if (DISABLE_MARKUP_WIKIWORD)
1063 $markup_types = array_remove($markup_types, 'wikiword');
1065 $action = $request->getArg('action');
1066 if ($action == 'SpellCheck' and $request->getArg('suggestions'))
1067 { // insert it after url
1068 array_splice($markup_types, 2, 1, array('url','spellcheck'));
1070 if (isset($request->_searchhighlight))
1071 { // insert it after url
1072 array_splice($markup_types, 2, 1, array('url','searchhighlight'));
1073 //$request->setArg('searchhighlight', false);
1076 $non_default = true;
1078 foreach ($markup_types as $mtype) {
1079 $class = "Markup_$mtype";
1080 $this->_addMarkup(new $class);
1082 if (ENABLE_MARKUP_DIVSPAN and !$non_default)
1083 $this->_addMarkup(new Markup_html_divspan);
1084 if (ENABLE_MARKUP_COLOR and !$non_default)
1085 $this->_addMarkup(new Markup_color);
1086 // Markup_wikicreole_preformatted must be before Markup_template_plugin
1087 $this->_addMarkup(new Markup_wikicreole_preformatted);
1088 if (ENABLE_MARKUP_TEMPLATE and !$non_default)
1089 $this->_addMarkup(new Markup_template_plugin);
1090 if (ENABLE_MARKUP_MEDIAWIKI_TABLE)
1091 $this->_addMarkup(new Markup_mediawikitable_plugin);
1092 // This does not work yet
1093 if (0 and PLUGIN_MARKUP_MAP and !$non_default)
1094 $this->_addMarkup(new Markup_xml_plugin);
1097 function _addMarkup ($markup) {
1098 if (isa($markup, 'SimpleMarkup'))
1099 $regexp = $markup->getMatchRegexp();
1101 $regexp = $markup->getStartRegexp();
1103 assert( !isset($this->_markup[$regexp]) );
1104 assert( strlen(trim($regexp)) > 0 );
1105 $this->_regexps[] = $regexp;
1106 $this->_markup[] = $markup;
1109 function parse (&$text, $end_regexps = array('$')) {
1110 $regexps = $this->_regexps;
1112 // $end_re takes precedence: "favor reduce over shift"
1113 array_unshift($regexps, $end_regexps[0]);
1114 //array_push($regexps, $end_regexps[0]);
1115 $regexps = new RegexpSet($regexps);
1118 $output = new XmlContent;
1120 $match = $regexps->match($input);
1123 if ($match->regexp_ind == 0) {
1124 // No start pattern found before end pattern.
1126 if (isset($markup) and is_object($markup)
1127 and isa($markup,'Markup_plugin'))
1129 $current =& $output->_content[count($output->_content)-1];
1130 $current->setTightness(true,true);
1132 $output->pushContent($match->prematch);
1133 $text = $match->postmatch;
1137 $markup = $this->_markup[$match->regexp_ind - 1];
1138 $body = $this->_parse_markup_body($markup, $match->match,
1139 $match->postmatch, $end_regexps);
1141 // Couldn't match balanced expression.
1142 // Ignore and look for next matching start regexp.
1143 $match = $regexps->nextMatch($input, $match);
1147 // Matched markup. Eat input, push output.
1148 // FIXME: combine adjacent strings.
1149 if (isa($markup, 'SimpleMarkup'))
1150 $current = $markup->markup($match->match);
1152 $current = $markup->markup($match->match, $body);
1153 $input = $match->postmatch;
1154 if (isset($markup) and is_object($markup)
1155 and isa($markup,'Markup_plugin'))
1157 $current->setTightness(true,true);
1159 $output->pushContent($match->prematch, $current);
1161 $match = $regexps->match($input);
1164 // No pattern matched, not even the end pattern.
1169 function _parse_markup_body ($markup, $match, &$text, $end_regexps) {
1170 if (isa($markup, 'SimpleMarkup'))
1171 return true; // Done. SimpleMarkup is simple.
1173 if (!is_object($markup)) return false; // Some error: Should assert
1174 array_unshift($end_regexps, $markup->getEndRegexp($match));
1176 // Optimization: if no end pattern in text, we know the
1177 // parse will fail. This is an important optimization,
1178 // e.g. when text is "*lots *of *start *delims *with
1179 // *no *matching *end *delims".
1180 $ends_pat = "/(?:" . join(").*(?:", $end_regexps) . ")/xs";
1181 if (!preg_match($ends_pat, $text))
1183 return $this->parse($text, $end_regexps);
1187 class LinkTransformer extends InlineTransformer
1189 function LinkTransformer () {
1190 $this->InlineTransformer(array('escape', 'wikicreolebracketlink', 'bracketlink', 'url',
1191 'semanticlink', 'interwiki', 'wikiword',
1196 class NowikiTransformer extends InlineTransformer
1198 function NowikiTransformer () {
1199 $this->InlineTransformer
1201 'html_emphasis', 'html_abbr', 'plugin', 'plugin_wikicreole',
1202 'isonumchars', 'isohexchars', /*'html_entities',*/
1207 function TransformInline($text, $markup = 2.0, $basepage=false) {
1209 $action = $GLOBALS['request']->getArg('action');
1210 if (empty($trfm) or $action == 'SpellCheck') {
1211 $trfm = new InlineTransformer;
1214 if ($markup < 2.0) {
1215 $text = ConvertOldMarkup($text, 'inline');
1219 return new CacheableMarkup($trfm->parse($text), $basepage);
1221 return $trfm->parse($text);
1224 function TransformLinks($text, $markup = 2.0, $basepage = false) {
1228 $trfm = new LinkTransformer;
1231 if ($markup < 2.0) {
1232 $text = ConvertOldMarkup($text, 'links');
1236 return new CacheableMarkup($trfm->parse($text), $basepage);
1238 return $trfm->parse($text);
1242 * Transform only html markup and entities.
1244 function TransformInlineNowiki($text, $markup = 2.0, $basepage=false) {
1248 $trfm = new NowikiTransformer;
1251 return new CacheableMarkup($trfm->parse($text), $basepage);
1253 return $trfm->parse($text);
1256 // (c-file-style: "gnu")
1260 // c-basic-offset: 4
1261 // c-hanging-comment-ender-p: nil
1262 // indent-tabs-mode: nil