3 /* Copyright (C) 2002 Geoffrey T. Dairiki <dairiki@dairiki.org>
4 * Copyright (C) 2004-2008 Reini Urban
5 * Copyright (C) 2008 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_monospace 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_superscript 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_subscript extends BalancedMarkup
656 var $_start_regexp = ",,";
658 function getEndRegexp ($match) {
662 function markup ($match, $body) {
664 return new HtmlElement($tag, $body);
668 class Markup_old_emphasis extends BalancedMarkup
670 var $_start_regexp = "''|__";
672 function getEndRegexp ($match) {
676 function markup ($match, $body) {
677 $tag = $match == "''" ? 'em' : 'strong';
678 return new HtmlElement($tag, $body);
682 class Markup_nestled_emphasis extends BalancedMarkup
684 function getStartRegexp() {
685 static $start_regexp = false;
687 if (!$start_regexp) {
688 // The three possible delimiters
689 // (none of which can be followed by itself.)
694 $any = "(?: ${i}|${b}|${tt})"; // any of the three.
696 // Any of [_*=] is okay if preceded by space or one of [-"'/:]
697 $start[] = "(?<= \\s|^|[-\"'\\/:]) ${any}";
699 // _ or * is okay after = as long as not immediately followed by =
700 $start[] = "(?<= =) (?: ${i}|${b}) (?! =)";
702 $start[] = "(?<= _) (?: ${b}|${tt}) (?! _)";
703 $start[] = "(?<= \\*) (?: ${i}|${tt}) (?! \\*)";
706 // any delimiter okay after an opening brace ( [{<(] )
707 // as long as it's not immediately followed by the matching closing
709 $start[] = "(?<= { ) ${any} (?! } )";
710 $start[] = "(?<= < ) ${any} (?! > )";
711 $start[] = "(?<= \\( ) ${any} (?! \\) )";
713 $start = "(?:" . join('|', $start) . ")";
715 // Any of the above must be immediately followed by non-whitespace.
716 $start_regexp = $start . "(?= \S)";
719 return $start_regexp;
722 function getEndRegexp ($match) {
723 $chr = preg_quote($match);
724 return "(?<= \S | ^ ) (?<! $chr) $chr (?! $chr) (?= \s | [-)}>\"'\\/:.,;!? _*=] | $)";
727 function markup ($match, $body) {
729 case '*': return new HtmlElement('b', $body);
730 case '=': return new HtmlElement('tt', $body);
731 case '_': return new HtmlElement('i', $body);
736 class Markup_html_emphasis extends BalancedMarkup
739 "<(?: b|big|i|small|tt|em|strong|cite|code|dfn|kbd|samp|s|strike|del|var|sup|sub )>";
741 function getEndRegexp ($match) {
742 return "<\\/" . substr($match, 1);
745 function markup ($match, $body) {
746 $tag = substr($match, 1, -1);
747 return new HtmlElement($tag, $body);
751 class Markup_html_divspan extends BalancedMarkup
754 "<(?: div|span )(?: \s[^>]*)?>";
756 function getEndRegexp ($match) {
757 if (substr($match,1,4) == 'span')
761 return "<\\/" . $tag . '>';
764 function markup ($match, $body) {
765 if (substr($match,1,4) == 'span')
769 $rest = substr($match,1+strlen($tag),-1);
771 list($key,$val) = explode("=",$rest);
772 $args = array($key => $val);
773 } else $args = array();
774 return new HtmlElement($tag, $args, $body);
779 class Markup_html_abbr extends BalancedMarkup
781 //rurban: abbr|acronym need an optional title tag.
783 var $_start_regexp = "<(?: abbr|acronym )(?: \stitle=[^>]*)?>";
785 function getEndRegexp ($match) {
786 if (substr($match,1,4) == 'abbr')
790 return "<\\/" . $tag . '>';
793 function markup ($match, $body) {
794 if (substr($match,1,4) == 'abbr')
798 $rest = substr($match,1+strlen($tag),-1);
800 list($key,$val) = explode("=",$rest);
801 $args = array($key => $val);
802 } else $args = array();
803 return new HtmlElement($tag, $args, $body);
807 /** ENABLE_MARKUP_COLOR
808 * See http://www.pmwiki.org/wiki/PmWiki/WikiStyles and
809 * http://www.flexwiki.com/default.aspx/FlexWiki/FormattingRules.html
811 class Markup_color extends BalancedMarkup {
812 // %color=blue% blue text %% and back to normal
813 var $_start_regexp = "%color=(?: [^%]*)%";
814 var $_end_regexp = "%%";
816 function markup ($match, $body) {
817 $color = strtoupper(substr($match, 7, -1));
818 if (strlen($color) != 7
819 and in_array($color, array('RED', 'BLUE', 'GRAY', 'YELLOW', 'GREEN', 'CYAN', 'BLACK')))
820 { // must be a valid color name
821 return new HtmlElement('font', array('color' => $color), $body);
822 } elseif ((substr($color,0,1) == '#')
823 and (strspn(substr($color,1),'0123456789ABCDEF') == strlen($color)-1)) {
824 return new HtmlElement('font', array('color' => $color), $body);
826 trigger_error(sprintf(_("unknown color %s ignored"), substr($match, 7, -1)), E_USER_WARNING);
832 // Special version for single-line plugins formatting,
833 // like: '<small>< ?plugin PopularNearby ? ></small>'
834 class Markup_plugin extends SimpleMarkup
836 var $_match_regexp = '<\?plugin(?:-form)?\s[^\n]+?\?>';
838 function markup ($match) {
839 //$xml = new Cached_PluginInvocation($match);
840 //$xml->setTightness(true,true);
841 return new Cached_PluginInvocation($match);
845 // Special version for plugins in xml syntax
846 // <name arg=value>body</name> or <name /> => < ? plugin pluginname arg=value body ? >
847 // PLUGIN_MARKUP_MAP = "html:RawHtml dot:GraphViz toc:CreateToc amath:AsciiMath richtable:RichTable include:IncludePage tex:TexToPng"
848 class Markup_xml_plugin extends BalancedMarkup
850 //var $_start_regexp = "<(?: ".join('|',PLUGIN_MARKUP_MAP)." )(?: \s[^>]*)>";
852 function getStartRegexp () {
853 global $PLUGIN_MARKUP_MAP;
854 static $_start_regexp;
855 if ($_start_regexp) return $_start_regexp;
856 if (empty($PLUGIN_MARKUP_MAP))
858 //"<(?: html|dot|toc|amath|richtable|include|tex )(?: \s[^>]*)>"
859 $_start_regexp = "<(?: ".join('|',array_keys($PLUGIN_MARKUP_MAP))." )(?: \s[^>]* | / )>";
860 return $_start_regexp;
862 function getEndRegexp ($match) {
863 return "<\\/" . $match . '>';
865 function markup ($match, $body) {
866 global $PLUGIN_MARKUP_MAP;
867 $name = substr($match,2,-2);
869 if (preg_match('/^(\S+)\|(.*)$/', $name, $_m)) {
871 $vars = $_m[2]; //str_replace(' ', '&', $_m[2]);
873 if (!isset($PLUGIN_MARKUP_MAP[$name])) {
874 trigger_error("No plugin for $name $vars defined.", E_USER_WARNING);
877 $plugin = $PLUGIN_MARKUP_MAP[$name];
878 return new Cached_PluginInvocation("<"."?plugin $plugin $vars $body ?".">");
883 * Wikicreole preformatted
887 class Markup_wikicreole_preformatted extends SimpleMarkup
889 var $_match_regexp = '\{\{\{.*?\}\}\}';
891 function markup ($match) {
892 // Remove {{{ and }}}
893 return new HtmlElement('pre', substr($match, 3, -3));
897 /** ENABLE_MARKUP_TEMPLATE
898 * Template syntax similar to Mediawiki
900 * => < ? plugin Template page=template ? >
901 * {{template|var1=value1|var2=value|...}}
902 * => < ? plugin Template page=template var=value ... ? >
904 class Markup_template_plugin extends SimpleMarkup
906 // patch #1732793: allow \n, mult. {{ }} in one line, and single letters
907 var $_match_regexp = '\{\{.*?\}\}';
909 function markup ($match) {
911 $page = substr($match,2,-2);
912 if (strpos($page, "|") === false) {
916 $imagename = substr($page, 0, strpos($page, "|"));
917 $alt = ltrim(strstr($page, "|"), "|");
920 // It's not a Mediawiki template, it's a Wikicreole image
921 if ((string_ends_with($imagename, ".jpg"))
922 or (string_ends_with($imagename, ".jpeg"))
923 or (string_ends_with($imagename, ".gif"))
924 or (string_ends_with($imagename, ".png"))) {
925 return LinkImage(UPLOAD_DATA_PATH . $imagename, $alt);
928 $page = str_replace("\n", "", $page);
931 if (preg_match('/^(\S+?)\|(.*)$/', $page, $_m)) {
933 $vars = '"' . preg_replace('/\|/', '" "', $_m[2]) . '"';
934 $vars = preg_replace('/"(\S+)=([^"]*)"/', '\\1="\\2"', $vars);
937 // page may contain a version number
939 // in that case, output is "page=foo rev=5"
940 if (strstr($page, "?")) {
941 $page = str_replace("?version=", "\" rev=\"", $page);
945 $s = '<'.'?plugin Template page="'.$page.'" '.$vars.' ?'.'>';
947 $s = '<'.'?plugin Template page="' . $page . '" ?'.'>';
948 return new Cached_PluginInvocation($s);
952 /** ENABLE_MARKUP_MEDIAWIKI_TABLE
953 * Table syntax similar to Mediawiki
955 * => <?plugin MediawikiTable
959 class Markup_mediawikitable_plugin extends SimpleMarkup
961 var $_match_regexp = '\{\|.*?\|\}';
963 function markup ($match) {
964 $s = '<'.'?plugin MediawikiTable ' . $match . '?'.'>';
965 return new Cached_PluginInvocation($s);
969 class Markup_wikicreoletable_plugin extends SimpleMarkup
971 var $_match_regexp = '^\|=.*?\?>';
973 function markup ($match) {
974 $s = '<'.'?plugin WikicreoleTable ' . $match . '?'.'>';
975 return new Cached_PluginInvocation($s);
979 // "..." => "…" browser specific display (not cached?)
980 // Support some HTML::Entities: (C) for copy, --- for mdash, -- for ndash
981 // TODO: "--" => "&emdash;" browser specific display (not cached?)
983 class Markup_html_entities extends SimpleMarkup {
984 //var $_match_regexp = '(: \.\.\.|\-\-|\-\-\-|\(C\) )';
986 function Markup_html_entities() {
987 $this->_entities = array('...' => '…',
991 '©' => '©',
992 '™' => '™',
994 $this->_match_regexp =
996 join('|', array_map('preg_quote', array_keys($this->_entities))) .
1000 function markup ($match) {
1001 return HTML::Raw($this->_entities[$match]);
1005 class Markup_isonumchars extends SimpleMarkup {
1006 var $_match_regexp = '\&\#\d{2,5};';
1008 function markup ($match) {
1009 return HTML::Raw($match);
1013 class Markup_isohexchars extends SimpleMarkup {
1014 // hexnums, like ¤ <=> ¤
1015 var $_match_regexp = '\&\#x[0-9a-fA-F]{2,4};';
1017 function markup ($match) {
1018 return HTML::Raw($match);
1022 // FIXME: Do away with magic phpwiki forms. (Maybe phpwiki: links too?)
1023 // FIXME: Do away with plugin-links. They seem not to be used.
1026 class InlineTransformer
1028 var $_regexps = array();
1029 var $_markup = array();
1031 function InlineTransformer ($markup_types = false) {
1033 // We need to extend the inline parsers by certain actions, like SearchHighlight,
1034 // SpellCheck and maybe CreateToc.
1035 if (!$markup_types) {
1036 $non_default = false;
1037 $markup_types = array
1038 ('escape', 'wikicreolebracketlink', 'bracketlink', 'url',
1039 'interwiki', 'semanticlink', 'wikiword', 'linebreak',
1040 'wikicreole_italics',
1041 'wikicreole_monospace', 'wikicreole_superscript',
1042 'wikicreole_subscript', 'old_emphasis', 'nestled_emphasis',
1043 'html_emphasis', 'html_abbr', 'plugin',
1044 'isonumchars', 'isohexchars', /*'html_entities'*/
1046 if (DISABLE_MARKUP_WIKIWORD)
1047 $markup_types = array_remove($markup_types, 'wikiword');
1049 $action = $request->getArg('action');
1050 if ($action == 'SpellCheck' and $request->getArg('suggestions'))
1051 { // insert it after url
1052 array_splice($markup_types, 2, 1, array('url','spellcheck'));
1054 if (isset($request->_searchhighlight))
1055 { // insert it after url
1056 array_splice($markup_types, 2, 1, array('url','searchhighlight'));
1057 //$request->setArg('searchhighlight', false);
1060 $non_default = true;
1062 foreach ($markup_types as $mtype) {
1063 $class = "Markup_$mtype";
1064 $this->_addMarkup(new $class);
1066 if (ENABLE_MARKUP_DIVSPAN and !$non_default)
1067 $this->_addMarkup(new Markup_html_divspan);
1068 if (ENABLE_MARKUP_COLOR and !$non_default)
1069 $this->_addMarkup(new Markup_color);
1070 $this->_addMarkup(new Markup_wikicreoletable_plugin);
1071 // Markup_wikicreole_preformatted must be before Markup_template_plugin
1072 $this->_addMarkup(new Markup_wikicreole_preformatted);
1073 if (ENABLE_MARKUP_TEMPLATE and !$non_default)
1074 $this->_addMarkup(new Markup_template_plugin);
1075 if (ENABLE_MARKUP_MEDIAWIKI_TABLE)
1076 $this->_addMarkup(new Markup_mediawikitable_plugin);
1077 // This does not work yet
1078 if (0 and PLUGIN_MARKUP_MAP and !$non_default)
1079 $this->_addMarkup(new Markup_xml_plugin);
1082 function _addMarkup ($markup) {
1083 if (isa($markup, 'SimpleMarkup'))
1084 $regexp = $markup->getMatchRegexp();
1086 $regexp = $markup->getStartRegexp();
1088 assert( !isset($this->_markup[$regexp]) );
1089 assert( strlen(trim($regexp)) > 0 );
1090 $this->_regexps[] = $regexp;
1091 $this->_markup[] = $markup;
1094 function parse (&$text, $end_regexps = array('$')) {
1095 $regexps = $this->_regexps;
1097 // $end_re takes precedence: "favor reduce over shift"
1098 array_unshift($regexps, $end_regexps[0]);
1099 //array_push($regexps, $end_regexps[0]);
1100 $regexps = new RegexpSet($regexps);
1103 $output = new XmlContent;
1105 $match = $regexps->match($input);
1108 if ($match->regexp_ind == 0) {
1109 // No start pattern found before end pattern.
1111 if (isset($markup) and is_object($markup)
1112 and isa($markup,'Markup_plugin'))
1114 $current =& $output->_content[count($output->_content)-1];
1115 $current->setTightness(true,true);
1117 $output->pushContent($match->prematch);
1118 $text = $match->postmatch;
1122 $markup = $this->_markup[$match->regexp_ind - 1];
1123 $body = $this->_parse_markup_body($markup, $match->match,
1124 $match->postmatch, $end_regexps);
1126 // Couldn't match balanced expression.
1127 // Ignore and look for next matching start regexp.
1128 $match = $regexps->nextMatch($input, $match);
1132 // Matched markup. Eat input, push output.
1133 // FIXME: combine adjacent strings.
1134 if (isa($markup, 'SimpleMarkup'))
1135 $current = $markup->markup($match->match);
1137 $current = $markup->markup($match->match, $body);
1138 $input = $match->postmatch;
1139 if (isset($markup) and is_object($markup)
1140 and isa($markup,'Markup_plugin'))
1142 $current->setTightness(true,true);
1144 $output->pushContent($match->prematch, $current);
1146 $match = $regexps->match($input);
1149 // No pattern matched, not even the end pattern.
1154 function _parse_markup_body ($markup, $match, &$text, $end_regexps) {
1155 if (isa($markup, 'SimpleMarkup'))
1156 return true; // Done. SimpleMarkup is simple.
1158 if (!is_object($markup)) return false; // Some error: Should assert
1159 array_unshift($end_regexps, $markup->getEndRegexp($match));
1161 // Optimization: if no end pattern in text, we know the
1162 // parse will fail. This is an important optimization,
1163 // e.g. when text is "*lots *of *start *delims *with
1164 // *no *matching *end *delims".
1165 $ends_pat = "/(?:" . join(").*(?:", $end_regexps) . ")/xs";
1166 if (!preg_match($ends_pat, $text))
1168 return $this->parse($text, $end_regexps);
1172 class LinkTransformer extends InlineTransformer
1174 function LinkTransformer () {
1175 $this->InlineTransformer(array('escape', 'wikicreolebracketlink', 'bracketlink', 'url',
1176 'semanticlink', 'interwiki', 'wikiword',
1181 class NowikiTransformer extends InlineTransformer
1183 function NowikiTransformer () {
1184 $this->InlineTransformer
1186 'html_emphasis', 'html_abbr', 'plugin',
1187 'isonumchars', 'isohexchars', /*'html_entities',*/
1192 function TransformInline($text, $markup = 2.0, $basepage=false) {
1194 $action = $GLOBALS['request']->getArg('action');
1195 if (empty($trfm) or $action == 'SpellCheck') {
1196 $trfm = new InlineTransformer;
1199 if ($markup < 2.0) {
1200 $text = ConvertOldMarkup($text, 'inline');
1204 return new CacheableMarkup($trfm->parse($text), $basepage);
1206 return $trfm->parse($text);
1209 function TransformLinks($text, $markup = 2.0, $basepage = false) {
1213 $trfm = new LinkTransformer;
1216 if ($markup < 2.0) {
1217 $text = ConvertOldMarkup($text, 'links');
1221 return new CacheableMarkup($trfm->parse($text), $basepage);
1223 return $trfm->parse($text);
1227 * Transform only html markup and entities.
1229 function TransformInlineNowiki($text, $markup = 2.0, $basepage=false) {
1233 $trfm = new NowikiTransformer;
1236 return new CacheableMarkup($trfm->parse($text), $basepage);
1238 return $trfm->parse($text);
1241 // (c-file-style: "gnu")
1245 // c-basic-offset: 4
1246 // c-hanging-comment-ender-p: nil
1247 // indent-tabs-mode: nil