1 <?php rcs_id('$Id: InlineParser.php,v 1.27 2003-02-26 00:39:30 dairiki Exp $');
2 /* Copyright (C) 2002, Geoffrey T. Dairiki <dairiki@dairiki.org>
4 * This file is part of PhpWiki.
6 * PhpWiki is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
11 * PhpWiki is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License
17 * along with PhpWiki; if not, write to the Free Software
18 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
21 * This is the code which deals with the inline part of the (new-style)
25 * @author Geoffrey T. Dairiki
31 * This is the character used in wiki markup to escape characters with
34 define('ESCAPE_CHAR', '~');
36 require_once('lib/HtmlElement.php');
37 require_once('lib/CachedMarkup.php');
38 require_once('lib/interwiki.php');
39 require_once('lib/stdlib.php');
42 function WikiEscape($text) {
43 return str_replace('#', ESCAPE_CHAR . '#', $text);
46 function UnWikiEscape($text) {
47 return preg_replace('/' . ESCAPE_CHAR . '(.)/', '\1', $text);
51 * Return type from RegexpSet::match and RegexpSet::nextMatch.
55 class RegexpSet_match {
57 * The text leading up the the next match.
67 * The text following the matched text.
72 * Index of the regular expression which matched.
78 * A set of regular expressions.
80 * This class is probably only useful for InlineTransformer.
86 * @param array $regexps A list of regular expressions. The
87 * regular expressions should not include any sub-pattern groups
88 * "(...)". (Anonymous groups, like "(?:...)", as well as
89 * look-ahead and look-behind assertions are okay.)
91 function RegexpSet ($regexps) {
92 $this->_regexps = $regexps;
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, $repeat);
143 function _match ($text, $regexps, $repeat) {
144 $pat= "/ ( . $repeat ) ( (" . join(')|(', $regexps) . ") ) /Axs";
146 if (! preg_match($pat, $text, $m)) {
150 $match = new RegexpSet_match;
151 $match->postmatch = substr($text, strlen($m[0]));
152 $match->prematch = $m[1];
153 $match->match = $m[2];
154 $match->regexp_ind = count($m) - 4;
157 PrintXML(HTML::dl(HTML::dt("input"),
158 HTML::dd(HTML::pre($text)),
160 HTML::dd(HTML::pre($match->match)),
162 HTML::dd(HTML::pre($regexps[$match->regexp_ind])),
163 HTML::dt("prematch"),
164 HTML::dd(HTML::pre($match->prematch))));
173 * A simple markup rule (i.e. terminal token).
175 * These are defined by a regexp.
177 * When a match is found for the regexp, the matching text is replaced.
178 * The replacement content is obtained by calling the SimpleMarkup::markup method.
186 * @return string Regexp which matches this token.
188 function getMatchRegexp () {
189 return $this->_match_regexp;
192 /** Markup matching text.
194 * @param string $match The text which matched the regexp
195 * (obtained from getMatchRegexp).
197 * @return mixed The expansion of the matched text.
199 function markup ($match /*, $body */) {
200 trigger_error("pure virtual", E_USER_ERROR);
205 * A balanced markup rule.
207 * These are defined by a start regexp, and and end regexp.
213 /** Get the starting regexp for this rule.
215 * @return string The starting regexp.
217 function getStartRegexp () {
218 return $this->_start_regexp;
221 /** Get the ending regexp for this rule.
223 * @param string $match The text which matched the starting regexp.
225 * @return string The ending regexp.
227 function getEndRegexp ($match) {
228 return $this->_end_regexp;
231 /** Get expansion for matching input.
233 * @param string $match The text which matched the starting regexp.
235 * @param mixed $body Transformed text found between the starting
236 * and ending regexps.
238 * @return mixed The expansion of the matched text.
240 function markup ($match, $body) {
241 trigger_error("pure virtual", E_USER_ERROR);
245 class Markup_escape extends SimpleMarkup
247 function getMatchRegexp () {
248 return ESCAPE_CHAR . ".";
251 function markup ($match) {
252 assert(strlen($match) == 2);
257 function LinkBracketLink($bracketlink) {
258 global $request, $AllowedProtocols, $InlineImages;
260 include_once("lib/interwiki.php");
261 $intermap = InterWikiMap::GetMap($request);
263 // $bracketlink will start and end with brackets; in between will
264 // be either a page name, a URL or both separated by a pipe.
266 // strip brackets and leading space
267 preg_match('/(\#?) \[\s* (?: (.*?) \s* (?<!' . ESCAPE_CHAR . ')(\|) )? \s* (.+?) \s*\]/x',
268 $bracketlink, $matches);
269 list (, $hash, $label, $bar, $rawlink) = $matches;
271 $label = UnWikiEscape($label);
272 $link = UnWikiEscape($rawlink);
274 // if label looks like a url to an image, we want an image link.
275 if (preg_match("/\\.($InlineImages)$/i", $label)) {
277 if (! preg_match("#^($AllowedProtocols):#", $imgurl)) {
278 // linkname like 'images/next.gif'.
280 $imgurl = $Theme->getImageURL($linkname);
282 $label = LinkImage($imgurl, $link);
286 // It's an anchor, not a link...
287 $id = MangleXmlIdentifier($link);
288 return HTML::a(array('name' => $id, 'id' => $id),
289 $bar ? $label : $link);
292 if (preg_match("#^($AllowedProtocols):#", $link)) {
293 // if it's an image, embed it; otherwise, it's a regular link
294 if (preg_match("/\\.($InlineImages)$/i", $link))
295 // no image link, just the src. see [img|link] above
296 return LinkImage($link, $label);
298 return new Cached_ExternalLink($link, $label);
300 elseif (preg_match("/^phpwiki:/", $link))
301 return new Cached_PhpwikiURL($link, $label);
302 elseif (preg_match("/^" . $intermap->getRegexp() . ":/", $link))
303 return new Cached_InterwikiLink($link, $label);
305 // Split anchor off end of pagename.
306 if (preg_match('/\A(.*)(?<!'.ESCAPE_CHAR.')#(.*?)\Z/', $rawlink, $m)) {
307 list(,$rawlink,$anchor) = $m;
308 $pagename = UnWikiEscape($rawlink);
309 $anchor = UnWikiEscape($anchor);
317 return new Cached_WikiLink($pagename, $label, $anchor);
321 class Markup_bracketlink extends SimpleMarkup
323 var $_match_regexp = "\\#? \\[ .*? [^]\\s] .*? \\]";
325 function markup ($match) {
326 $link = LinkBracketLink($match);
327 assert($link->isInlineElement());
332 class Markup_url extends SimpleMarkup
334 function getMatchRegexp () {
335 global $AllowedProtocols;
336 return "(?<![[:alnum:]]) (?:$AllowedProtocols) : [^\s<>\"']+ (?<![ ,.?; \] \) ])";
339 function markup ($match) {
340 return new Cached_ExternalLink(UnWikiEscape($match));
345 class Markup_interwiki extends SimpleMarkup
347 function getMatchRegexp () {
349 $map = InterWikiMap::GetMap($request);
350 return "(?<! [[:alnum:]])" . $map->getRegexp(). ": \S+ (?<![ ,.?;! \] \) \" \' ])";
353 function markup ($match) {
355 $map = InterWikiMap::GetMap($request);
356 return new Cached_InterwikiLink(UnWikiEscape($match));
360 class Markup_wikiword extends SimpleMarkup
362 function getMatchRegexp () {
363 global $WikiNameRegexp;
364 return " $WikiNameRegexp";
367 function markup ($match) {
368 return new Cached_WikiLink($match);
372 class Markup_linebreak extends SimpleMarkup
374 var $_match_regexp = "(?: (?<! %) %%% (?! %) | <(?:br|BR)> )";
381 class Markup_old_emphasis extends BalancedMarkup
383 var $_start_regexp = "''|__";
385 function getEndRegexp ($match) {
389 function markup ($match, $body) {
390 $tag = $match == "''" ? 'em' : 'strong';
391 return new HtmlElement($tag, $body);
395 class Markup_nestled_emphasis extends BalancedMarkup
397 function getStartRegexp() {
398 static $start_regexp = false;
400 if (!$start_regexp) {
401 // The three possible delimiters
402 // (none of which can be followed by itself.)
407 $any = "(?: ${i}|${b}|${tt})"; // any of the three.
409 // Any of [_*=] is okay if preceded by space or one of [-"'/:]
410 $start[] = "(?<= \\s|^|[-\"'\\/:]) ${any}";
412 // _ or * is okay after = as long as not immediately followed by =
413 $start[] = "(?<= =) (?: ${i}|${b}) (?! =)";
415 $start[] = "(?<= _) (?: ${b}|${tt}) (?! _)";
416 $start[] = "(?<= \\*) (?: ${i}|${tt}) (?! \\*)";
419 // any delimiter okay after an opening brace ( [{<(] )
420 // as long as it's not immediately followed by the matching closing
422 $start[] = "(?<= { ) ${any} (?! } )";
423 $start[] = "(?<= < ) ${any} (?! > )";
424 $start[] = "(?<= \\( ) ${any} (?! \\) )";
426 $start = "(?:" . join('|', $start) . ")";
428 // Any of the above must be immediately followed by non-whitespace.
429 $start_regexp = $start . "(?= \S)";
432 return $start_regexp;
435 function getEndRegexp ($match) {
436 $chr = preg_quote($match);
437 return "(?<= \S | ^ ) (?<! $chr) $chr (?! $chr) (?= \s | [-)}>\"'\\/:.,;!? _*=] | $)";
440 function markup ($match, $body) {
442 case '*': return new HtmlElement('b', $body);
443 case '=': return new HtmlElement('tt', $body);
444 case '_': return new HtmlElement('i', $body);
449 class Markup_html_emphasis extends BalancedMarkup
451 var $_start_regexp = "<(?: b|big|i|small|tt|
453 abbr|acronym|cite|code|dfn|kbd|samp|var|
456 function getEndRegexp ($match) {
457 return "<\\/" . substr($match, 1);
460 function markup ($match, $body) {
461 $tag = substr($match, 1, -1);
462 return new HtmlElement($tag, $body);
466 // FIXME: Do away with magic phpwiki forms. (Maybe phpwiki: links too?)
467 // FIXME: Do away with plugin-links. They seem not to be used.
471 class InlineTransformer
473 var $_regexps = array();
474 var $_markup = array();
476 function InlineTransformer ($markup_types = false) {
478 $markup_types = array('escape', 'bracketlink', 'url',
479 'interwiki', 'wikiword', 'linebreak',
480 'old_emphasis', 'nestled_emphasis',
483 foreach ($markup_types as $mtype) {
484 $class = "Markup_$mtype";
485 $this->_addMarkup(new $class);
489 function _addMarkup ($markup) {
490 if (isa($markup, 'SimpleMarkup'))
491 $regexp = $markup->getMatchRegexp();
493 $regexp = $markup->getStartRegexp();
495 assert(!isset($this->_markup[$regexp]));
496 $this->_regexps[] = $regexp;
497 $this->_markup[] = $markup;
500 function parse (&$text, $end_regexps = array('$')) {
501 $regexps = $this->_regexps;
503 // $end_re takes precedence: "favor reduce over shift"
504 array_unshift($regexps, $end_regexps[0]);
505 $regexps = new RegexpSet($regexps);
508 $output = new XmlContent;
510 $match = $regexps->match($input);
513 if ($match->regexp_ind == 0) {
514 // No start pattern found before end pattern.
516 $output->pushContent($match->prematch);
517 $text = $match->postmatch;
521 $markup = $this->_markup[$match->regexp_ind - 1];
522 $body = $this->_parse_markup_body($markup, $match->match, $match->postmatch, $end_regexps);
524 // Couldn't match balanced expression.
525 // Ignore and look for next matching start regexp.
526 $match = $regexps->nextMatch($input, $match);
530 // Matched markup. Eat input, push output.
531 // FIXME: combine adjacent strings.
532 $input = $match->postmatch;
533 $output->pushContent($match->prematch,
534 $markup->markup($match->match, $body));
536 $match = $regexps->match($input);
539 // No pattern matched, not even the end pattern.
544 function _parse_markup_body ($markup, $match, &$text, $end_regexps) {
545 if (isa($markup, 'SimpleMarkup'))
546 return true; // Done. SimpleMarkup is simple.
548 array_unshift($end_regexps, $markup->getEndRegexp($match));
549 // Optimization: if no end pattern in text, we know the
550 // parse will fail. This is an important optimization,
551 // e.g. when text is "*lots *of *start *delims *with
552 // *no *matching *end *delims".
553 $ends_pat = "/(?:" . join(").*(?:", $end_regexps) . ")/xs";
554 if (!preg_match($ends_pat, $text))
556 return $this->parse($text, $end_regexps);
560 class LinkTransformer extends InlineTransformer
562 function LinkTransformer () {
563 $this->InlineTransformer(array('escape', 'bracketlink', 'url',
564 'interwiki', 'wikiword'));
568 function TransformInline($text, $markup = 2.0, $basepage=false) {
572 $trfm = new InlineTransformer;
576 $text = ConvertOldMarkup($text, 'inline');
580 return new CacheableMarkup($trfm->parse($text), $basepage);
582 return $trfm->parse($text);
585 function TransformLinks($text, $markup = 2.0, $basepage = false) {
589 $trfm = new LinkTransformer;
593 $text = ConvertOldMarkup($text, 'links');
597 return new CacheableMarkup($trfm->parse($text), $basepage);
599 return $trfm->parse($text);
602 // (c-file-style: "gnu")
607 // c-hanging-comment-ender-p: nil
608 // indent-tabs-mode: nil