Context Navigation

wakka.php @ 761

Visit:

Revision 761, 30.2 KB (checked in by DarTar, 6 years ago)
Replacing CSS selector for data tables (`.data` instead of `.wikka`) (missed one occurrence), refs #380 and #230
Property svn:keywords set to `Id`

Line
1	<?php
2	/**
3	* Wikka Formatting Engine
4	*
5	* This is the main formatting engine used by Wikka to parse wiki markup and render valid XHTML.
6	*
7	* @package Formatters
8	* @version $Id$
9	* @license http://www.gnu.org/copyleft/gpl.html GNU General Public License
10	* @filesource
11	*
12	* @author {@link http://wikkawiki.org/JsnX Jason Tourtelotte}
13	* @author {@link http://wikkawiki.org/DotMG Mahefa Randimbisoa}
14	* @author {@link http://wikkawiki.org/JavaWoman Marjolein Katsma}
15	* @author {@link http://wikkawiki.org/NilsLindenberg Nils Lindenberg} (code cleanup)
16	* @author {@link http://wikkawiki.org/DarTar Dario Taraborelli} (grab handler and filename support for codeblocks)
17	* @author {@link http://wikkawiki.org/TormodHaugen Tormod Haugen} (table formatter support)
18	*
19	* @uses Wakka::htmlspecialchars_ent()
20	*
21	* @todo add support for formatter plugins;
22	* @todo use a central RegEx library #34;
23	* @todo add further improvements from ImprovedFormatter
24	*/
25
26	/**#@+
27	* Code block pattern.
28	*/
29	if (!defined('PATTERN_OPEN_BRACKET')) define('PATTERN_OPEN_BRACKET', '\(');
30	if (!defined('PATTERN_FORMATTER')) define('PATTERN_FORMATTER', '([^;\)]+)');
31	if (!defined('PATTERN_LINE_NUMBER')) define('PATTERN_LINE_NUMBER', '(;(\d*?))?');
32	if (!defined('PATTERN_FILENAME')) define('PATTERN_FILENAME', '(;([^\)\x01-\x1f\\?\"<>\\|])([^\)]*))?');
33	if (!defined('PATTERN_CLOSE_BRACKET')) define('PATTERN_CLOSE_BRACKET', '\)');
34	if (!defined('PATTERN_CODE')) define('PATTERN_CODE', '(.*)');
35	/*#@-/
36
37	/**
38	* Match heading tags.
39	*
40	* - $result[0] : the entire node representation, including the closing tag
41	* - $result[1] : the nodename (h1, h2, .. , h6)
42	* - $result[2] : the heading attribute, ie all the strings after the tagname and before the first ">" character
43	* - $result[3] : the content of the heading tag, just like the innerHTML method in DOM.
44	* This pattern will match only if the text it is applied to is valid XHTML: it should use lowercase in the tagName,
45	* it should not contain the character ">" inside attributes.
46	*/
47	if (!defined('PATTERN_MATCH_HEADINGS')) define('PATTERN_MATCH_HEADINGS', '#^<(h[1-6])(.?)>(.?)</\\1>$#s');
48	/**
49	* Match id in attributes.
50	*
51	* - $result[0] : a string like <code>id="h1_id"</code>, starting with the letters id=, and followed by a string
52	* enclosed in either single or double quote. It doesn't match if the term id is not preceded by any whitespace.
53	* - $result[1] : The single character used to enclose the string, either a single or a double quote.
54	* - $result[2] : The content of the string, ie the value of the id attribute.
55	* The RE uses a backref to match both single and double enclosing quotes.
56	*/
57	if (!defined('PATTERN_MATCH_ID_ATTRIBUTES')) define('PATTERN_MATCH_ID_ATTRIBUTES', '/(?<=\\s)id=("\|\')(.*?)\\1/');
58	/**
59	* The string $format_option is a semicolon separated list of strings, including the word `page'
60	*/
61	if (!defined('PATTERN_MATCH_PAGE_FORMATOPTION')) define('PATTERN_MATCH_PAGE_FORMATOPTION', '/(^\|;)page(;\|$)/');
62
63	// @@@ is this condition handy? would prevent generating IDs on a page fragment
64	// - unless formatter is called on those with an explicit foprmat option!
65	if (isset($format_option) && preg_match(PATTERN_MATCH_PAGE_FORMATOPTION, $format_option))
66	{
67	if (!function_exists('wakka3callback'))
68	{
69	/**
70	* "Afterburner" formatting: extra handling of already-generated XHTML code.
71	*
72	* 1. For headings:
73	* a) Use heading to derive a document title
74	* b) Ensure every heading has an id, either specified or generated. (May be
75	* extended to generate section TOC data.)
76	* If an id is already specified, that is used without any modification.
77	* If no id is specified, it is generated on the basis of the heading context:
78	* - any image tag is replaced by its alt text (if specified)
79	* - all tags are stripped
80	* - all characters that are not valid in an ID are stripped (except whitespace)
81	* - the resulting string is then used by makedId() to generate an id out of it
82	*
83	* @access private
84	* @uses Wakka::HasPageTitle()
85	* @uses Wakka::SetPageTitle()
86	* @uses Wakka::CleanTextNode()
87	* @uses Wakka::makeId()
88	*
89	* @param array $things required: matches of the regex in the preg_replace_callback
90	* @return string heading with an id attribute
91	*/
92	function wakka3callback($things)
93	{
94	global $wakka;
95	$thing = $things[1];
96
97	// heading
98	if (preg_match(PATTERN_MATCH_HEADINGS, $thing, $matches))
99	{
100	list($h_element, $h_tagname, $h_attribs, $h_heading) = $matches;
101	// @@@ apply nodeToTextOnly() on $h_heading so stored title is always valid
102	if ((!$wakka->HasPageTitle()) && ('h5' > $h_tagname))
103	{
104	$wakka->SetPageTitle($h_heading);
105	}
106
107	if (preg_match(PATTERN_MATCH_ID_ATTRIBUTES, $h_attribs))
108	{
109	// existing id attribute: nothing to do (assume already treated as embedded code)
110	// @@@ we may want to gather ids and heading text for a TOC here ...
111	// heading text should then get partly the same treatment as when we're creating ids:
112	// at least replace images and strip tags - we can leave entities etc. alone - so we end up with
113	// plain text-only
114	// do this if we have a condition set to generate a TOC
115	return $h_element;
116	}
117	else
118	{
119	// no id: we'll have to create one
120	$headingtext = $wakka->CleanTextNode($h_heading); // @@@ replace with headingToTextOnly()
121	// now create id based on resulting heading text
122	$h_id = $wakka->makeId('hn', $headingtext);
123
124	// rebuild element, adding id
125	return '<'.$h_tagname.$h_attribs.' id="'.$h_id.'">'.$h_heading.'</'.$h_tagname.'>';
126	}
127	}
128	// other elements to be treated go here (tables, images, code sections...)
129	}
130	}
131	}
132	// Note: all possible formatting tags have to be in a single regular expression for this to work correctly.
133
134	if (!function_exists("wakka2callback")) # DotMG [many lines] : Unclosed tags fix!
135	# JW: NOT a complete fix!
136	# see http://wikkawiki.org/ImprovedFormatter#hn_Not_a_complete_solution
137	{
138	function wakka2callback($things)
139	{
140	$thing = $things[0];
141	$result='';
142
143	static $oldIndentLevel = 0;
144	static $oldIndentLength= 0;
145	static $indentClosers = array();
146	static $newIndentSpace= array();
147	static $br = 1;
148	static $trigger_table = 0;
149	static $trigger_rowgroup = 0;
150	static $trigger_colgroup = 0;
151	static $trigger_bold = 0;
152	static $trigger_italic = 0;
153	static $trigger_underline = 0;
154	static $trigger_monospace = 0;
155	static $trigger_notes = 0;
156	static $trigger_strike = 0;
157	static $trigger_inserted = 0;
158	static $trigger_deleted = 0;
159	static $trigger_floatl = 0;
160	static $trigger_floatr = 0;
161	static $trigger_keys = 0;
162	static $trigger_strike = 0;
163	static $trigger_inserted = 0;
164	static $trigger_center = 0;
165	static $trigger_l = array(-1, 0, 0, 0, 0, 0);
166	static $output = '';
167	static $valid_filename = '';
168	static $invalid = '';
169	static $curIndentType;
170
171	global $wakka;
172
173	// @@@ inline elements should be closed before block-level elements
174	// (see ImprovedFormatter solution again)
175	// TEST: are indents closed at end of page now??? (see... again)
176	// @@@ <kbd> is missing
177	if ((!is_array($things)) && ($things == 'closetags'))
178	{
179	$result = '';
180	if (3 < $trigger_table){
181	$result .= '</caption>';
182	}
183	elseif (2 < $trigger_table)
184	{
185	$result .= '</th></tr>';
186	}
187	elseif (1 < $trigger_table)
188	{
189	$result .= '</td></tr>';
190	}
191	if (2 < $trigger_rowgroup)
192	{
193	$result .= '</tbody>';
194	}
195	elseif (1 < $trigger_rowgroup)
196	{
197	$result .= '</tfoot>';
198	}
199	elseif (0 < $trigger_rowgroup)
200	{
201	$result .= '</thead>';
202	}
203	if (0 < $trigger_table)
204	{
205	$result .= '</table>';
206	}
207
208	if ($trigger_strike % 2)
209	{
210	$result .= '</span>';
211	}
212	if ($trigger_notes % 2)
213	{
214	$result .= '</span>';
215	}
216	if ($trigger_inserted % 2)
217	{
218	$result .= '</span>';
219	}
220	if ($trigger_underline % 2)
221	{
222	$result .= '</span>';
223	}
224	if ($trigger_floatl % 2)
225	{
226	$result .= '</div>';
227	}
228	if ($trigger_floatr % 2)
229	{
230	$result .= '</div>';
231	}
232	if ($trigger_center % 2)
233	{
234	$result .= '</div>';
235	}
236	if ($trigger_italic % 2)
237	{
238	$result .= '</em>';
239	}
240	if ($trigger_monospace % 2)
241	{
242	$result .= '</tt>';
243	}
244	if ($trigger_bold % 2)
245	{
246	$result .= '</strong>';
247	}
248
249	for ($i = 1; $i<=5; $i ++)
250	{
251	if ($trigger_l[$i] % 2) $result .= "</h$i>";
252	}
253
254	$trigger_bold = $trigger_center = $trigger_floatl = $trigger_floatr = $trigger_inserted = $trigger_deleted = $trigger_italic = $trigger_keys = $trigger_table = 0;
255	$trigger_l = array(-1, 0, 0, 0, 0, 0);
256	$trigger_monospace = $trigger_notes = $trigger_strike = $trigger_underline = 0;
257	return $result;
258	}
259	// Ignore the closing delimiter if there is nothing to close.
260	elseif ( preg_match("/^\\|\\|\n$/", $thing, $matches) && $trigger_table == 1 )
261	{
262	return '';
263	}
264
265	// $matches[1] is element, $matches[2] is attributes, $matches[3] is styles and $matches[4] is linebreak
266	elseif ( preg_match("/^\\|([^\\|])?\\|($.?$)?(\{.?\})?(\n)?$/", $thing, $matches) )
267	{
268	for ( $i = 1; $i < 5; $i++ ) #38
269	{
270	if (!isset($matches[$i])) $matches[$i] = '';
271	}
272	//Set up the variables that will aggregate the html markup
273	$close_part = '';
274	$open_part = '';
275	$linebreak_after_open = '';
276	$selfclose = '';
277
278	// $trigger_table == 0 means no table, 1 means in table but no cell, 2 is in datacell, 3 is in headercell, 4 is in caption.
279
280	//If we have parsed the caption, close it, set trigger = 1 and return.
281	if ( $trigger_table == 4 )
282	{
283	$close_part = '</caption>'."\n";
284	$trigger_table = 1;
285	return $close_part;
286	}
287
288	//If we have parsed a cell - close it, go on to open new.
289	if ( $trigger_table == 3 )
290	{
291	$close_part = '</th>';
292	}
293	elseif ( $trigger_table == 2 )
294	{
295	$close_part = '</td>';
296	}
297	// If no cell, or we want to open a table; then there is nothing to close
298	elseif ( $trigger_table == 1 \|\| $matches[1] == '!')
299	{
300	$close_part = '';
301	}
302	else
303	{
304	//This is actually opening the table (i.e. nothing at all to close). Go on to open a cell.
305	$trigger_table = 1;
306	$close_part = '<table class="data">'."\n";
307	}
308
309	//If we are in a cell and there is a linebreak - then it is end of row.
310	if ( $trigger_table > 1 && $matches[4] == "\n" )
311	{
312	$trigger_table = 1;
313	return $close_part .= '</tr>'."\n"; //Can return here, it is closed-
314	}
315
316	//If we were in a colgroup and there is a linebreak, then it is the end.
317	if ( $trigger_colgroup == 1 && $matches[4] == "\n" )
318	{
319	$trigger_colgroup = 0;
320	return $close_part .= '</colgroup>'."\n"; //Can return here, it is closed-
321	}
322
323	//We want to start a new table, and most likely have attributes to parse.
324	//TODO: Need to find out if class="data" should be auto added, and if so - put it in the attribute list to add up.
325	if ( $matches[1] == '!' )
326	{
327	$trigger_table = 1;
328	$open_part = '<table class="data"';
329	$linebreak_after_open = "\n";
330	}
331	//Open a caption.
332	elseif ( $matches[1] == '?' )
333	{
334	$trigger_table = 4;
335	$open_part = '<caption';
336	}
337	//Start a rowgroup.
338	elseif ( $matches[1] == '#' \|\| $matches[1] == '[' \|\| $matches[1] == ']' )
339	{
340	//If we're here, we want to close any open rowgroup.
341	if (2 < $trigger_rowgroup)
342	{
343	$close_part .= '</tbody>'."\n";
344	}
345	elseif (1 < $trigger_rowgroup)
346	{
347	$close_part .= '</tfoot>'."\n";
348	}
349	elseif (0 < $trigger_rowgroup)
350	{
351	$close_part .= '</thead>'."\n";
352	}
353
354	//Then open the appropriate rowgroup.
355	if ($matches[1] == '[' )
356	{
357	$open_part .= '<thead';
358	$trigger_rowgroup = 1;
359	}
360	elseif ($matches[1] == ']' )
361	{
362	$open_part .= '<tfoot';
363	$trigger_rowgroup = 2;
364	}
365	else
366	{
367	$open_part .= '<tbody';
368	$trigger_rowgroup = 3;
369	}
370
371	$linebreak_after_open = "\n";
372	}
373	//Here we want to add colgroup.
374	elseif ( $matches[1] == '_' )
375	{
376	//close any open colgroup
377	if ( $trigger_colgroup == 1 )
378	{
379	$close_part .= '</colgroup>'."\n";
380	}
381
382	$trigger_colgroup = 1;
383	$open_part .= '<colgroup';
384	}
385	//And col elements
386	elseif ( $matches[1] == '-' )
387	{
388	$open_part .= '<col';
389	$selfclose = ' /';
390	if ( $matches[4] ) $linebreak_after_open = "\n";
391	}
392	//Ok, then it is cells.
393	else
394	{
395	$open_part = '';
396	//Need a tbody if no other rowgroup open.
397	if ($trigger_rowgroup == 0)
398	{
399	$open_part .= '<tbody>'."\n";
400	$trigger_rowgroup = 3;
401	}
402
403	//If no row, open a new one.
404	if ( $trigger_table == 1 )
405	{
406	$open_part .= '<tr>';
407	}
408
409	//Header cell.
410	if ( $matches[1] == '=' )
411	{
412	$trigger_table = 3;
413	$open_part .= '<th';
414	}
415	//Datacell
416	else
417	{
418	$trigger_table = 2;
419	$open_part .= '<td';
420	}
421	}
422
423	//If attributes...
424	if ( preg_match("/$(.*)$/", $matches[2], $attribs ) )
425	{
426	// $hints = array('core' => 'core', 'i18n' => 'i18n');
427	$hints = array();
428	//allow / disallow different attribute keys. (ie. data/header cell only.
429	if ($trigger_table == 2 \|\| $trigger_table == 3)
430	{
431	$hints['cell'] = 'cell';
432	}
433	else
434	{
435	$hints['other_table'] = 'other_table';
436	}
437	$open_part .= parse_attributes($attribs[1], $hints);
438	}
439
440	//If styles, just make attribute of it and parse again.
441	if ( preg_match("/\{(.*)\}/", $matches[3], $attribs ) )
442	{
443	$attribs = "s:".$attribs[1];
444	$open_part .= parse_attributes($attribs, array() );
445	}
446
447	//the variable $selfclose is "/" if this is a <col/> element.
448	$open_part .= $selfclose.'>';
449	return $close_part . $open_part . $linebreak_after_open;
450	}
451	//Are in table, no cell - but not asked to open new: please close and parse again. ;)
452	else if ( $trigger_table == 1 )
453	{
454	$close_part = '';
455	if (2 < $trigger_rowgroup)
456	{
457	$close_part .= '</tbody>'."\n";
458	}
459	elseif (1 < $trigger_rowgroup)
460	{
461	$close_part .= '</tfoot>'."\n";
462	}
463	elseif (0 < $trigger_rowgroup)
464	{
465	$close_part .= '</thead>'."\n";
466	}
467
468	$close_part .= '</table>'."\n";
469
470	$trigger_table = $trigger_rowgroup = 0;
471
472	//And remember to parse what we got.
473	return $close_part.wakka2callback($things);
474	}
475
476	// convert HTML thingies
477	if ($thing == "<")
478	{
479	return "<";
480	}
481	elseif ($thing == ">")
482	{
483	return ">";
484	}
485	// float box left
486	elseif ($thing == "<<")
487	{
488	return (++$trigger_floatl % 2 ? '<div class="floatl">' : '</div>');
489	}
490	// float box right
491	elseif ($thing == ">>")
492	{
493	return (++$trigger_floatr % 2 ? '<div class="floatr">' : '</div>');
494	}
495	// clear floated element
496	elseif ($thing == "::c::")
497	{
498	return ("<div class=\"clear\"> </div>\n");
499	}
500	// keyboard
501	elseif ($thing == "#%")
502	{
503	return (++$trigger_keys % 2 ? "<kbd class=\"keys\">" : "</kbd>");
504	}
505	// bold
506	elseif ($thing == "**")
507	{
508	return (++$trigger_bold % 2 ? "<strong>" : "</strong>");
509	}
510	// italic
511	elseif ($thing == "//")
512	{
513	return (++$trigger_italic % 2 ? "<em>" : "</em>");
514	}
515	// underlinue
516	elseif ($thing == "__")
517	{
518	return (++$trigger_underline % 2 ? "<span class=\"underline\">" : "</span>");
519	}
520	// monospace
521	elseif ($thing == "##")
522	{
523	return (++$trigger_monospace % 2 ? "<tt>" : "</tt>");
524	}
525	// notes
526	elseif ($thing == "''")
527	{
528	return (++$trigger_notes % 2 ? "<span class=\"notes\">" : "</span>");
529	}
530	// strikethrough
531	elseif ($thing == "++")
532	{
533	return (++$trigger_strike % 2 ? "<span class=\"strikethrough\">" : "</span>");
534	}
535	// additions
536	elseif ($thing == "££")
537	{
538	return (++$trigger_inserted % 2 ? "<ins>" : "</ins>");
539	}
540	// deletions
541	elseif ($thing == "¥¥")
542	{
543	return (++$trigger_deleted % 2 ? "<del>" : "</del>");
544	}
545	// center
546	elseif ($thing == "@@")
547	{
548	return (++$trigger_center % 2 ? "<div class=\"center\">\n" : "\n</div>\n");
549	}
550	// urls
551	elseif (preg_match("/^([a-z]+:\/\/\S+?)([^[:alnum:]^\/])?$/", $thing, $matches))
552	{
553	$url = $matches[1];
554	/* Inline images are disabled for security reason, use {{image action}} #142
555	But if you still need this functionality, update this file like below
556	if (preg_match("/\.(gif\|jpg\|png\|svg)$/si", $url)) {
557	return '<img src="'.$wakka->Link($url).'" alt="image" />'.$wakka->htmlspecialchars_ent($matches[2]);
558	} else */
559	// Mind Mapping Mod
560	if (preg_match("/\.(mm)$/si", $url)) { #145
561	return $wakka->Action("mindmap ".$url);
562	} else
563	return $wakka->Link($url).(isset($matches[2]) ? $matches[2] : ''); #38
564	}
565	// header level 5
566	elseif ($thing == "==")
567	{
568	$br = 0;
569	return (++$trigger_l[5] % 2 ? "<h5>" : "</h5>\n");
570	}
571	// header level 4
572	elseif ($thing == "===")
573	{
574	$br = 0;
575	return (++$trigger_l[4] % 2 ? "<h4>" : "</h4>\n");
576	}
577	// header level 3
578	elseif ($thing == "====")
579	{
580	$br = 0;
581	return (++$trigger_l[3] % 2 ? "<h3>" : "</h3>\n");
582	}
583	// header level 2
584	elseif ($thing == "=====")
585	{
586	$br = 0;
587	return (++$trigger_l[2] % 2 ? "<h2>" : "</h2>\n");
588	}
589	// header level 1
590	elseif ($thing == "======")
591	{
592	$br = 0;
593	return (++$trigger_l[1] % 2 ? "<h1>" : "</h1>\n");
594	}
595	// forced line breaks
596	elseif ($thing == "---")
597	{
598	return "<br />";
599	}
600	// escaped text
601	elseif (preg_match("/^\"\"(.*)\"\"$/s", $thing, $matches))
602	{
603	$ddquotes_policy = $wakka->GetConfigValue("double_doublequote_html");
604	$embedded = $matches[1];
605	if (($ddquotes_policy == 'safe') \|\| ($ddquotes_policy == 'raw'))
606	{
607	// get tags with id attributes
608	# use backref to match both single and double quotes
609	$patTagWithId = '((<[a-z][^>])((?<=\\s)id=("\|\')(.?)\\4)(.*?>))'; // @@@ #34
610	// with PREG_SET_ORDER we get an array for each match: easy to use with list()!
611	// we do the match case-insensitive so we catch uppercase HTML as well;
612	// SafeHTML will treat this but 'raw' may end up with invalid code!
613	$tags2 = preg_match_all('/'.$patTagWithId.'/i', $embedded, $matches2, PREG_SET_ORDER);
614	// step through code, replacing tags with ids with tags with new ('repaired') ids
615	$tmpembedded = $embedded;
616	$newembedded = '';
617	for ($i=0; $i < $tags2; $i++)
618	{
619	list( , $tag, $tagstart, $attrid, $quote, $id, $tagend) = $matches2[$i]; # $attrid not needed, just for clarity
620	$parts = explode($tag, $tmpembedded, 2); # split in two at matched tag
621	if ($id != ($newid = $wakka->makeId('embed', $id))) # replace if we got a new value
622	{
623	$tag = $tagstart.'id='.$quote.$newid.$quote.$tagend;
624	}
625	$newembedded .= $parts[0].$tag; # append (replacement) tag to first part
626	$tmpembedded = $parts[1]; # after tag: next bit to handle
627	}
628	$newembedded .= $tmpembedded; # add last part
629	}
630	switch ($ddquotes_policy)
631	{
632	case 'safe':
633	return $wakka->ReturnSafeHTML($newembedded);
634	case 'raw':
635	return $newembedded; # may still be invalid code - 'raw' will not be corrected!
636	default:
637	return $wakka->htmlspecialchars_ent($embedded); # display only
638	}
639	}
640	// code text
641	elseif (preg_match("/^%%(.*?)%%$/s", $thing, $matches))
642	{
643	/*
644	* Note: this routine is rewritten such that (new) language formatters
645	* will automatically be found, whether they are GeSHi language config files
646	* or "internal" Wikka formatters.
647	* Path to GeSHi language files and Wikka formatters MUST be defined in config.
648	* For line numbering (GeSHi only) a starting line can be specified after the language
649	* code, separated by a ; e.g., %%(php;27)....%%.
650	* Specifying >= 1 turns on line numbering if this is enabled in the configuration.
651	* An optional filename can be specified as well, e.g. %%(php;27;myfile.php)....%%
652	* This filename will be used by the grabcode handler.
653	*/
654	$output = ''; //reinitialize variables
655	$filename = '';
656	$valid_filename = '';
657	$code = $matches[1];
658	// if configuration path isn't set, make sure we'll get an invalid path so we
659	// don't match anything in the home directory
660	$geshi_hi_path = isset($wakka->config['geshi_languages_path']) ? $wakka->config['geshi_languages_path'] : '/:/';
661	$wikka_hi_path = isset($wakka->config['wikka_highlighters_path']) ? $wakka->config['wikka_highlighters_path'] : '/:/';
662	// check if a language (and an optional starting line or filename) has been specified
663	if (preg_match('/^'.PATTERN_OPEN_BRACKET.PATTERN_FORMATTER.PATTERN_LINE_NUMBER.PATTERN_FILENAME.PATTERN_CLOSE_BRACKET.PATTERN_CODE.'$/s', $code, $matches))
664	{
665	list(, $language, , $start, , $filename, $invalid, $code) = $matches;
666	}
667	// get rid of newlines at start and end (and preceding/following whitespace)
668	// Note: unlike trim(), this preserves any tabs at the start of the first "real" line
669	$code = preg_replace('/^\s\n+\|\n+\s$/','',$code);
670
671	// check if GeSHi path is set and we have a GeSHi highlighter for this language
672	if (isset($language) &&
673	isset($wakka->config['geshi_path']) &&
674	file_exists($geshi_hi_path.DIRECTORY_SEPARATOR.$language.'.php'))
675	{
676	// check if specified filename is valid and generate code block header
677	if (isset($filename) &&
678	strlen($filename) > 0 &&
679	strlen($invalid) == 0) # #34 TODO: use central regex library for filename validation
680	{
681	$valid_filename = $filename;
682	// create code block header
683	$output .= '<div class="code_header">';
684	// display filename and start line, if specified
685	$output .= $filename;
686	if (strlen($start)>0)
687	{
688	$output .= ' (line '.$start.')';
689	}
690	$output .= '</div>'."\n";
691	}
692	// use GeSHi for highlighting
693	$output .= $wakka->GeSHi_Highlight($code, $language, $start);
694	}
695	// check Wikka highlighter path is set and if we have an internal Wikka highlighter
696	elseif (isset($language) &&
697	isset($wakka->config['wikka_formatter_path']) &&
698	file_exists($wikka_hi_path.DIRECTORY_SEPARATOR.$language.'.php') &&
699	'wakka' != $language)
700	{
701	// use internal Wikka highlighter
702	$output = '<div class="code">'."\n";
703	$output .= $wakka->Format($code, $language);
704	$output .= "</div>\n";
705	}
706	// no language defined or no formatter found: make default code block;
707	// IncludeBuffered() will complain if 'code' formatter doesn't exist!
708	else
709	{
710	$output = '<div class="code">'."\n";
711	$output .= $wakka->Format($code, 'code');
712	$output .= "</div>\n";
713	}
714
715	// display grab button if option is set in the config file
716	if ($wakka->GetConfigValue('grabcode_button') == '1') // @@@ cast to boolean and compare to TRUE
717	{
718	$output .= $wakka->FormOpen("grabcode");
719	// build form
720	$output .= '<input type="submit" class="grabcode" name="save" value="'.GRABCODE_BUTTON.'" title="'.rtrim(sprintf(GRABCODE_BUTTON_TITLE, $valid_filename)).'" />';
721	$output .= '<input type="hidden" name="filename" value="'.urlencode($valid_filename).'" />';
722	$output .= '<input type="hidden" name="code" value="'.urlencode($code).'" />';
723	$output .= $wakka->FormClose();
724	}
725	// output
726	return $output;
727	}
728	// forced links
729	// \S : any character that is not a whitespace character
730	// \s : any whitespace character
731	// @@@ regex accepts NO non-whitespace before whitespace, surely not correct? [[ something]]
732	else if (preg_match("/^\[\[(\S*)(\s+(.+))?\]\]$/s", $thing, $matches)) # recognize forced links across lines
733	{
734	if (!isset($matches[1])) $matches[1] = ''; #38
735	if (!isset($matches[3])) $matches[3] = ''; #38
736	list (, $url, , $text) = $matches;
737	if ($url)
738	{
739	//if ($url!=($url=(preg_replace("/@@\|££\|\|\[\[/","",$url))))$result="</span>";
740	if (!$text) $text = $url;
741	//$text=preg_replace("/@@\|££\|\[\[/","",$text);
742	return $result.$wakka->Link($url, "", $text);
743	}
744	else
745	{
746	return "";
747	}
748	}
749	// indented text
750	elseif (preg_match("/(^\|\n)([\t~]+)(-\|&\|([0-9a-zA-Z]+)\))?(\n\|$)/s", $thing, $matches))
751	{
752	// new line
753	$result .= ($br ? "<br />\n" : "\n");
754
755	// we definitely want no line break in this one.
756	$br = 0;
757
758	// find out which indent type we want
759	$newIndentType = $matches[3];
760
761	if (!$newIndentType)
762	{
763	$opener = "<div class=\"indent\">";
764	$closer = "</div>"; $br = 1;
765	}
766	elseif ($newIndentType == "-")
767	{
768	$opener = "<ul><li>";
769	$closer = "</li></ul>";
770	$li = 1;
771	}
772	elseif ($newIndentType == "&")
773	{
774	$opener = "<ul class=\"thread\"><li>";
775	$closer = "</li></ul>";
776	$li = 1;
777	} #inline comments
778	else
779	{
780	if (ereg('[0-9]', $newIndentType[0])) { $newIndentType = '1'; }
781	elseif (ereg('[IVX]', $newIndentType[0])) { $newIndentType = 'I'; }
782	elseif (ereg('[ivx]', $newIndentType[0])) { $newIndentType = 'i'; }
783	elseif (ereg('[A-Z]', $newIndentType[0])) { $newIndentType = 'A'; }
784	elseif (ereg('[a-z]', $newIndentType[0])) { $newIndentType = 'a'; }
785
786	$opener = '<ol type="'.$newIndentType.'"><li>';
787	$closer = '</li></ol>';
788	$li = 1;
789	}
790
791	// get new indent level
792	$newIndentLevel = strlen($matches[2]);
793	if (($newIndentType != $curIndentType) && ($oldIndentLevel > 0))
794	{
795	for (; $oldIndentLevel > 0; $oldIndentLevel--)
796	{
797	$result .= array_pop($indentClosers);
798	}
799	}
800	if ($newIndentLevel > $oldIndentLevel)
801	{
802	for ($i = 0; $i < $newIndentLevel - $oldIndentLevel; $i++)
803	{
804	$result .= $opener;
805	array_push($indentClosers, $closer);
806	}
807	}
808	elseif ($newIndentLevel < $oldIndentLevel)
809	{
810	for ($i = 0; $i < $oldIndentLevel - $newIndentLevel; $i++)
811	{
812	$result .= array_pop($indentClosers);
813	}
814	}
815
816	$oldIndentLevel = $newIndentLevel;
817
818	if (isset($li) && !preg_match("/".str_replace(")", "\)", $opener)."$/", $result))
819	{
820	$result .= "</li><li>";
821	}
822
823	$curIndentType = $newIndentType;
824	return $result;
825	}
826	// new lines
827	else if ($thing == "\n")
828	{
829	// if we got here, there was no tab in the next line; this means that we can close all open indents.
830	$c = count($indentClosers);
831	for ($i = 0; $i < $c; $i++)
832	{
833	$result .= array_pop($indentClosers);
834	$br = 0;
835	}
836	$oldIndentLevel = 0;
837	$oldIndentLength= 0;
838	$newIndentSpace=array();
839
840	$result .= ($br ? "<br />\n" : "\n");
841	$br = 1;
842	return $result;
843	}
844	// Actions
845	elseif (preg_match("/^\{\{(.*?)\}\}$/s", $thing, $matches))
846	{
847	if ($matches[1])
848	{
849	return $wakka->Action($matches[1]);
850	}
851	else
852	{
853	return "{{}}";
854	}
855	}
856	// interwiki links!
857	elseif (preg_match("/^[A-ZÄÖÜ][A-Za-zÄÖÜßäöü]+[:]\S*$/s", $thing))
858	{
859	return $wakka->Link($thing);
860	}
861	// wiki links!
862	elseif (preg_match("/^[A-ZÄÖÜ]+[a-zßäöü]+[A-Z0-9ÄÖÜ][A-Za-z0-9ÄÖÜßäöü]*$/s", $thing))
863	{
864	return $wakka->Link($thing);
865	}
866	// separators
867	elseif (preg_match("/-{4,}/", $thing, $matches))
868	{
869	// TODO: This could probably be improved for situations where someone puts text on the same line as a separator.
870	// Which is a stupid thing to do anyway! HAW HAW! Ahem.
871	$br = 0;
872	return "<hr />\n";
873	}
874	// mind map xml
875	elseif (preg_match("/^<map.*<\/map>$/s", $thing))
876	{
877	return $wakka->Action("mindmap ".$wakka->Href()."/mindmap.mm");
878	}
879	elseif ($thing[0] == '&')
880	{
881	return ($wakka->htmlspecialchars_ent($thing));
882	}
883	// if we reach this point, it must have been an accident.
884	return $thing;
885	}
886	}
887
888	if (!function_exists('parse_attributes'))
889	{
890	function parse_attributes($attribs, $hints) {
891
892	//Sort different attributes / keys to use for different elements.
893	static $attributes = array(
894	'core' => array( 'c' => 'class','i' => 'id','s' => 'style','t' => 'title'),
895	'i18n' => array( 'd' => 'dir','l' => 'xml:lang'),
896	'cell' => array( 'a' => 'abbr','h' => 'headers','o' => 'scope','x' => 'colspan','y' => 'rowspan','z' => 'axis'),
897	'other_table' => array( 'p' => 'span','u' => 'summary')
898	);
899
900	//adds in default hints ( core + i18n )
901	$hints['core'] = 'core';
902	$hints['i18n'] = 'i18n';
903
904	$attribs = preg_split('/;(?=.:)/', $attribs);
905	$return_value = '';
906
907	foreach ( $attribs as $attrib )
908	{
909	list ($key, $value) = explode(':', $attrib, 2);
910	foreach ( $hints as $hint )
911	{
912	$temp = $attributes[$hint];
913	if ($temp) $a = $temp[$key];
914	if ($a) break;
915	}
916
917	if (!$a)
918	{
919	//This attribute isn't allowed here / is wrong.
920	// WARNING: JS vulnerability: two minus signs are not allowed in a comment, so we replace any occurence of them by underscore.
921	// Consider the code \|\|(p--><font size=1px><a href=...<!--:blabla
922	// When migrating to UTF-8, we could use str_replace('--', 'ââ', $key) to make things more pretty. //TODO garbled ... mdash?
923	echo '<!--Cannot find attribute for key "'.str_replace('--', '__', $key).'" from hints given.-->'."\n"; #i18n
924	}
925	else
926	{
927	// WARNING: JS vulnerability: use htmlspecialchars_ent to prevent JS attack!
928	$return_value .= ' '.$a.'="'.$GLOBALS['wakka']->htmlspecialchars_ent($value).'"';
929	}
930	}
931
932	return $return_value;
933	}
934	}
935
936	$text = str_replace("\r\n", "\n", $text);
937
938	// replace 4 consecutive spaces at the beginning of a line with tab character
939	// $text = preg_replace("/\n[ ]{4}/", "\n\t", $text); // moved to edit.php
940
941	if ($this->handler == "show") $mind_map_pattern = "<map.*?<\/map>\|"; else $mind_map_pattern = "";
942
943	$text = preg_replace_callback(
944	"/".
945	"%%.*?%%\|". # code
946	"\"\".*?\"\"\|". # literal
947	$mind_map_pattern.
948	"\[\[[^\[]*?\]\]\|". # forced link
949	"-{3,}\|". # forced linebreak and hr
950	"\b[a-z]+:\/\/\S+\|". # URL
951	"\\\|\'\'\|\#\#\|\#\%\|@@\|::c::\|\>\>\|\<\<\|££\|¥¥\|\+\+\|__\|<\|>\|\/\/\|". # Wiki markup
952	"======\|=====\|====\|===\|==\|". # headings
953	"(^\|\n)[\t~]+(-(?!-)\|&\|([0-9]+\|[a-zA-Z]+)\))?\|". # indents and lists
954	"\\|(?:[^\\|])?\\|(?:$.?$)?(?:\{[^\{\}]?\})?(?:\n)?\|". # Simple Tables
955	"\{\{.*?\}\}\|". # action
956	"\b[A-ZÄÖÜ][A-Za-zÄÖÜßäöü]+[:](?![=_])\S*\b\|". # InterWiki link
957	"\b([A-ZÄÖÜ]+[a-zßäöü]+[A-Z0-9ÄÖÜ][A-Za-z0-9ÄÖÜßäöü]*)\b\|". # CamelWords
958	'\\&([#a-zA-Z0-9]+;)?\|'. #ampersands! Track single ampersands or any htmlentity-like (&...;)
959	"\n". # new line
960	"/ms", "wakka2callback", $text."\n"); #append \n (#444)
961
962	// we're cutting the last <br />
963	$text = preg_replace("/<br \/>$/","", $text);
964
965	// @@@ don't report generation time unless some "debug mode" is on
966	if (isset($format_option) && preg_match(PATTERN_MATCH_PAGE_FORMATOPTION, $format_option))
967	{
968	$text .= wakka2callback('closetags'); // attempt close open tags @@@ may be needed for more than whole page!
969	$idstart = getmicrotime(TRUE);
970	$text = preg_replace_callback(
971	'#('.
972	'<h[1-6].?>.?</h[1-6]>'.
973	// other elements to be treated go here
974	')#ms','wakka3callback', $text);
975	printf('<!-- Header ID generation took %.6f seconds -->', (getmicrotime(TRUE) - $idstart)); #i18n
976	}
977	echo $text;
978	?>

Note: See TracBrowser for help on using the browser.

Context Navigation

root/trunk/formatters/wakka.php @ 761

Download in other formats: