Context Navigation

Back to Ticket #45

Ticket #45: 0000045-markdown.phps

File 0000045-markdown.phps, 32.0 KB (added by rakaur, 20 years ago)

Line
1	<?php
2
3	#
4	# Markdown - A text-to-HTML conversion tool for web writers
5	#
6	# Copyright (c) 2004 John Gruber
7	# <http://daringfireball.net/projects/markdown/>
8	#
9	# Copyright (c) 2004 Michel Fortin - Translation to PHP
10	# <http://www.michelf.com/>
11	#
12
13
14
15	global $MarkdownPHPVersion, $MarkdownSyntaxVersion,
16	$md_empty_element_suffix, $md_tab_width,
17	$md_nested_brackets_depth, $md_nested_brackets,
18	$md_escape_table, $md_backslash_escape_table;
19
20
21	$MarkdownPHPVersion = '1.0b7'; # Sun 6 Jun 2004
22	$MarkdownSyntaxVersion = '1.0b7'; # Fri 29 Apr 2004
23
24
25	#
26	# Global default settings:
27	#
28	$md_empty_element_suffix = " />"; # Change to ">" for HTML output
29	$md_tab_width = 4;
30
31
32	# -- WordPress plugin interface -----------------------------------------------
33	/*
34	Plugin Name: Markdown
35	Plugin URI: http://www.michelf.com/projects/php-markdown/
36	Description: <a href="http://daringfireball.net/projects/markdown/syntax">Markdown syntax</a> allows you to write using an easy-to-read, easy-to-write plain text format. This plugin <strong>enables Markdown for your posts and comments</strong>. Based on the original Perl version by <a href="http://daringfireball.net/">John Gruber</a>. Thanks to <a href="http://photomatt.net/">Matt</a> for making the first Markdown WP plugin. If you use this plugin you should disable Textile 1 and 2 because they do not play well with Markdown.
37	Version: 1.0b7
38	Author: Michel Fortin
39	Author URI: http://www.michelf.com/
40	*/
41	if (isset($wp_version)) {
42	# Remove default WordPress autop filter
43	remove_filter('the_content', 'wpautop');
44	remove_filter('the_excerpt', 'wpautop');
45	// remove_filter('comment_text', 'wpautop', 30); // 30 because of a bug in 1.2
46	# Add Markdown filter with priority 6 (same as Texturize).
47	add_filter('the_content', 'Markdown', 6);
48	add_filter('the_excerpt', 'Markdown', 6);
49	// add_filter('comment_text', 'Markdown', 30); // 30 because of the same bug
50	}
51
52	# -- Textile Compatibility Mode -----------------------------------------------
53	# Rename this file to "classTextile.php" and it can replace Textile anywhere.
54	if (strcasecmp(substr(__FILE__, -16), "classTextile.php") == 0) {
55	# Try to include PHP SmartyPants. Should be in the same directory.
56	@include_once 'smartypants.php';
57	# Fake Textile class. It calls Markdown instead.
58	class Textile {
59	function TextileThis($text, $lite='', $encode='', $noimage='', $strict='') {
60	if ($lite == '' && $encode == '') $text = Markdown($text);
61	if (function_exists('SmartyPants')) $text = SmartyPants($text);
62	return $text;
63	}
64	}
65	}
66
67
68	#
69	# Globals:
70	#
71
72	# Regex to match balanced [brackets].
73	# Needed to insert a maximum bracked depth while converting to PHP.
74	$md_nested_brackets_depth = 6;
75	$md_nested_brackets =
76	str_repeat('(?>[^\[\]]+\|\[', $md_nested_brackets_depth).
77	str_repeat('\])*', $md_nested_brackets_depth);
78
79	# Table of hash values for escaped characters:
80	$md_escape_table = array(
81	"\\" => md5("\\"),
82	"`" => md5("`"),
83	"" => md5(""),
84	"_" => md5("_"),
85	"{" => md5("{"),
86	"}" => md5("}"),
87	"[" => md5("["),
88	"]" => md5("]"),
89	"(" => md5("("),
90	")" => md5(")"),
91	"#" => md5("#"),
92	"." => md5("."),
93	"!" => md5("!")
94	);
95	# Create an identical table but for escaped characters.
96	$md_backslash_escape_table;
97	foreach ($md_escape_table as $key => $char)
98	$md_backslash_escape_table["\\$key"] = $char;
99
100
101	function Markdown($text) {
102	#
103	# Main function. The order in which other subs are called here is
104	# essential. Link and image substitutions need to happen before
105	# _EscapeSpecialChars(), so that any *'s or _'s in the <a>
106	# and <img> tags get encoded.
107	#
108	# Clear the global hashes. If we don't clear these, you get conflicts
109	# from other articles when generating a page which contains more than
110	# one article (e.g. an index page that shows the N most recent
111	# articles):
112	global $md_urls, $md_titles, $md_html_blocks;
113	$md_urls = array();
114	$md_titles = array();
115	$md_html_blocks = array();
116
117	# Standardize line endings:
118	# DOS to Unix and Mac to Unix
119	$text = str_replace(array("\r\n", "\r"), "\n", $text);
120
121	# Make sure $text ends with a couple of newlines:
122	$text .= "\n\n";
123
124	# Convert all tabs to spaces.
125	$text = _Detab($text);
126
127	# Strip any lines consisting only of spaces and tabs.
128	# This makes subsequent regexen easier to write, because we can
129	# match consecutive blank lines with /\n+/ instead of something
130	# contorted like /[ \t]*\n+/ .
131	$text = preg_replace('/^[ \t]+$/m', '', $text);
132
133	# Turn block-level HTML blocks into hash entries
134	$text = _HashHTMLBlocks($text);
135
136	# Strip link definitions, store in hashes.
137	$text = _StripLinkDefinitions($text);
138
139	# _EscapeSpecialChars() must be called very early, to get
140	# backslash escapes processed.
141	$text = _EscapeSpecialChars($text);
142
143	$text = _RunBlockGamut($text);
144
145	$text = _UnescapeSpecialChars($text);
146
147	return $text . "\n";
148	}
149
150
151	function _StripLinkDefinitions($text) {
152	#
153	# Strips link definitions from text, stores the URLs and titles in
154	# hash references.
155	#
156	# Link defs are in the form: ^[id]: url "optional title"
157	$text = preg_replace_callback('{
158	^[ \t]*\[(.+)\]: # id = $1
159	[ \t]*
160	\n? # maybe one newline
161	[ \t]*
162	<?(\S+?)>? # url = $2
163	[ \t]*
164	\n? # maybe one newline
165	[ \t]*
166	(?:
167	# Todo: Titles are delimited by "quotes" or (parens).
168	["(]
169	(.+?) # title = $3
170	[")]
171	[ \t]*
172	)? # title is optional
173	(?:\n+\|\Z)
174	}xm',
175	'_StripLinkDefinitions_callback',
176	$text);
177	return $text;
178	}
179	function _StripLinkDefinitions_callback($matches) {
180	global $md_urls, $md_titles;
181	$link_id = strtolower($matches[1]);
182	$md_urls[$link_id] = _EncodeAmpsAndAngles($matches[2]);
183	if (isset($matches[3]))
184	$md_titles[$link_id] = htmlentities($matches[3]);
185	return ''; # String that will replace the block
186	}
187
188
189	function _HashHTMLBlocks($text) {
190	# Hashify HTML blocks:
191	# We only want to do this for block-level HTML tags, such as headers,
192	# lists, and tables. That's because we still want to wrap <p>s around
193	# "paragraphs" that are wrapped in non-block-level tags, such as anchors,
194	# phrase emphasis, and spans. The list of tags we're looking for is
195	# hard-coded:
196	$block_tags_a = 'p\|div\|h[1-6]\|blockquote\|pre\|table\|dl\|ol\|ul\|script\|ins\|del';
197	$block_tags_b = 'p\|div\|h[1-6]\|blockquote\|pre\|table\|dl\|ol\|ul\|script';
198
199	# First, look for nested blocks, e.g.:
200	# <div>
201	# <div>
202	# tags for inner block must be indented.
203	# </div>
204	# </div>
205	#
206	# The outermost tags must start at the left margin for this to match, and
207	# the inner nested divs must be indented.
208	# We need to do this before the next, more liberal match, because the next
209	# match will start at the first `<div>` and stop at the first `</div>`.
210	$text = preg_replace_callback("{
211	( # save in $1
212	^ # start of line (with /m)
213	<($block_tags_a) # start tag = $2
214	\\b # word break
215	(.\\n)? # any number of lines, minimally matching
216	</\\2> # the matching end tag
217	[ \\t]* # trailing spaces/tabs
218	(?=\\n+\|\\Z) # followed by a newline or end of document
219	)
220	}xm",
221	'_HashHTMLBlocks_callback',
222	$text);
223
224	#
225	# Now match more liberally, simply from `\n<tag>` to `</tag>\n`
226	#
227	$text = preg_replace_callback("{
228	( # save in $1
229	^ # start of line (with /m)
230	<($block_tags_b) # start tag = $2
231	\\b # word break
232	(.\\n)? # any number of lines, minimally matching
233	.*</\\2> # the matching end tag
234	[ \\t]* # trailing spaces/tabs
235	(?=\\n+\|\\Z) # followed by a newline or end of document
236	)
237	}xm",
238	'_HashHTMLBlocks_callback',
239	$text);
240
241	# Special case just for <hr />. It was easier to make a special case than
242	# to make the other regex more complicated.
243	$text = preg_replace_callback('{
244	(?:
245	(?<=\n\n) # Starting after a blank line
246	\| # or
247	\A\n? # the beginning of the doc
248	)
249	( # save in $1
250	[ \t]*
251	<(hr) # start tag = $2
252	\b # word break
253	([^<>])*? #
254	/?> # the matching end tag
255	(?=\n{2,}\|\Z) # followed by a blank line or end of document
256	)
257	}x',
258	'_HashHTMLBlocks_callback',
259	$text);
260
261	return $text;
262	}
263	function _HashHTMLBlocks_callback($matches) {
264	global $md_html_blocks;
265	$text = $matches[1];
266	$key = md5($text);
267	$md_html_blocks[$key] = $text;
268	return "\n\n$key\n\n"; # String that will replace the block
269	}
270
271
272	function _RunBlockGamut($text) {
273	#
274	# These are all the transformations that form block-level
275	# tags like paragraphs, headers, and list items.
276	#
277	global $md_empty_element_suffix;
278
279	$text = _DoHeaders($text);
280
281	# Do Horizontal Rules:
282	$text = preg_replace(
283	array('/^( ?\* ?){3,}$/m',
284	'/^( ?- ?){3,}$/m',
285	'/^( ?_ ?){3,}$/m'),
286	"\n<hr$md_empty_element_suffix\n",
287	$text);
288
289	$text = _DoLists($text);
290
291	$text = _DoCodeBlocks($text);
292
293	$text = _DoBlockQuotes($text);
294
295	# Make links out of things like `<http://example.com/>`
296	$text = _DoAutoLinks($text);
297
298	# We already ran _HashHTMLBlocks() before, in Markdown(), but that
299	# was to escape raw HTML in the original Markdown source. This time,
300	# we're escaping the markup we've just created, so that we don't wrap
301	# <p> tags around block-level tags.
302	$text = _HashHTMLBlocks($text);
303
304	$text = _FormParagraphs($text);
305
306	return $text;
307	}
308
309
310	function _RunSpanGamut($text) {
311	#
312	# These are all the transformations that occur within block-level
313	# tags like paragraphs, headers, and list items.
314	#
315	global $md_empty_element_suffix;
316	$text = _DoCodeSpans($text);
317
318	# Fix unencoded ampersands and <'s:
319	$text = _EncodeAmpsAndAngles($text);
320
321	# Process anchor and image tags. Images must come first,
322	# because ![foo][f] looks like an anchor.
323	$text = _DoImages($text);
324	$text = _DoAnchors($text);
325
326
327	$text = _DoItalicsAndBold($text);
328
329	# Do hard breaks:
330	$text = preg_replace('/ {2,}\n/', "<br$md_empty_element_suffix\n", $text);
331
332	return $text;
333	}
334
335
336	function _EscapeSpecialChars($text) {
337	global $md_escape_table;
338	$tokens = _TokenizeHTML($text);
339
340	$text = ''; # rebuild $text from the tokens
341	$in_pre = 0; # Keep track of when we're inside <pre> or <code> tags.
342	$tags_to_skip = "!<(/?)(?:pre\|code\|kbd\|script\|math)[\s>]!";
343
344	foreach ($tokens as $cur_token) {
345	if ($cur_token[0] == 'tag') {
346	# Within tags, encode * and _ so they don't conflict
347	# with their use in Markdown for italics and strong.
348	# We're replacing each such character with its
349	# corresponding MD5 checksum value; this is likely
350	# overkill, but it should prevent us from colliding
351	# with the escape values by accident.
352	$cur_token[1] = str_replace(array('*', '_'),
353	array($md_escape_table['*'], $md_escape_table['_']),
354	$cur_token[1]);
355	$text .= $cur_token[1];
356	} else {
357	$t = $cur_token[1];
358	if (! $in_pre) {
359	$t = _EncodeBackslashEscapes($t);
360	# $t =~ s{([a-z])/([a-z])}{$1 / $2}ig;
361	}
362	$text .= $t;
363	}
364	}
365	return $text;
366	}
367
368
369	function _DoAnchors($text) {
370	#
371	# Turn Markdown link shortcuts into XHTML <a> tags.
372	#
373	global $md_nested_brackets;
374	#
375	# First, handle reference-style links: [link text] [id]
376	#
377	$text = preg_replace_callback("{
378	( # wrap whole match in $1
379	\\[
380	($md_nested_brackets) # link text = $2
381	\\]
382
383	[ ]? # one optional space
384	(?:\\n[ ]*)? # one optional newline followed by spaces
385
386	\\[
387	(.*?) # id = $3
388	\\]
389	)
390	}xs",
391	'_DoAnchors_reference_callback', $text);
392
393	#
394	# Next, inline-style links: [link text](url "optional title")
395	#
396	$text = preg_replace_callback("{
397	( # wrap whole match in $1
398	\\[
399	($md_nested_brackets) # link text = $2
400	\\]
401	\\( # literal paren
402	[ \\t]*
403	<?(.+?)>? # href = $3
404	[ \\t]*
405	( # title = $4
406	(['\"]) # quote char = $5
407	.*?
408	\\5 # matching quote
409	)? # title is optional
410	\\)
411	)
412	}xs",
413	'_DoAnchors_inline_callback', $text);
414
415	return $text;
416	}
417	function _DoAnchors_reference_callback($matches) {
418	global $md_urls, $md_titles, $md_escape_table;
419	$whole_match = $matches[1];
420	$link_text = $matches[2];
421	$link_id = strtolower($matches[3]);
422
423	if ($link_id == "") {
424	$link_id = strtolower($link_text); # for shortcut links like [this][].
425	}
426
427	if (isset($md_urls[$link_id])) {
428	$url = $md_urls[$link_id];
429	# We've got to encode these to avoid conflicting with italics/bold.
430	$url = str_replace(array('*', '_'),
431	array($md_escape_table['*'], $md_escape_table['_']),
432	$url);
433	$result = "<a href=\"$url\"";
434	if ( isset( $md_titles[$link_id] ) ) {
435	$title = $md_titles[$link_id];
436	$title = str_replace(array('*', '_'),
437	array($md_escape_table['*'],
438	$md_escape_table['_']), $title);
439	$result .= " title=\"$title\"";
440	}
441	$result .= ">$link_text</a>";
442	}
443	else {
444	$result = $whole_match;
445	}
446	return $result;
447	}
448	function _DoAnchors_inline_callback($matches) {
449	global $md_escape_table;
450	$whole_match = $matches[1];
451	$link_text = $matches[2];
452	$url = $matches[3];
453	$title = $matches[4];
454
455	# We've got to encode these to avoid conflicting with italics/bold.
456	$url = str_replace(array('*', '_'),
457	array($md_escape_table['*'], $md_escape_table['_']),
458	$url);
459	$result = "<a href=\"$url\"";
460	if ($title) {
461	$title = str_replace(array('*', '_'),
462	array($md_escape_table['*'], $md_escape_table['_']),
463	$title);
464	$result .= " title=$title";
465	}
466	$result .= ">$link_text</a>";
467
468	return $result;
469	}
470
471
472	function _DoImages($text) {
473	#
474	# Turn Markdown image shortcuts into <img> tags.
475	#
476	#
477	# First, handle reference-style labeled images: ![alt text][id]
478	#
479	$text = preg_replace_callback('{
480	( # wrap whole match in $1
481	!\[
482	(.*?) # alt text = $2
483	\]
484
485	[ ]? # one optional space
486	(?:\n[ ]*)? # one optional newline followed by spaces
487
488	\[
489	(.*?) # id = $3
490	\]
491
492	)
493	}xs',
494	'_DoImages_reference_callback', $text);
495
496	#
497	# Next, handle inline images: ![alt text](url "optional title")
498	# Don't forget: encode * and _
499
500	$text = preg_replace_callback("{
501	( # wrap whole match in $1
502	!\\[
503	(.*?) # alt text = $2
504	\\]
505	\\( # literal paren
506	[ \\t]*
507	<?(\S+?)>? # src url = $3
508	[ \\t]*
509	( # title = $4
510	(['\"]) # quote char = $5
511	.*?
512	\\5 # matching quote
513	[ \\t]*
514	)? # title is optional
515	\\)
516	)
517	}xs",
518	'_DoImages_inline_callback', $text);
519
520	return $text;
521	}
522	function _DoImages_reference_callback($matches) {
523	global $md_urls, $md_titles, $md_empty_element_suffix, $md_escape_table;
524	$whole_match = $matches[1];
525	$alt_text = $matches[2];
526	$link_id = strtolower($matches[3]);
527
528	if ($link_id == "") {
529	$link_id = strtolower($alt_text); # for shortcut links like ![this][].
530	}
531
532	if (isset($md_urls[$link_id])) {
533	$url = $md_urls[$link_id];
534	# We've got to encode these to avoid conflicting with italics/bold.
535	$url = str_replace(array('*', '_'),
536	array($md_escape_table['*'], $md_escape_table['_']),
537	$url);
538	$result = "<img src=\"$url\" alt=\"$alt_text\"";
539	if (isset($md_titles[$link_id])) {
540	$title = $md_titles[$link_id];
541	$title = str_replace(array('*', '_'),
542	array($md_escape_table['*'],
543	$md_escape_table['_']), $title);
544	$result .= " title=\"$title\"";
545	}
546	$result .= $md_empty_element_suffix;
547	}
548	else {
549	# If there's no such link ID, leave intact:
550	$result = $whole_match;
551	}
552
553	return $result;
554	}
555	function _DoImages_inline_callback($matches) {
556	global $md_empty_element_suffix, $md_escape_table;
557	$whole_match = $matches[1];
558	$alt_text = $matches[2];
559	$url = $matches[3];
560	$title = $matches[4];
561
562	# We've got to encode these to avoid conflicting with italics/bold.
563	$url = str_replace(array('*', '_'),
564	array($md_escape_table['*'], $md_escape_table['_']),
565	$url);
566	$result = "<img src=\"$url\" alt=\"$alt_text\"";
567	if (isset($title)) {
568	$title = str_replace(array('*', '_'),
569	array($md_escape_table['*'], $md_escape_table['_']),
570	$title);
571	$result .= " title=$title"; # $title already quoted
572	}
573	$result .= $md_empty_element_suffix;
574
575	return $result;
576	}
577
578
579	function _DoHeaders($text) {
580	# Setext-style headers:
581	# Header 1
582	# ========
583	#
584	# Header 2
585	# --------
586	#
587	$text = preg_replace(
588	array("/(.+)[ \t]\n=+[ \t]\n+/e",
589	"/(.+)[ \t]\n-+[ \t]\n+/e"),
590	array("'<h1>'._RunSpanGamut(_UnslashQuotes('\\1')).'</h1>\n\n'",
591	"'<h2>'._RunSpanGamut(_UnslashQuotes('\\1')).'</h2>\n\n'"),
592	$text);
593
594	# atx-style headers:
595	# # Header 1
596	# ## Header 2
597	# ## Header 2 with closing hashes ##
598	# ...
599	# ###### Header 6
600	#
601	$text = preg_replace("{
602	^(\\#{1,6}) # $1 = string of #'s
603	[ \\t]*
604	(.+?) # $2 = Header text
605	[ \\t]*
606	\\#* # optional closing #'s (not counted)
607	\\n+
608	}xme",
609	"'<h'.strlen('\\1').'>'._RunSpanGamut(_UnslashQuotes('\\2')).'</h'.strlen('\\1').'>\n\n'",
610	$text);
611
612	return $text;
613	}
614
615
616	function _DoLists($text) {
617	#
618	# Form HTML ordered (numbered) and unordered (bulleted) lists.
619	#
620	global $md_tab_width;
621	$less_than_tab = $md_tab_width - 1;
622
623	$text = preg_replace_callback("{
624	(
625	(
626	^[ ]{0,$less_than_tab}
627	(\\*\|\\d+[.])
628	[ \\t]+
629	)
630	(?s:.+?)
631	(
632	\\z
633	\|
634	\\n{2,}
635	(?=\\S)
636	(?![ \\t]* (\\*\|\\d+[.]) [ \\t]+)
637	)
638	)
639	}xm",
640	'_DoLists_callback', $text);
641
642	return $text;
643	}
644	function _DoLists_callback($matches) {
645	$list_type = ($matches[3] == "*") ? "ul" : "ol";
646	$list = $matches[1];
647	# Turn double returns into triple returns, so that we can make a
648	# paragraph for the last item in a list, if necessary:
649	$list = preg_replace("/\n{2,}/", "\n\n\n", $list);
650	$result = _ProcessListItems($list);
651	$result = "<$list_type>\n" . $result . "</$list_type>\n";
652	return $result;
653	}
654
655
656	function _ProcessListItems($list_str) {
657	# trim trailing blank lines:
658	$list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str);
659
660	$list_str = preg_replace_callback('{
661	(\n)? # leading line = $1
662	(^[ \t]*) # leading whitespace = $2
663	(\*\|\d+[.]) [ \t]+ # list marker = $3
664	((?s:.+?) # list item text = $4
665	(\n{1,2}))
666	(?= \n* (\z \| \2 (\*\|\d+[.]) [ \t]+))
667	}xm',
668	'_ProcessListItems_callback', $list_str);
669
670	return $list_str;
671	}
672	function _ProcessListItems_callback($matches) {
673	$item = $matches[4];
674	$leading_line = $matches[1];
675	$leading_space = $matches[2];
676
677	if ($leading_line \|\| preg_match('/\n{2,}/', $item)) {
678	$item = _RunBlockGamut(_Outdent($item));
679	#$item =~ s/\n+/\n/g;
680	}
681	else {
682	# Recursion for sub-lists:
683	$item = _DoLists(_Outdent($item));
684	$item = rtrim($item, "\n");
685	$item = _RunSpanGamut($item);
686	}
687
688	return "<li>" . $item . "</li>\n";
689	}
690
691
692	function _DoCodeBlocks($text) {
693	#
694	# Process Markdown `<pre><code>` blocks.
695	#
696	global $md_tab_width;
697	$text = preg_replace_callback("{
698	(.?) # $1 = preceding character
699	(:) # $2 = colon delimiter
700	(\\n+) # $3 = newlines after colon
701	( # $4 = the code block -- one or more lines, starting with a space/tab
702	(?:
703	(?:[ ]\{$md_tab_width} \| \\t) # Lines must start with a tab or a tab-width of spaces
704	.*\\n+
705	)+
706	)
707	((?=^[ ]{0,$md_tab_width}\\S)\|\\Z) # Lookahead for non-space at line-start, or end of doc
708	}xm",
709	'_DoCodeBlocks_callback', $text);
710
711	return $text;
712	}
713	function _DoCodeBlocks_callback($matches) {
714	$prevchar = $matches[1];
715	$newlines = $matches[2];
716	$codeblock = $matches[4];
717
718	#
719	# Check the preceding character before the ":". If it's not
720	# whitespace, then the ":" remains; if it is whitespace,
721	# the ":" disappears completely, along with the space character.
722	#
723	$prefix = "";
724	if (!(preg_match('/\s/', $prevchar) \|\| ($prevchar == ""))) {
725	$prefix = "$prevchar:";
726	}
727	$codeblock = _EncodeCode(_Outdent($codeblock));
728	$codeblock = _Detab($codeblock);
729	# trim leading newlines and trailing whitespace
730	$codeblock = preg_replace(array('/\A\n+/', '/\s+\z/'), '', $codeblock);
731
732	$result = $prefix . "\n\n<pre><code>" . $codeblock . "\n</code></pre>\n\n";
733
734	return $result;
735	}
736
737
738	function _DoCodeSpans($text) {
739	#
740	# * Backtick quotes are used for <code></code> spans.
741	#
742	# * You can use multiple backticks as the delimiters if you want to
743	# include literal backticks in the code span. So, this input:
744	#
745	# Just type ``foo `bar` baz`` at the prompt.
746	#
747	# Will translate to:
748	#
749	# <p>Just type <code>foo `bar` baz</code> at the prompt.</p>
750	#
751	# There's no arbitrary limit to the number of backticks you
752	# can use as delimters. If you need three consecutive backticks
753	# in your code, use four for delimiters, etc.
754	#
755	# * You can use spaces to get literal backticks at the edges:
756	#
757	# ... type `` `bar` `` ...
758	#
759	# Turns to:
760	#
761	# ... type <code>`bar`</code> ...
762	#
763	$text = preg_replace_callback("@
764	(`+) # Opening run of `
765	(.+?) # the code block
766	(?<!`)
767	\\1
768	(?!`)
769	@xs",
770	'_DoCodeSpans_callback', $text);
771
772	return $text;
773	}
774	function _DoCodeSpans_callback($matches) {
775	$c = $matches[2];
776	$c = preg_replace('/^[ \t]*/', '', $c); # leading whitespace
777	$c = preg_replace('/[ \t]*$/', '', $c); # trailing whitespace
778	$c = _EncodeCode($c);
779	return "<code>$c</code>";
780	}
781
782
783	function _EncodeCode($_) {
784	#
785	# Encode/escape certain characters inside Markdown code runs.
786	# The point is that in code, these characters are literals,
787	# and lose their special Markdown meanings.
788	#
789	global $md_escape_table;
790
791	# Encode all ampersands; HTML entities are not
792	# entities within a Markdown code span.
793	$_ = str_replace('&', '&', $_);
794
795	# Do the angle bracket song and dance:
796	$_ = str_replace(array('<', '>'),
797	array('<', '>'), $_);
798
799	# Now, escape characters that are magic in Markdown:
800	$_ = str_replace(array_keys($md_escape_table),
801	array_values($md_escape_table), $_);
802
803	return $_;
804	}
805
806
807	function _DoItalicsAndBold($text) {
808	# <strong> must go first:
809	$text = preg_replace('{ (\\\|__) (?=\S) (.+?) (?<=\S) \1 }sx',
810	'<strong>\2</strong>', $text);
811	# Then <em>:
812	$text = preg_replace('{ (\*\|_) (?=\S) (.+?) (?<=\S) \1 }sx',
813	'<em>\2</em>', $text);
814
815	return $text;
816	}
817
818
819	function _DoBlockQuotes($text) {
820	$text = preg_replace_callback('/
821	( # Wrap whole match in $1
822	(
823	^[ \t]*>[ \t]? # ">" at the start of a line
824	.+\n # rest of the first line
825	(.+\n)* # subsequent consecutive lines
826	\n* # blanks
827	)+
828	)
829	/xm',
830	'_DoBlockQuotes_callback', $text);
831
832	return $text;
833	}
834	function _DoBlockQuotes_callback($matches) {
835	$bq = $matches[1];
836	# trim one level of quoting - trim whitespace-only lines
837	$bq = preg_replace(array('/^[ \t]*>[ \t]?/m', '/^[ \t]+$/m'), '', $bq);
838	$bq = _RunBlockGamut($bq); # recurse
839	$bq = preg_replace('/^/m', "\t", $bq);
840
841	return "<blockquote>\n$bq\n</blockquote>\n\n";
842	}
843
844
845	function _FormParagraphs($text) {
846	#
847	# Params:
848	# $text - string to process with html <p> tags
849	#
850	global $md_html_blocks;
851
852	# Strip leading and trailing lines:
853	$text = preg_replace(array('/\A\n+/', '/\n+\z/'), '', $text);
854
855	$grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY);
856	$count = count($grafs);
857
858	#
859	# Wrap <p> tags.
860	#
861	foreach ($grafs as $key => $value) {
862	if (!isset( $md_html_blocks[$value] )) {
863	$value = _RunSpanGamut($value);
864	$value = preg_replace('/^([ \t]*)/', '<p>', $value);
865	$value .= "</p>";
866	$grafs[$key] = $value;
867	}
868	}
869
870	#
871	# Unhashify HTML blocks
872	#
873	foreach ($grafs as $key => $value) {
874	if (isset( $md_html_blocks[$value] )) {
875	$grafs[$key] = $md_html_blocks[$value];
876	}
877	}
878
879	return implode("\n\n", $grafs);
880	}
881
882
883	function _EncodeAmpsAndAngles($text) {
884	# Smart processing for ampersands and angle brackets that need to be encoded.
885
886	# Ampersand-encoding based entirely on Nat Irons's Amputator MT plugin:
887	# http://bumppo.net/projects/amputator/
888	$text = preg_replace('/&(?!#?[xX]?(?:[0-9a-fA-F]+\|\w+);)/',
889	'&', $text);;
890
891	# Encode naked <'s
892	$text = preg_replace('{<(?![a-z/?\$!])}i', '<', $text);
893
894	return $text;
895	}
896
897
898	function _EncodeBackslashEscapes($text) {
899	#
900	# Parameter: String.
901	# Returns: The string, with after processing the following backslash
902	# escape sequences.
903	#
904	global $md_escape_table, $md_backslash_escape_table;
905	# Must process escaped backslashes first.
906	return str_replace(array_keys($md_backslash_escape_table),
907	array_values($md_backslash_escape_table), $text);
908	}
909
910
911	function _DoAutoLinks($text) {
912	$text = preg_replace("!<((https?\|ftp):[^'\">\\s]+)>!",
913	'<a href="\1">\1</a>', $text);
914
915	# Email addresses: <address@domain.foo>
916	$text = preg_replace('{
917	<
918	(
919	[-.\w]+
920	\@
921	[-a-z0-9]+(\.[-a-z0-9]+)*\.[a-z]+
922	)
923	>
924	}exi',
925	"_EncodeEmailAddress(_UnescapeSpecialChars(_UnslashQuotes('\\1')))",
926	$text);
927
928	return $text;
929	}
930
931
932	function _EncodeEmailAddress($addr) {
933	#
934	# Input: an email address, e.g. "foo@example.com"
935	#
936	# Output: the email address as a mailto link, with each character
937	# of the address encoded as either a decimal or hex entity, in
938	# the hopes of foiling most address harvesting spam bots. E.g.:
939	#
940	# <a href="mailto:foo@e
941	# xample.com">foo
942	# @example.com</a>
943	#
944	# Based by a filter by Matthew Wickline, posted to the BBEdit-Talk
945	# mailing list: <http://tinyurl.com/yu7ue>
946	#
947	$addr = "mailto:" . $addr;
948	$length = strlen($addr);
949
950	# leave ':' alone (to spot mailto: later)
951	$addr = preg_replace_callback('/([^\:])/',
952	'_EncodeEmailAddress_callback', $addr);
953
954	$addr = "<a href=\"$addr\">$addr</a>";
955	# strip the mailto: from the visible part
956	$addr = preg_replace('/">.+?:/', '">', $addr);
957
958	return $addr;
959	}
960	function _EncodeEmailAddress_callback($matches) {
961	$char = $matches[1];
962	$r = rand(0, 100);
963	# roughly 10% raw, 45% hex, 45% dec
964	# '@' must be encoded. I insist.
965	if ($r > 90 && $char != '@') return $char;
966	if ($r < 45) return '&#x'.dechex(ord($char)).';';
967	return '&#'.ord($char).';';
968	}
969
970
971	function _UnescapeSpecialChars($text) {
972	#
973	# Swap back in all the special characters we've hidden.
974	#
975	global $md_escape_table;
976	return str_replace(array_values($md_escape_table),
977	array_keys($md_escape_table), $text);
978	}
979
980
981	# Tokenize_HTML is shared between PHP Markdown and PHP SmartyPants.
982	# We only define it if it is not already defined.
983	if (!function_exists('_TokenizeHTML')) {
984	function _TokenizeHTML($str) {
985	#
986	# Parameter: String containing HTML markup.
987	# Returns: An array of the tokens comprising the input
988	# string. Each token is either a tag (possibly with nested,
989	# tags contained therein, such as <a href="<MTFoo>">, or a
990	# run of text between tags. Each element of the array is a
991	# two-element array; the first is either 'tag' or 'text';
992	# the second is the actual value.
993	#
994	#
995	# Regular expression derived from the _tokenize() subroutine in
996	# Brad Choate's MTRegex plugin.
997	# <http://www.bradchoate.com/past/mtregex.php>
998	#
999	$index = 0;
1000	$tokens = array();
1001
1002	$depth = 6;
1003	$nested_tags = str_repeat('(?:<[a-z\/!$](?:[^<>]\|',$depth)
1004	.str_repeat(')*>)', $depth);
1005	$match = "(?s:<!(?:--.?--\s)+>)\|". # comment
1006	"(?s:<\?.*?\?>)\|". # processing instruction
1007	"$nested_tags"; # nested tags
1008
1009	$parts = preg_split("/($match)/", $str, -1, PREG_SPLIT_DELIM_CAPTURE);
1010
1011	foreach ($parts as $part) {
1012	if (++$index % 2 && $part != '')
1013	array_push($tokens, array('text', $part));
1014	else
1015	array_push($tokens, array('tag', $part));
1016	}
1017
1018	return $tokens;
1019	}
1020	}
1021
1022	function _Outdent($text) {
1023	#
1024	# Remove one level of line-leading tabs or spaces
1025	#
1026	global $md_tab_width;
1027	return preg_replace("/^(\\t\|[ ]{1,$md_tab_width})/m", "", $text);
1028	}
1029
1030
1031	function _Detab($text) {
1032	#
1033	# Inspired from a post by Bart Lateur:
1034	# <http://www.nntp.perl.org/group/perl.macperl.anyperl/154>
1035	#
1036	global $md_tab_width;
1037	$text = preg_replace(
1038	"/(.*?)\t/e",
1039	"'\\1'.str_repeat(' ', $md_tab_width - strlen('\\1') % $md_tab_width)",
1040	$text);
1041	return $text;
1042	}
1043
1044
1045	function _UnslashQuotes($text) {
1046	#
1047	# This function is useful to remove automaticaly slashed double quotes
1048	# when using preg_replace and evaluating an expression.
1049	# Parameter: String.
1050	# Returns: The string with any slash-double-quote (\") sequence replaced
1051	# by a single double quote.
1052	#
1053	return str_replace('\"', '"', $text);
1054	}
1055
1056
1057	/*
1058
1059	PHP Markdown
1060	============
1061
1062	Description
1063	-----------
1064
1065	This is a PHP translation of the original Markdown parser written in
1066	Perl by John Gruber.
1067
1068	Markdown is a text-to-HTML filter; it translates an easy-to-read /
1069	easy-to-write structured text format into HTML. Markdown's text format
1070	is most similar to that of plain text email, and supports features such
1071	as headers, emphasis, code blocks, blockquotes, and links.
1072
1073	Markdown's syntax is designed not as a generic markup language, but
1074	specifically to serve as a front-end to (X)HTML. You can use span-level
1075	HTML tags anywhere in a Markdown document, and you can use block level
1076	HTML tags (like <div> and <table> as well).
1077
1078	For more information about Markdown's syntax, see:
1079
1080	<http://daringfireball.net/projects/markdown/>
1081
1082
1083	Bugs
1084	----
1085
1086	To file bug reports please send email to:
1087
1088	<michel.fortin@michelf.com>
1089
1090	Please include with your report: (1) the example input; (2) the output you
1091	expected; (3) the output Markdown actually produced.
1092
1093
1094	Version history
1095	---------------
1096
1097	1.0b7: Sat 12 Jun 2004
1098
1099	* Added 'math' to $tags_to_skip pattern, for MathML users.
1100
1101	* Tweaked regex for identifying HTML entities in
1102	_EncodeAmpsAndAngles(), so as to allow for the very long entity
1103	names used by MathML. (Thanks to Jacques Distler for the patch.)
1104
1105	* Changed the internals of _TokenizeHTML to lower the PHP version
1106	requirement to PHP 4.0.5.
1107
1108
1109	1.0b6: Sun 6 Jun 2004
1110
1111	* Added a WordPress plugin interface. This means that you can
1112	directly put the `markdown.php` file into the `wp-content/plugins`
1113	directory and then activate it from the administrative interface.
1114
1115	* Added a Textile compatibility interface. Rename this file to
1116	`classTextile.php` and it can replace Textile anywhere.
1117
1118	* The title attribute of reference-style links were ignored.
1119	This is now fixed.
1120
1121	* Changed internal variables names so that they begin with `md_`
1122	instead of `g_`. This should reduce the risk of name collision with
1123	other programs.
1124
1125
1126	1.0b5: Sun 2 May 2004
1127
1128	* Workaround for supporting <ins> and <del> as block-level tags.
1129	This only works if the start and end tags are on lines by
1130	themselves.
1131
1132	* Three or more underscores can now be used for horizontal rules.
1133
1134	* Lines containing only whitespace are trimmed from blockquotes.
1135
1136	* You can now optionally wrap URLs with angle brackets -- like so:
1137	`<http://example.com>` -- in link definitions and inline links and
1138	images.
1139
1140	* `_` and `*` characters in links and images are no longer escaped
1141	as HTML entities. Instead, we use the ridiculous but effective MD5
1142	hashing trick that's used to hide these characters elsewhere. The
1143	end result is that the HTML output uses the literal `*` and `_`
1144	characters, rather than the ugly entities.
1145
1146	* Passing an empty string to the Markdown function no longer creates
1147	an empty paragraph.
1148
1149	* Added a global declaration at the beginning of the file. This
1150	means you can now `include 'markdown.php'` from inside a function.
1151
1152
1153	1.0b4.1: Sun 4 Apr 2004
1154
1155	* Fixed a bug where image tags did not close.
1156
1157	* Fixed a bug where brakets `[]` inside a link caused the link to be
1158	ignored. PHP Markdown support only 6 (!) level of brakets inside a link
1159	(while John's original version of Markdown in Perl support much more).
1160
1161
1162	1.0b4: Sat 27 Mar 2004
1163
1164	* First release of PHP Markdown, based on the 1.0b4 release.
1165
1166
1167	Author & contributors
1168	---------------------
1169
1170	Original version by John Gruber
1171	<http://daringfireball.net/>
1172
1173	PHP translation by Michel Fortin
1174	<http://www.michelf.com/>
1175
1176	First WordPress plugin interface written by Matt Mullenweg
1177	(<http://photomatt.net/>)
1178
1179
1180	Copyright and license
1181	---------------------
1182
1183	Copyright (c) 2003-2004 John Gruber
1184	<http://daringfireball.net/>
1185	All rights reserved.
1186
1187	Copyright (c) 2004 Michel Fortin
1188	<http://www.michelf.com/>
1189
1190	Markdown is free software; you can redistribute it and/or modify it
1191	under the terms of the GNU General Public License as published by the
1192	Free Software Foundation; either version 2 of the License, or (at your
1193	option) any later version.
1194
1195	Markdown is distributed in the hope that it will be useful, but WITHOUT
1196	ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
1197	FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
1198	for more details.
1199
1200	*/
1201	?>

Trac UI Preferences

Download in other formats:

Original Format

Make WordPress Core

Context Navigation

Ticket #45: 0000045-markdown.phps

Download in other formats: