Context Navigation

source: subversion/trunk/roundcubemail/program/lib/html2text.php @ 4013

Last change on this file since 4013 was 4013, checked in by alec, 3 years ago
Fix HTML to plain text conversion doesn't handle citation blocks (#1486921)
Property svn:eol-style set to `native` Property svn:keywords set to `Author Date Id Revision`
File size: 22.8 KB

Line
1	<?php
2
3	/*************************************************************************
4	* *
5	* class.html2text.inc *
6	* *
7	*************************************************************************
8	* *
9	* Converts HTML to formatted plain text *
10	* *
11	* Copyright (c) 2005-2007 Jon Abernathy <jon@chuggnutt.com> *
12	* All rights reserved. *
13	* *
14	* This script is free software; you can redistribute it and/or modify *
15	* it under the terms of the GNU General Public License as published by *
16	* the Free Software Foundation; either version 2 of the License, or *
17	* (at your option) any later version. *
18	* *
19	* The GNU General Public License can be found at *
20	* http://www.gnu.org/copyleft/gpl.html. *
21	* *
22	* This script is distributed in the hope that it will be useful, *
23	* but WITHOUT ANY WARRANTY; without even the implied warranty of *
24	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
25	* GNU General Public License for more details. *
26	* *
27	* Author(s): Jon Abernathy <jon@chuggnutt.com> *
28	* *
29	* Last modified: 08/08/07 *
30	* *
31	*************************************************************************/
32
33
34	/**
35	* Takes HTML and converts it to formatted, plain text.
36	*
37	* Thanks to Alexander Krug (http://www.krugar.de/) to pointing out and
38	* correcting an error in the regexp search array. Fixed 7/30/03.
39	*
40	* Updated set_html() function's file reading mechanism, 9/25/03.
41	*
42	* Thanks to Joss Sanglier (http://www.dancingbear.co.uk/) for adding
43	* several more HTML entity codes to the $search and $replace arrays.
44	* Updated 11/7/03.
45	*
46	* Thanks to Darius Kasperavicius (http://www.dar.dar.lt/) for
47	* suggesting the addition of $allowed_tags and its supporting function
48	* (which I slightly modified). Updated 3/12/04.
49	*
50	* Thanks to Justin Dearing for pointing out that a replacement for the
51	* <TH> tag was missing, and suggesting an appropriate fix.
52	* Updated 8/25/04.
53	*
54	* Thanks to Mathieu Collas (http://www.myefarm.com/) for finding a
55	* display/formatting bug in the _build_link_list() function: email
56	* readers would show the left bracket and number ("[1") as part of the
57	* rendered email address.
58	* Updated 12/16/04.
59	*
60	* Thanks to Wojciech Bajon (http://histeria.pl/) for submitting code
61	* to handle relative links, which I hadn't considered. I modified his
62	* code a bit to handle normal HTTP links and MAILTO links. Also for
63	* suggesting three additional HTML entity codes to search for.
64	* Updated 03/02/05.
65	*
66	* Thanks to Jacob Chandler for pointing out another link condition
67	* for the _build_link_list() function: "https".
68	* Updated 04/06/05.
69	*
70	* Thanks to Marc Bertrand (http://www.dresdensky.com/) for
71	* suggesting a revision to the word wrapping functionality; if you
72	* specify a $width of 0 or less, word wrapping will be ignored.
73	* Updated 11/02/06.
74	*
75	* *** Big housecleaning updates below:
76	*
77	* Thanks to Colin Brown (http://www.sparkdriver.co.uk/) for
78	* suggesting the fix to handle </li> and blank lines (whitespace).
79	* Christian Basedau (http://www.movetheweb.de/) also suggested the
80	* blank lines fix.
81	*
82	* Special thanks to Marcus Bointon (http://www.synchromedia.co.uk/),
83	* Christian Basedau, Norbert Laposa (http://ln5.co.uk/),
84	* Bas van de Weijer, and Marijn van Butselaar
85	* for pointing out my glaring error in the <th> handling. Marcus also
86	* supplied a host of fixes.
87	*
88	* Thanks to Jeffrey Silverman (http://www.newtnotes.com/) for pointing
89	* out that extra spaces should be compressed--a problem addressed with
90	* Marcus Bointon's fixes but that I had not yet incorporated.
91	*
92	* Thanks to Daniel Schledermann (http://www.typoconsult.dk/) for
93	* suggesting a valuable fix with <a> tag handling.
94	*
95	* Thanks to Wojciech Bajon (again!) for suggesting fixes and additions,
96	* including the <a> tag handling that Daniel Schledermann pointed
97	* out but that I had not yet incorporated. I haven't (yet)
98	* incorporated all of Wojciech's changes, though I may at some
99	* future time.
100	*
101	* *** End of the housecleaning updates. Updated 08/08/07.
102	*
103	* @author Jon Abernathy <jon@chuggnutt.com>
104	* @version 1.0.0
105	* @since PHP 4.0.2
106	*/
107	class html2text
108	{
109
110	/**
111	* Contains the HTML content to convert.
112	*
113	* @var string $html
114	* @access public
115	*/
116	var $html;
117
118	/**
119	* Contains the converted, formatted text.
120	*
121	* @var string $text
122	* @access public
123	*/
124	var $text;
125
126	/**
127	* Maximum width of the formatted text, in columns.
128	*
129	* Set this value to 0 (or less) to ignore word wrapping
130	* and not constrain text to a fixed-width column.
131	*
132	* @var integer $width
133	* @access public
134	*/
135	var $width = 70;
136
137	/**
138	* List of preg* regular expression patterns to search for,
139	* used in conjunction with $replace.
140	*
141	* @var array $search
142	* @access public
143	* @see $replace
144	*/
145	var $search = array(
146	"/\r/", // Non-legal carriage return
147	"/[\n\t]+/", // Newlines and tabs
148	'/[ ]{2,}/', // Runs of spaces, pre-handling
149	'/<script[^>]>.?<\/script>/i', // <script>s -- which strip_tags supposedly has problems with
150	'/<style[^>]>.?<\/style>/i', // <style>s -- which strip_tags supposedly has problems with
151	'/<p[^>]*>/i', // <P>
152	'/<br[^>]*>/i', // <br>
153	'/<i[^>]>(.?)<\/i>/i', // <i>
154	'/<em[^>]>(.?)<\/em>/i', // <em>
155	'/(<ul[^>]*>\|<\/ul>)/i', // <ul> and </ul>
156	'/(<ol[^>]*>\|<\/ol>)/i', // <ol> and </ol>
157	'/<li[^>]>(.?)<\/li>/i', // <li> and </li>
158	'/<li[^>]*>/i', // <li>
159	'/<hr[^>]*>/i', // <hr>
160	'/<div[^>]*>/i', // <div>
161	'/(<table[^>]*>\|<\/table>)/i', // <table> and </table>
162	'/(<tr[^>]*>\|<\/tr>)/i', // <tr> and </tr>
163	'/<td[^>]>(.?)<\/td>/i', // <td> and </td>
164	'/&(nbsp\|#160);/i', // Non-breaking space
165	'/&(quot\|rdquo\|ldquo\|#8220\|#8221\|#147\|#148);/i',
166	// Double quotes
167	'/&(apos\|rsquo\|lsquo\|#8216\|#8217);/i', // Single quotes
168	'/>/i', // Greater-than
169	'/</i', // Less-than
170	'/&(amp\|#38);/i', // Ampersand
171	'/&(copy\|#169);/i', // Copyright
172	'/&(trade\|#8482\|#153);/i', // Trademark
173	'/&(reg\|#174);/i', // Registered
174	'/&(mdash\|#151\|#8212);/i', // mdash
175	'/&(ndash\|minus\|#8211\|#8722);/i', // ndash
176	'/&(bull\|#149\|#8226);/i', // Bullet
177	'/&(pound\|#163);/i', // Pound sign
178	'/&(euro\|#8364);/i', // Euro sign
179	'/[ ]{2,}/' // Runs of spaces, post-handling
180	);
181
182	/**
183	* List of pattern replacements corresponding to patterns searched.
184	*
185	* @var array $replace
186	* @access public
187	* @see $search
188	*/
189	var $replace = array(
190	'', // Non-legal carriage return
191	' ', // Newlines and tabs
192	' ', // Runs of spaces, pre-handling
193	'', // <script>s -- which strip_tags supposedly has problems with
194	'', // <style>s -- which strip_tags supposedly has problems with
195	"\n\n", // <P>
196	"\n", // <br>
197	'_\\1_', // <i>
198	'_\\1_', // <em>
199	"\n\n", // <ul> and </ul>
200	"\n\n", // <ol> and </ol>
201	"\t* \\1\n", // <li> and </li>
202	"\n\t* ", // <li>
203	"\n-------------------------\n", // <hr>
204	"<div>\n", // <div>
205	"\n\n", // <table> and </table>
206	"\n", // <tr> and </tr>
207	"\t\t\\1\n", // <td> and </td>
208	' ', // Non-breaking space
209	'"', // Double quotes
210	"'", // Single quotes
211	'>',
212	'<',
213	'&',
214	'(c)',
215	'(tm)',
216	'(R)',
217	'--',
218	'-',
219	'*',
220	'Â£',
221	'EUR', // Euro sign. ?
222	' ' // Runs of spaces, post-handling
223	);
224
225	/**
226	* List of preg* regular expression patterns to search for
227	* and replace using callback function.
228	*
229	* @var array $callback_search
230	* @access public
231	*/
232	var $callback_search = array(
233	'/<(h)[123456][^>]>(.?)<\/h[123456]>/i', // H1 - H3
234	'/<(b)[^>]>(.?)<\/b>/i', // <b>
235	'/<(strong)[^>]>(.?)<\/strong>/i', // <strong>
236	'/<(a) [^>]href=("\|\')([^"\']+)\2[^>]>(.*?)<\/a>/i',
237	// <a href="">
238	'/<(th)[^>]>(.?)<\/th>/i', // <th> and </th>
239	);
240
241	/**
242	* List of preg* regular expression patterns to search for in PRE body,
243	* used in conjunction with $pre_replace.
244	*
245	* @var array $pre_search
246	* @access public
247	* @see $pre_replace
248	*/
249	var $pre_search = array(
250	"/\n/",
251	"/\t/",
252	'/ /',
253	'/<pre[^>]*>/',
254	'/<\/pre>/'
255	);
256
257	/**
258	* List of pattern replacements corresponding to patterns searched for PRE body.
259	*
260	* @var array $pre_replace
261	* @access public
262	* @see $pre_search
263	*/
264	var $pre_replace = array(
265	'<br>',
266	'    ',
267	' ',
268	'',
269	''
270	);
271
272	/**
273	* Contains a list of HTML tags to allow in the resulting text.
274	*
275	* @var string $allowed_tags
276	* @access public
277	* @see set_allowed_tags()
278	*/
279	var $allowed_tags = '';
280
281	/**
282	* Contains the base URL that relative links should resolve to.
283	*
284	* @var string $url
285	* @access public
286	*/
287	var $url;
288
289	/**
290	* Indicates whether content in the $html variable has been converted yet.
291	*
292	* @var boolean $_converted
293	* @access private
294	* @see $html, $text
295	*/
296	var $_converted = false;
297
298	/**
299	* Contains URL addresses from links to be rendered in plain text.
300	*
301	* @var string $_link_list
302	* @access private
303	* @see _build_link_list()
304	*/
305	var $_link_list = '';
306
307	/**
308	* Number of valid links detected in the text, used for plain text
309	* display (rendered similar to footnotes).
310	*
311	* @var integer $_link_count
312	* @access private
313	* @see _build_link_list()
314	*/
315	var $_link_count = 0;
316
317	/**
318	* Boolean flag, true if a table of link URLs should be listed after the text.
319	*
320	* @var boolean $_do_links
321	* @access private
322	* @see html2text()
323	*/
324	var $_do_links = true;
325
326	/**
327	* Constructor.
328	*
329	* If the HTML source string (or file) is supplied, the class
330	* will instantiate with that source propagated, all that has
331	* to be done it to call get_text().
332	*
333	* @param string $source HTML content
334	* @param boolean $from_file Indicates $source is a file to pull content from
335	* @param boolean $do_links Indicate whether a table of link URLs is desired
336	* @param integer $width Maximum width of the formatted text, 0 for no limit
337	* @access public
338	* @return void
339	*/
340	function html2text( $source = '', $from_file = false, $do_links = true, $width = 75 )
341	{
342	if ( !empty($source) ) {
343	$this->set_html($source, $from_file);
344	}
345
346	$this->set_base_url();
347	$this->_do_links = $do_links;
348	$this->width = $width;
349	}
350
351	/**
352	* Loads source HTML into memory, either from $source string or a file.
353	*
354	* @param string $source HTML content
355	* @param boolean $from_file Indicates $source is a file to pull content from
356	* @access public
357	* @return void
358	*/
359	function set_html( $source, $from_file = false )
360	{
361	if ( $from_file && file_exists($source) ) {
362	$this->html = file_get_contents($source);
363	}
364	else
365	$this->html = $source;
366
367	$this->_converted = false;
368	}
369
370	/**
371	* Returns the text, converted from HTML.
372	*
373	* @access public
374	* @return string
375	*/
376	function get_text()
377	{
378	if ( !$this->_converted ) {
379	$this->_convert();
380	}
381
382	return $this->text;
383	}
384
385	/**
386	* Prints the text, converted from HTML.
387	*
388	* @access public
389	* @return void
390	*/
391	function print_text()
392	{
393	print $this->get_text();
394	}
395
396	/**
397	* Alias to print_text(), operates identically.
398	*
399	* @access public
400	* @return void
401	* @see print_text()
402	*/
403	function p()
404	{
405	print $this->get_text();
406	}
407
408	/**
409	* Sets the allowed HTML tags to pass through to the resulting text.
410	*
411	* Tags should be in the form "<p>", with no corresponding closing tag.
412	*
413	* @access public
414	* @return void
415	*/
416	function set_allowed_tags( $allowed_tags = '' )
417	{
418	if ( !empty($allowed_tags) ) {
419	$this->allowed_tags = $allowed_tags;
420	}
421	}
422
423	/**
424	* Sets a base URL to handle relative links.
425	*
426	* @access public
427	* @return void
428	*/
429	function set_base_url( $url = '' )
430	{
431	if ( empty($url) ) {
432	if ( !empty($_SERVER['HTTP_HOST']) ) {
433	$this->url = 'http://' . $_SERVER['HTTP_HOST'];
434	} else {
435	$this->url = '';
436	}
437	} else {
438	// Strip any trailing slashes for consistency (relative
439	// URLs may already start with a slash like "/file.html")
440	if ( substr($url, -1) == '/' ) {
441	$url = substr($url, 0, -1);
442	}
443	$this->url = $url;
444	}
445	}
446
447	/**
448	* Workhorse function that does actual conversion (calls _converter() method).
449	*
450	* @access private
451	* @return void
452	*/
453	function _convert()
454	{
455	// Variables used for building the link list
456	$this->_link_count = 0;
457	$this->_link_list = '';
458
459	$text = trim(stripslashes($this->html));
460
461	// Convert HTML to TXT
462	$this->_converter($text);
463
464	// Add link list
465	if ( !empty($this->_link_list) ) {
466	$text .= "\n\nLinks:\n------\n" . $this->_link_list;
467	}
468
469	$this->text = $text;
470
471	$this->_converted = true;
472	}
473
474	/**
475	* Workhorse function that does actual conversion.
476	*
477	* First performs custom tag replacement specified by $search and
478	* $replace arrays. Then strips any remaining HTML tags, reduces whitespace
479	* and newlines to a readable format, and word wraps the text to
480	* $width characters.
481	*
482	* @param string Reference to HTML content string
483	*
484	* @access private
485	* @return void
486	*/
487	function _converter(&$text)
488	{
489	// Convert <BLOCKQUOTE> (before PRE!)
490	$this->_convert_blockquotes($text);
491
492	// Convert <PRE>
493	$this->_convert_pre($text);
494
495	// Run our defined search-and-replace
496	$text = preg_replace($this->search, $this->replace, $text);
497
498	// Replace known html entities
499	$text = html_entity_decode($text, ENT_COMPAT, 'UTF-8');
500
501	// Run our defined search-and-replace with callback
502	$text = preg_replace_callback($this->callback_search, array('html2text', '_preg_callback'), $text);
503
504	// Remove unknown/unhandled entities (this cannot be done in search-and-replace block)
505	$text = preg_replace('/&#?[a-z0-9]{2,7};/i', '', $text);
506
507	// Strip any other HTML tags
508	$text = strip_tags($text, $this->allowed_tags);
509
510	// Bring down number of empty lines to 2 max
511	$text = preg_replace("/\n\s+\n/", "\n\n", $text);
512	$text = preg_replace("/[\n]{3,}/", "\n\n", $text);
513
514	// Wrap the text to a readable format
515	// for PHP versions >= 4.0.2. Default width is 75
516	// If width is 0 or less, don't wrap the text.
517	if ( $this->width > 0 ) {
518	$text = wordwrap($text, $this->width);
519	}
520	}
521
522	/**
523	* Helper function called by preg_replace() on link replacement.
524	*
525	* Maintains an internal list of links to be displayed at the end of the
526	* text, with numeric indices to the original point in the text they
527	* appeared. Also makes an effort at identifying and handling absolute
528	* and relative links.
529	*
530	* @param string $link URL of the link
531	* @param string $display Part of the text to associate number with
532	* @access private
533	* @return string
534	*/
535	function _build_link_list( $link, $display )
536	{
537	if ( !$this->_do_links )
538	return $display;
539
540	if ( substr($link, 0, 7) == 'http://' \|\| substr($link, 0, 8) == 'https://' \|\|
541	substr($link, 0, 7) == 'mailto:'
542	) {
543	$this->_link_count++;
544	$this->_link_list .= '[' . $this->_link_count . "] $link\n";
545	$additional = ' [' . $this->_link_count . ']';
546	} elseif ( substr($link, 0, 11) == 'javascript:' ) {
547	// Don't count the link; ignore it
548	$additional = '';
549	// what about href="#anchor" ?
550	} else {
551	$this->_link_count++;
552	$this->_link_list .= '[' . $this->_link_count . '] ' . $this->url;
553	if ( substr($link, 0, 1) != '/' ) {
554	$this->_link_list .= '/';
555	}
556	$this->_link_list .= "$link\n";
557	$additional = ' [' . $this->_link_count . ']';
558	}
559
560	return $display . $additional;
561	}
562
563	/**
564	* Helper function for PRE body conversion.
565	*
566	* @param string HTML content
567	* @access private
568	*/
569	function _convert_pre(&$text)
570	{
571	while (preg_match('/<pre[^>]>(.)<\/pre>/ismU', $text, $matches)) {
572	$result = preg_replace($this->pre_search, $this->pre_replace, $matches[1]);
573	$text = preg_replace('/<pre[^>]>.<\/pre>/ismU', '<div><br>' . $result . '<br></div>', $text, 1);
574	}
575	}
576
577	/**
578	* Helper function for BLOCKQUOTE body conversion.
579	*
580	* @param string HTML content
581	* @access private
582	*/
583	function _convert_blockquotes(&$text)
584	{
585	if (preg_match_all('/<\/blockquote[^>]>/i', $text, $matches, PREG_OFFSET_CAPTURE)) {
586	$level = 0;
587	$diff = 0;
588	foreach ($matches[0] as $m) {
589	if ($m[0][0] == '<' && $m[0][1] == '/') {
590	$level--;
591	if ($level < 0) {
592	$level = 0; // malformed HTML: go to next blockquote
593	}
594	else if ($level > 0) {
595	// skip inner blockquote
596	}
597	else {
598	$end = $m[1];
599	$len = $end - $taglen - $start;
600	// Get blockquote content
601	$body = substr($text, $start + $taglen - $diff, $len);
602
603	// Set text width
604	$p_width = $this->width;
605	if ($this->width > 0) $this->width -= 2;
606	// Convert blockquote content
607	$body = trim($body);
608	$this->_converter($body);
609	// Add citation markers and create PRE block
610	$body = preg_replace('/((^\|\n)>*)/', '\\1> ', trim($body));
611	$body = '<pre>' . htmlspecialchars($body) . '</pre>';
612	// Re-set text width
613	$this->width = $p_width;
614	// Replace content
615	$text = substr($text, 0, $start - $diff)
616	. $body . substr($text, $end + strlen($m[0]) - $diff);
617
618	$diff = $len + $taglen + strlen($m[0]) - strlen($body);
619	unset($body);
620	}
621	}
622	else {
623	if ($level == 0) {
624	$start = $m[1];
625	$taglen = strlen($m[0]);
626	}
627	$level ++;
628	}
629	}
630	}
631	}
632
633	/**
634	* Callback function for preg_replace_callback use.
635	*
636	* @param array PREG matches
637	* @return string
638	* @access private
639	*/
640	function _preg_callback($matches)
641	{
642	switch($matches[1]) {
643	case 'b':
644	case 'strong':
645	return $this->_strtoupper($matches[2]);
646	case 'th':
647	return $this->_strtoupper("\t\t". $matches[2] ."\n");
648	case 'h':
649	return $this->_strtoupper("\n\n". $matches[2] ."\n\n");
650	case 'a':
651	return $this->_build_link_list($matches[3], $matches[4]);
652	}
653	}
654
655	/**
656	* Strtoupper multibyte wrapper function
657	*
658	* @param string
659	* @return string
660	* @access private
661	*/
662	function _strtoupper($str)
663	{
664	if (function_exists('mb_strtoupper'))
665	return mb_strtoupper($str);
666	else
667	return strtoupper($str);
668	}
669	}

Note: See TracBrowser for help on using the repository browser.

Download in other formats: