3 /*************************************************************************
5 * class.html2text.inc *
7 *************************************************************************
9 * Converts HTML to formatted plain text *
11 * Copyright (c) 2005-2007 Jon Abernathy <jon@chuggnutt.com> *
12 * All rights reserved. *
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. *
19 * The GNU General Public License can be found at *
20 * http://www.gnu.org/copyleft/gpl.html. *
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. *
27 * Author(s): Jon Abernathy <jon@chuggnutt.com> *
29 * Last modified: 08/08/07 *
31 *************************************************************************/
35 * Takes HTML and converts it to formatted, plain text.
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.
40 * Updated set_html() function's file reading mechanism, 9/25/03.
42 * Thanks to Joss Sanglier (http://www.dancingbear.co.uk/) for adding
43 * several more HTML entity codes to the $search and $replace arrays.
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.
50 * Thanks to Justin Dearing for pointing out that a replacement for the
51 * <TH> tag was missing, and suggesting an appropriate fix.
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.
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.
66 * Thanks to Jacob Chandler for pointing out another link condition
67 * for the _build_link_list() function: "https".
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.
75 * *** Big housecleaning updates below:
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
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.
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.
92 * Thanks to Daniel Schledermann (http://www.typoconsult.dk/) for
93 * suggesting a valuable fix with <a> tag handling.
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
101 * *** End of the housecleaning updates. Updated 08/08/07.
103 * Modified 2015-10-04 by Rod Roark <rod@sunsetsystems.com>
104 * Added an argument to the _convert function to allow omission of the
107 * @author Jon Abernathy <jon@chuggnutt.com>
115 * Contains the HTML content to convert.
123 * Contains the converted, formatted text.
131 * Maximum width of the formatted text, in columns.
133 * Set this value to 0 (or less) to ignore word wrapping
134 * and not constrain text to a fixed-width column.
136 * @var integer $width
142 * List of preg* regular expression patterns to search for,
143 * used in conjunction with $replace.
150 "/\r/", // Non-legal carriage return
151 "/[\n\t]+/", // Newlines and tabs
152 '/[ ]{2,}/', // Runs of spaces, pre-handling
153 '/<script[^>]*>.*?<\/script>/i', // <script>s -- which strip_tags supposedly has problems with
154 '/<style[^>]*>.*?<\/style>/i', // <style>s -- which strip_tags supposedly has problems with
155 //'/<!-- .* -->/', // Comments -- which strip_tags might have problem a with
156 '/<h[123][^>]*>(.*?)<\/h[123]>/ie', // H1 - H3
157 '/<h[456][^>]*>(.*?)<\/h[456]>/ie', // H4 - H6
158 '/<p[^>]*>/i', // <P>
159 '/<br[^>]*>/i', // <br>
160 '/<b[^>]*>(.*?)<\/b>/ie', // <b>
161 '/<strong[^>]*>(.*?)<\/strong>/ie', // <strong>
162 '/<i[^>]*>(.*?)<\/i>/i', // <i>
163 '/<em[^>]*>(.*?)<\/em>/i', // <em>
164 '/(<ul[^>]*>|<\/ul>)/i', // <ul> and </ul>
165 '/(<ol[^>]*>|<\/ol>)/i', // <ol> and </ol>
166 '/<li[^>]*>(.*?)<\/li>/i', // <li> and </li>
167 '/<li[^>]*>/i', // <li>
168 '/<a [^>]*href="([^"]+)"[^>]*>(.*?)<\/a>/ie',
170 '/<hr[^>]*>/i', // <hr>
171 '/(<table[^>]*>|<\/table>)/i', // <table> and </table>
172 '/(<tr[^>]*>|<\/tr>)/i', // <tr> and </tr>
173 '/<td[^>]*>(.*?)<\/td>/i', // <td> and </td>
174 '/<th[^>]*>(.*?)<\/th>/ie', // <th> and </th>
175 '/&(nbsp|#160);/i', // Non-breaking space
176 '/&(quot|rdquo|ldquo|#8220|#8221|#147|#148);/i',
178 '/&(apos|rsquo|lsquo|#8216|#8217);/i', // Single quotes
179 '/>/i', // Greater-than
180 '/</i', // Less-than
181 '/&(amp|#38);/i', // Ampersand
182 '/&(copy|#169);/i', // Copyright
183 '/&(trade|#8482|#153);/i', // Trademark
184 '/&(reg|#174);/i', // Registered
185 '/&(mdash|#151|#8212);/i', // mdash
186 '/&(ndash|minus|#8211|#8722);/i', // ndash
187 '/&(bull|#149|#8226);/i', // Bullet
188 '/&(pound|#163);/i', // Pound sign
189 '/&(euro|#8364);/i', // Euro sign
190 '/&[^&;]+;/i', // Unknown/unhandled entities
191 '/[ ]{2,}/' // Runs of spaces, post-handling
195 * List of pattern replacements corresponding to patterns searched.
197 * @var array $replace
201 var $replace = array(
202 '', // Non-legal carriage return
203 ' ', // Newlines and tabs
204 ' ', // Runs of spaces, pre-handling
205 '', // <script>s -- which strip_tags supposedly has problems with
206 '', // <style>s -- which strip_tags supposedly has problems with
207 //'', // Comments -- which strip_tags might have problem a with
208 "strtoupper(\"\n\n\\1\n\n\")", // H1 - H3
209 "ucwords(\"\n\n\\1\n\n\")", // H4 - H6
212 'strtoupper("\\1")', // <b>
213 'strtoupper("\\1")', // <strong>
216 "\n\n", // <ul> and </ul>
217 "\n\n", // <ol> and </ol>
218 "\t* \\1\n", // <li> and </li>
220 '$this->_build_link_list("\\1", "\\2")',
222 "\n-------------------------\n", // <hr>
223 "\n\n", // <table> and </table>
224 "\n", // <tr> and </tr>
225 "\t\t\\1\n", // <td> and </td>
226 "strtoupper(\"\t\t\\1\n\")", // <th> and </th>
227 ' ', // Non-breaking space
228 '"', // Double quotes
229 "'", // Single quotes
240 'EUR', // Euro sign. € ?
241 '', // Unknown/unhandled entities
242 ' ' // Runs of spaces, post-handling
246 * Contains a list of HTML tags to allow in the resulting text.
248 * @var string $allowed_tags
250 * @see set_allowed_tags()
252 var $allowed_tags = '';
255 * Contains the base URL that relative links should resolve to.
263 * Indicates whether content in the $html variable has been converted yet.
265 * @var boolean $_converted
269 var $_converted = false;
272 * Contains URL addresses from links to be rendered in plain text.
274 * @var string $_link_list
276 * @see _build_link_list()
278 var $_link_list = '';
281 * Number of valid links detected in the text, used for plain text
282 * display (rendered similar to footnotes).
284 * @var integer $_link_count
286 * @see _build_link_list()
288 var $_link_count = 0;
293 * If the HTML source string (or file) is supplied, the class
294 * will instantiate with that source propagated, all that has
295 * to be done it to call get_text().
297 * @param string $source HTML content
298 * @param boolean $from_file Indicates $source is a file to pull content from
302 function __construct($source = '', $from_file = false)
304 if (!empty($source)) {
305 $this->set_html($source, $from_file);
308 $this->set_base_url();
312 * Loads source HTML into memory, either from $source string or a file.
314 * @param string $source HTML content
315 * @param boolean $from_file Indicates $source is a file to pull content from
319 function set_html($source, $from_file = false)
321 $this->html
= $source;
323 if ($from_file && file_exists($source)) {
324 $fp = fopen($source, 'r');
325 $this->html
= fread($fp, filesize($source));
329 $this->_converted
= false;
333 * Returns the text, converted from HTML.
340 if (!$this->_converted
) {
348 * Prints the text, converted from HTML.
353 function print_text()
355 print $this->get_text();
359 * Alias to print_text(), operates identically.
367 print $this->get_text();
371 * Sets the allowed HTML tags to pass through to the resulting text.
373 * Tags should be in the form "<p>", with no corresponding closing tag.
378 function set_allowed_tags($allowed_tags = '')
380 if (!empty($allowed_tags)) {
381 $this->allowed_tags
= $allowed_tags;
386 * Sets a base URL to handle relative links.
391 function set_base_url($url = '')
394 if (!empty($_SERVER['HTTP_HOST'])) {
395 $this->url
= 'http://' . $_SERVER['HTTP_HOST'];
400 // Strip any trailing slashes for consistency (relative
401 // URLs may already start with a slash like "/file.html")
402 if (substr($url, -1) == '/') {
403 $url = substr($url, 0, -1);
411 * Workhorse function that does actual conversion.
413 * First performs custom tag replacement specified by $search and
414 * $replace arrays. Then strips any remaining HTML tags, reduces whitespace
415 * and newlines to a readable format, and word wraps the text to
421 function _convert($ll = true)
423 // Variables used for building the link list
424 $this->_link_count
= 0;
425 $this->_link_list
= '';
427 $text = trim(stripslashes($this->html
));
429 // Run our defined search-and-replace
430 $text = preg_replace($this->search
, $this->replace
, $text);
432 // Strip any other HTML tags
433 $text = strip_tags($text, $this->allowed_tags
);
435 // Bring down number of empty lines to 2 max
436 $text = preg_replace("/\n\s+\n/", "\n\n", $text);
437 $text = preg_replace("/[\n]{3,}/", "\n\n", $text);
440 if ($ll && !empty($this->_link_list
)) {
441 $text .= "\n\nLinks:\n------\n" . $this->_link_list
;
444 // Wrap the text to a readable format
445 // for PHP versions >= 4.0.2. Default width is 75
446 // If width is 0 or less, don't wrap the text.
447 if ($this->width
> 0) {
448 $text = wordwrap($text, $this->width
);
453 $this->_converted
= true;
457 * Helper function called by preg_replace() on link replacement.
459 * Maintains an internal list of links to be displayed at the end of the
460 * text, with numeric indices to the original point in the text they
461 * appeared. Also makes an effort at identifying and handling absolute
462 * and relative links.
464 * @param string $link URL of the link
465 * @param string $display Part of the text to associate number with
469 function _build_link_list($link, $display)
471 if (substr($link, 0, 7) == 'http://' ||
substr($link, 0, 8) == 'https://' ||
472 substr($link, 0, 7) == 'mailto:' ) {
473 $this->_link_count++
;
474 $this->_link_list
.= "[" . $this->_link_count
. "] $link\n";
475 $additional = ' [' . $this->_link_count
. ']';
476 } elseif (substr($link, 0, 11) == 'javascript:') {
477 // Don't count the link; ignore it
479 // what about href="#anchor" ?
481 $this->_link_count++
;
482 $this->_link_list
.= "[" . $this->_link_count
. "] " . $this->url
;
483 if (substr($link, 0, 1) != '/') {
484 $this->_link_list
.= '/';
487 $this->_link_list
.= "$link\n";
488 $additional = ' [' . $this->_link_count
. ']';
491 return $display . $additional;