Markdown.php

   1 <?php
   2 /*=====================================================================*\
   3 || ###################################################################
   4 || # Blue Static ISSO Framework
   5 || # Copyright ©2002-[#]year[#] Blue Static
   6 || #
   7 || # This program is free software; you can redistribute it and/or modify
   8 || # it under the terms of the GNU General Public License as published by
   9 || # the Free Software Foundation; version [#]gpl[#] of the License.
  10 || #
  11 || # This program is distributed in the hope that it will be useful, but
  12 || # WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
  13 || # or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
  14 || # more details.
  15 || #
  16 || # You should have received a copy of the GNU General Public License along
  17 || # with this program; if not, write to the Free Software Foundation, Inc.,
  18 || # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
  19 || ###################################################################
  20 \*=====================================================================*/
  21
  22 require_once('ISSO/Functions.php');
  23
  24 /**
  25 * Markdown
  26 *
  27 * This class is a PHP implementation of John Gruber's Markdown text processing
  28 * system. It is not a port of his Perl version, but rather a creation based
  29 * on the rules outlined here: http://daringfireball.net/projects/markdown/.
  30 *
  31 * @author               Blue Static
  32 * @copyright    Copyright (c)2002 - [#]year[#], Blue Static
  33 * @version              $Revision$
  34 * @package              ISSO
  35 *
  36 */
  37 class BSMarkdown
  38 {
  39         /**
  40         * Remove HTML rather than ignoring it?
  41         * @var bool
  42         */
  43         public $removeHtml = false;
  44
  45         /**
  46         * Map of HTML blocks extracted from the text
  47         * @var array
  48         */
  49         private $htmlBlockMap = array();
  50
  51         /**
  52         * The working parser text that at any given moment is only partially parsed
  53         * @var string
  54         */
  55         private $text;
  56
  57         /**
  58         * HTML block element tags
  59         * @var array
  60         */
  61         private $htmlBockTags = array(
  62                 'p', 'div', 'blockquote',
  63                 'pre', 'table', 'form', 'ol',
  64                 'ul', 'h1', 'h2', 'h3', 'h4',
  65                 'h5', 'h6', 'iframe', 'script',
  66                 'noscript'
  67         );
  68
  69         /**
  70         * Map of link IDs to their link
  71         * @var string
  72         */
  73         private $linkMap = array();
  74
  75         // ###################################################################
  76         /**
  77         * Quick parsing function that uses the system defaults for parsing.
  78         *
  79         * @param        string  Original text
  80         *
  81         * @return       string  HTML output
  82         */
  83         public static function Parse($text)
  84         {
  85                 $parser = new BSMarkdown();
  86                 return $parser->transform($text);
  87         }
  88
  89         // ###################################################################
  90         /**
  91         * Resets the parser
  92         */
  93         public function __construct()
  94         {
  95                 $this->htmlBlockMap = array();
  96                 $this->text = '';
  97         }
  98
  99         // ###################################################################
 100         /**
 101         * Transforms the Markdown text into XHTML with the parser's set options
 102         *
 103         * @param        string  Text to transform
 104         *
 105         * @return       string  HTML output
 106         */
 107         public function transform($text)
 108         {
 109                 // reset the data arrays
 110                 $this->__construct();
 111
 112                 // convert line breaks and remove empty lines of whitespace
 113                 $this->text = BSFunctions::ConvertLineBreaks($text);
 114                 $this->text = preg_replace('/^\s*?$/m', '', $this->text);
 115
 116                 $this->_extractHtmlBlocks();
 117                 $this->_extractLinkMap();
 118
 119                 $this->_convertHardLineBreaks();
 120                 $this->_convertAtxHeaders();
 121                 $this->_convertSetextHeaders();
 122                 $this->_convertHorizontalRules();
 123                 $this->_formatParagraphs();
 124                 $this->text = $this->_expandHtmlBlocks($this->text);
 125
 126                 // convert entitites
 127                 $this->_convertFloatingEntities();
 128
 129                 return $this->text;
 130         }
 131
 132         // ###################################################################
 133         /**
 134         * Description of the function
 135         *
 136         * @param        string  A string param
 137         *
 138         * @return       integer Return value
 139         */
 140         private function _extractHtmlBlocks()
 141         {
 142                 $start = -1;
 143                 $blockStart = 0;
 144                 $nest = array();
 145                 $matches = array();
 146
 147                 // find the first insance of a block tag
 148                 $regex = implode('|', $this->htmlBockTags);
 149                 while ($start <= strlen($this->text))
 150                 {
 151                         $start++;
 152                         if (preg_match("/^<($regex)/i", substr($this->text, $start)) === 1)
 153                         {
 154                                 if (sizeof($nest) == 0)
 155                                 {
 156                                         $blockStart = $start;
 157                                 }
 158                                 array_push($nest, $start);
 159                         }
 160                         else if (preg_match("#^</($regex)>#i", substr($this->text, $start), $matches) === 1)
 161                         {
 162                                 array_pop($nest);
 163                                 if (sizeof($nest) == 0)
 164                                 {
 165                                         $block = substr($this->text, $blockStart, $start - $blockStart + strlen($matches[0]));
 166                                         $hash = md5($block . microtime());
 167                                         $this->htmlBlockMap[$hash] = $block;
 168                                         $this->text = substr_replace($this->text, $hash, $blockStart, strlen($block));
 169                                         $start = $blockStart;
 170                                 }
 171                         }
 172                 }
 173         }
 174
 175         // ###################################################################
 176         /**
 177         * Expands the hashed HTML blocks back into their originial form
 178         */
 179         private function _expandHtmlBlocks($text)
 180         {
 181                 return str_replace(array_keys($this->htmlBlockMap), array_values($this->htmlBlockMap), $text);
 182         }
 183
 184         // ###################################################################
 185         /**
 186         * Extracts all links in the "[id]: link" form
 187         */
 188         public function _extractLinkMap()
 189         {
 190                 $this->text = preg_replace_callback('/\[(\w+)\]:\s*<?(\S+)>?(\s*(("|\')|\()(.*?)(\4|\)))?\n?/', array(&$this, '_extractLinkMapCallback'), $this->text);
 191         }
 192
 193         // ###################################################################
 194         /**
 195         * Converts extracted link definitions into the map
 196         *
 197         * @param        array   Matches array
 198         */
 199         private function _extractLinkMapCallback($matches)
 200         {
 201                 $this->linkMap[$matches[1]] = array($matches[2], $matches[6]);
 202         }
 203
 204         // ###################################################################
 205         /**
 206         * Converts text surrounded by #sings to headers (## Heading 2)
 207         */
 208         private function _convertAtxHeaders()
 209         {
 210                 $this->text = preg_replace_callback('/^(\#{1,6})\s*(.+)(\s*\#+)?$/', array(&$this, '_convertAtxHeadersCallback'), $this->text);
 211         }
 212
 213         // ###################################################################
 214         /**
 215         * Callback function for preg_replace() in _convertAtxHeaders()
 216         *
 217         * @param        array   Matches
 218         */
 219         private function _convertAtxHeadersCallback($matches)
 220         {
 221                 // var_dump($matches);
 222                 $html = '<h' . strlen($matches[1]) . '>' . $this->_expandHtmlBlocks($matches[2]) . '</h' . strlen($matches[1]) . '>';
 223                 $hash = md5($html . microtime());
 224                 $this->htmlBlockMap[$hash] = $html;
 225                 return $hash;
 226         }
 227
 228         // ###################################################################
 229         /**
 230         * Converts headers that are formed by underlines into headings
 231         */
 232         private function _convertSetextHeaders()
 233         {
 234                 $this->text = preg_replace_callback('/(.+)\n(-|=){1,}$/m', array(&$this, '_convertSetextHeadersCallback'), $this->text);
 235         }
 236
 237         // ###################################################################
 238         /**
 239         * Callback function for _convertSetextHeaders(). This does the actual
 240         * conversion and then hashes it into a block
 241         *
 242         * @param        array   Matches from the preg_replace_callback()
 243         */
 244         private function _convertSetextHeadersCallback($matches)
 245         {
 246                 $text = $this->_expandHtmlBlocks($matches[1]);
 247                 if ($matches[2][0] == '=')
 248                 {
 249                         $text = '<h1>' . $text . '</h1>';
 250                 }
 251                 else
 252                 {
 253                         $text = '<h2>' . $text . '</h2>';
 254                 }
 255                 $hash = md5($text . microtime());
 256                 $this->htmlBlockMap[$hash] = $text;
 257                 return $hash;
 258         }
 259
 260         // ###################################################################
 261         /**
 262         * Converts three stars or dashes (optionally separated by spaces) into
 263         * a horizontal rule.
 264         */
 265         private function _convertHorizontalRules()
 266         {
 267                 $hash = md5('<hr />' . microtime());
 268                 $count = 0;
 269                 $this->text = preg_replace('/((-|\*) ?){3,}/', $hash, $this->text, -1, $count);
 270                 if ($count > 0)
 271                 {
 272                         $this->htmlBlockMap[$hash] = '<hr />';
 273                 }
 274         }
 275
 276         // ###################################################################
 277         /**
 278         * Wraps blocks into paragraphs
 279         */
 280         private function _formatParagraphs()
 281         {
 282                 $blocks = preg_split('/\n{2,}/', $this->text, -1, PREG_SPLIT_NO_EMPTY);
 283                 foreach ($blocks AS $key => $value)
 284                 {
 285                         if (!isset($this->htmlBlockMap[$value]))
 286                         {
 287                                 $blocks[$key] = "<p>$value</p>";
 288                         }
 289                 }
 290
 291                 $this->text = implode("\n\n", $blocks);
 292         }
 293
 294         // ###################################################################
 295         /**
 296         * Converts two spaces followed by a new line with text on it to
 297         * a hard line break (<br />)
 298         */
 299         private function _convertHardLineBreaks()
 300         {
 301                 $this->text = preg_replace('/ {2,}\n/m', " <br />\n", $this->text);
 302         }
 303
 304         // ###################################################################
 305         /**
 306         * Converts all angle brackets and ampersands that are floating into
 307         * HTML entities
 308         */
 309         private function _convertFloatingEntities()
 310         {
 311                 // encode ampersands
 312                 $this->text = preg_replace('/&(?!#?x?[0-9a-z]+;)/i', '&amp;', $this->text);
 313
 314                 // encode brackets who aren't followed by text
 315                 $this->text = preg_replace('#<(?![/?a-z])#i', '&lt;', $this->text);
 316         }
 317 }
 318
 319 /*=====================================================================*\
 320 || ###################################################################
 321 || # $HeadURL$
 322 || # $Id$
 323 || ###################################################################
 324 \*=====================================================================*/
 325 ?>