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 $htmlBockTags = array(
  62                 'p', 'div', 'blockquote',
  63                 'pre', 'table', 'form', 'ol',
  64                 'ul', 'h1', 'h2', 'h3', 'h4',
  65                 'h5', 'h6', 'iframe', 'script',
  66                 'noscript'
  67         );
  68
  69         // ###################################################################
  70         /**
  71         * Quick parsing function that uses the system defaults for parsing.
  72         *
  73         * @param        string  Original text
  74         *
  75         * @return       string  HTML output
  76         */
  77         public static function Parse($text)
  78         {
  79                 $parser = new BSMarkdown();
  80                 return $parser->transform($text);
  81         }
  82
  83         // ###################################################################
  84         /**
  85         * Resets the parser
  86         */
  87         public function __construct()
  88         {
  89                 $this->htmlBlockMap = array();
  90                 $this->text = '';
  91         }
  92
  93         // ###################################################################
  94         /**
  95         * Transforms the Markdown text into XHTML with the parser's set options
  96         *
  97         * @param        string  Text to transform
  98         *
  99         * @return       string  HTML output
 100         */
 101         public function transform($text)
 102         {
 103                 // reset the data arrays
 104                 $this->__construct();
 105
 106                 // convert line breaks and remove empty lines of whitespace
 107                 $this->text = BSFunctions::ConvertLineBreaks($text);
 108                 $this->text = preg_replace('/^\s*?$/m', '', $this->text);
 109
 110                 $this->_extractHtmlBlocks();
 111
 112                 $this->_convertHardLineBreaks();
 113                 $this->_convertAtxHeaders();
 114                 $this->_convertSetextHeaders();
 115                 $this->_formatParagraphs();
 116                 $this->text = $this->_expandHtmlBlocks($this->text);
 117
 118                 // convert entitites
 119                 $this->_convertFloatingEntities();
 120
 121                 return $this->text;
 122         }
 123
 124         // ###################################################################
 125         /**
 126         * Description of the function
 127         *
 128         * @param        string  A string param
 129         *
 130         * @return       integer Return value
 131         */
 132         private function _extractHtmlBlocks()
 133         {
 134                 $start = -1;
 135                 $blockStart = 0;
 136                 $nest = array();
 137                 $matches = array();
 138
 139                 // find the first insance of a block tag
 140                 $regex = implode('|', $this->htmlBockTags);
 141                 while ($start <= strlen($this->text))
 142                 {
 143                         $start++;
 144                         if (preg_match("/^<($regex)/i", substr($this->text, $start)) === 1)
 145                         {
 146                                 if (sizeof($nest) == 0)
 147                                 {
 148                                         $blockStart = $start;
 149                                 }
 150                                 array_push($nest, $start);
 151                         }
 152                         else if (preg_match("#^</($regex)>#i", substr($this->text, $start), $matches) === 1)
 153                         {
 154                                 array_pop($nest);
 155                                 if (sizeof($nest) == 0)
 156                                 {
 157                                         $block = substr($this->text, $blockStart, $start - $blockStart + strlen($matches[0]));
 158                                         $hash = md5($block . microtime());
 159                                         $this->htmlBlockMap[$hash] = $block;
 160                                         $this->text = substr_replace($this->text, $hash, $blockStart, strlen($block));
 161                                         $start = $blockStart;
 162                                 }
 163                         }
 164                 }
 165         }
 166
 167         // ###################################################################
 168         /**
 169         * Expands the hashed HTML blocks back into their originial form
 170         */
 171         private function _expandHtmlBlocks($text)
 172         {
 173                 return str_replace(array_keys($this->htmlBlockMap), array_values($this->htmlBlockMap), $text);
 174         }
 175
 176         // ###################################################################
 177         /**
 178         * Converts text surrounded by #sings to headers (## Heading 2)
 179         */
 180         private function _convertAtxHeaders()
 181         {
 182                 $this->text = preg_replace_callback('/^(\#{1,6})\s*(.+)(\s*\#+)?$/', array(&$this, '_convertAtxHeadersCallback'), $this->text);
 183         }
 184
 185         // ###################################################################
 186         /**
 187         * Callback function for preg_replace() in _convertAtxHeaders()
 188         *
 189         * @param        array   Matches
 190         */
 191         private function _convertAtxHeadersCallback($matches)
 192         {
 193                 var_dump($matches);
 194                 $html = '<h' . strlen($matches[1]) . '>' . $this->_expandHtmlBlocks($matches[2]) . '</h' . strlen($matches[1]) . '>';
 195                 $hash = md5($html . microtime());
 196                 $this->htmlBlockMap[$hash] = $html;
 197                 return $hash;
 198         }
 199
 200         // ###################################################################
 201         /**
 202         * Converts headers that are formed by underlines into headings
 203         */
 204         private function _convertSetextHeaders()
 205         {
 206                 $this->text = preg_replace_callback('/(.+)\n(-|=){1,}$/m', array(&$this, '_convertSetextHeadersCallback'), $this->text);
 207         }
 208
 209         // ###################################################################
 210         /**
 211         * Callback function for _convertSetextHeaders(). This does the actual
 212         * conversion and then hashes it into a block
 213         *
 214         * @param        array   Matches from the preg_replace_callback()
 215         */
 216         private function _convertSetextHeadersCallback($matches)
 217         {
 218                 $text = $this->_expandHtmlBlocks($matches[1]);
 219                 if ($matches[2][0] == '=')
 220                 {
 221                         $text = '<h1>' . $text . '</h1>';
 222                 }
 223                 else
 224                 {
 225                         $text = '<h2>' . $text . '</h2>';
 226                 }
 227                 $hash = md5($text . microtime());
 228                 $this->htmlBlockMap[$hash] = $text;
 229                 return $hash;
 230         }
 231
 232         // ###################################################################
 233         /**
 234         * Wraps blocks into paragraphs
 235         */
 236         private function _formatParagraphs()
 237         {
 238                 $blocks = preg_split('/\n{2,}/', $this->text, -1, PREG_SPLIT_NO_EMPTY);
 239                 foreach ($blocks AS $key => $value)
 240                 {
 241                         if (!isset($this->htmlBlockMap[$value]))
 242                         {
 243                                 $blocks[$key] = "<p>$value</p>";
 244                         }
 245                 }
 246
 247                 $this->text = implode("\n\n", $blocks);
 248         }
 249
 250         // ###################################################################
 251         /**
 252         * Converts two spaces followed by a new line with text on it to
 253         * a hard line break (<br />)
 254         */
 255         private function _convertHardLineBreaks()
 256         {
 257                 $this->text = preg_replace('/ {2,}\n/m', " <br />\n", $this->text);
 258         }
 259
 260         // ###################################################################
 261         /**
 262         * Converts all angle brackets and ampersands that are floating into
 263         * HTML entities
 264         */
 265         private function _convertFloatingEntities()
 266         {
 267                 // encode ampersands
 268                 $this->text = preg_replace('/&(?!#?x?[0-9a-z]+;)/i', '&amp;', $this->text);
 269
 270                 // encode brackets who aren't followed by text
 271                 $this->text = preg_replace('#<(?![/?a-z])#i', '&lt;', $this->text);
 272         }
 273 }
 274
 275 /*=====================================================================*\
 276 || ###################################################################
 277 || # $HeadURL$
 278 || # $Id$
 279 || ###################################################################
 280 \*=====================================================================*/
 281 ?>