├── README.md ├── demo ├── index.php └── style.css ├── extension-meta.php └── markdown.php /README.md: -------------------------------------------------------------------------------- 1 | WordPress Package Parser 2 | ======================== 3 | 4 | A PHP library for parsing WordPress plugin and theme metadata. Point it at a ZIP package and it will: 5 | 6 | - Tell you whether it contains a plugin or a theme. 7 | - Give you the metadata from the comment header (Version, Description, Author URI, etc). 8 | - Parse readme.txt into a list of headers and sections. 9 | - Convert readme.txt contents from Markdown to HTML (optional). 10 | 11 | Basic usage 12 | ----------- 13 | 14 | Extract plugin metadata: 15 | 16 | ```php 17 | require 'wp-extension-meta/extension-meta.php'; 18 | $package = WshWordPressPackageParser::parsePackage('sample-plugin.zip', true); 19 | print_r($package); 20 | ``` 21 | 22 | Sample output: 23 | 24 | ``` 25 | Array 26 | ( 27 | [header] => Array 28 | ( 29 | [Name] => Plugin Name 30 | [PluginURI] => http://example.com/ 31 | [Version] => 1.7 32 | [Description] => This plugin does stuff. 33 | [Author] => Yahnis Elsts 34 | [AuthorURI] => http://w-shadow.com/ 35 | [TextDomain] => sample-plugin 36 | [DomainPath] => 37 | [Network] => 38 | [Title] => Plugin Name 39 | ) 40 | 41 | [readme] => Array 42 | ( 43 | [name] => Plugin Name 44 | [contributors] => Array 45 | ( 46 | [0] => whiteshadow 47 | ) 48 | 49 | [tags] => Array 50 | ( 51 | [0] => sample 52 | [1] => tag 53 | [2] => stuff 54 | ) 55 | 56 | [requires] => 3.2 57 | [tested] => 3.5.1 58 | [stable] => 1.7 59 | [short_description] => This is the short description from the readme. 60 | [sections] => Array 61 | ( 62 | [Description] => This is the Description section of the readme. 63 | [Installation] => ... 64 | [Changelog] => ... 65 | ) 66 | ) 67 | [pluginFile] => sample-plugin/sample-plugin.php 68 | [stylesheet] => 69 | [type] => plugin 70 | ) 71 | ``` 72 | 73 | Requirements 74 | ------------ 75 | PHP 5.2. 76 | 77 | Credits 78 | ------- 79 | Partially based on plugin header parsing code from the WordPress core. 80 | -------------------------------------------------------------------------------- /demo/index.php: -------------------------------------------------------------------------------- 1 | 2 | 3 | Extract Plugin Meta 4 | 5 | 6 | 7 | 8 | 9 |
10 | 11 | 12 |
13 | 24 |
25 | 31 | 32 |
33 | 34 |
35 | 36 |
37 | 38 |
39 | 40 |
41 | $sectionContent) { 44 | printf( 45 | "

%s

\n%s\n", 46 | ucwords(str_replace('_', ' ', $sectionHeader)), 47 | $sectionContent 48 | ); 49 | } 50 | } else { 51 | ?> 52 | No sections found in readme.txt 53 | 56 |
57 | 58 |
59 | $sectionContent) { 62 | printf( 63 | "

%s

", 64 | ucwords(str_replace('_', ' ', $sectionHeader)), 65 | htmlentities($sectionContent) 66 | ); 67 | } 68 | } else { 69 | ?> 70 | No sections found in readme.txt 71 | 74 |
75 |
76 | 77 | 82 | 83 | 92 | 93 | -------------------------------------------------------------------------------- /demo/style.css: -------------------------------------------------------------------------------- 1 | .ui-tabs-nav { 2 | margin: 0; 3 | padding: 0 0 0.2em 0; 4 | } 5 | 6 | .ui-tabs-nav li { 7 | display: inline-block; 8 | } 9 | 10 | .ui-tabs-nav li a { 11 | display: inline-block; 12 | padding: 0.5em; 13 | border: 1px solid #bbb; 14 | font-weight: bold; 15 | color: #bbb; 16 | background: #f5f5f5; 17 | text-decoration: none; 18 | 19 | border-radius: 5px; 20 | -moz-border-radius: 5px; 21 | text-shadow: #ffffff 1px 1px 1px; 22 | } 23 | 24 | .ui-tabs-nav .ui-state-hover a { 25 | color: #888; 26 | } 27 | 28 | .ui-tabs-nav .ui-tabs-selected a { 29 | color: black; 30 | background: white; 31 | } 32 | 33 | .ui-tabs-hide { 34 | display: none; 35 | } 36 | 37 | textarea { 38 | width: 70em; 39 | height: 30em; 40 | } -------------------------------------------------------------------------------- /extension-meta.php: -------------------------------------------------------------------------------- 1 | open($packageFilename) !== true ){ 34 | return false; 35 | } 36 | 37 | //Find and parse the plugin or theme file and (optionally) readme.txt. 38 | $header = null; 39 | $readme = null; 40 | $pluginFile = null; 41 | $stylesheet = null; 42 | $type = null; 43 | 44 | for ( $fileIndex = 0; ($fileIndex < $zip->numFiles) && (empty($readme) || empty($header)); $fileIndex++ ){ 45 | $info = $zip->statIndex($fileIndex); 46 | 47 | //Normalize filename: convert backslashes to slashes, remove leading slashes. 48 | $fileName = trim(str_replace('\\', '/', $info['name']), '/'); 49 | $fileName = ltrim($fileName, '/'); 50 | 51 | $parts = explode('.', $fileName); 52 | $extension = strtolower(end($parts)); 53 | $depth = substr_count($fileName, '/'); 54 | 55 | //Skip empty files, directories and everything that's more than 1 sub-directory deep. 56 | if ( ($depth > 1) || ($info['size'] == 0) ) { 57 | continue; 58 | } 59 | 60 | //readme.txt (for plugins)? 61 | if ( empty($readme) && (strtolower(basename($fileName)) == 'readme.txt') ){ 62 | //Try to parse the readme. 63 | $readme = self::parseReadme($zip->getFromIndex($fileIndex), $applyMarkdown); 64 | } 65 | 66 | //Theme stylesheet? 67 | if ( empty($header) && (strtolower(basename($fileName)) == 'style.css') ) { 68 | $fileContents = substr($zip->getFromIndex($fileIndex), 0, 8*1024); 69 | $header = self::getThemeHeaders($fileContents); 70 | if ( !empty($header) ){ 71 | $stylesheet = $fileName; 72 | $type = 'theme'; 73 | } 74 | } 75 | 76 | //Main plugin file? 77 | if ( empty($header) && ($extension === 'php') ){ 78 | $fileContents = substr($zip->getFromIndex($fileIndex), 0, 8*1024); 79 | $header = self::getPluginHeaders($fileContents); 80 | if ( !empty($header) ){ 81 | $pluginFile = $fileName; 82 | $type = 'plugin'; 83 | } 84 | } 85 | } 86 | 87 | if ( empty($type) ){ 88 | return false; 89 | } else { 90 | return compact('header', 'readme', 'pluginFile', 'stylesheet', 'type'); 91 | } 92 | } 93 | 94 | /** 95 | * Parse a plugin's readme.txt to extract various plugin metadata. 96 | * 97 | * Returns an array with the following fields: 98 | * 'name' - Name of the plugin. 99 | * 'contributors' - An array of wordpress.org usernames. 100 | * 'donate' - The plugin's donation link. 101 | * 'tags' - An array of the plugin's tags. 102 | * 'requires' - The minimum version of WordPress that the plugin will run on. 103 | * 'tested' - The latest version of WordPress that the plugin has been tested on. 104 | * 'stable' - The SVN tag of the latest stable release, or 'trunk'. 105 | * 'short_description' - The plugin's "short description". 106 | * 'sections' - An associative array of sections present in the readme.txt. 107 | * Case and formatting of section headers will be preserved. 108 | * 109 | * Be warned that this function does *not* perfectly emulate the way that WordPress.org 110 | * parses plugin readme's. In particular, it may mangle certain HTML markup that wp.org 111 | * handles correctly. 112 | * 113 | * @see http://wordpress.org/extend/plugins/about/readme.txt 114 | * 115 | * @param string $readmeTxtContents The contents of a plugin's readme.txt file. 116 | * @param bool $applyMarkdown Whether to transform Markdown used in readme.txt sections to HTML. Defaults to false. 117 | * @return array|null Associative array, or NULL if the input isn't a valid readme.txt file. 118 | */ 119 | public static function parseReadme($readmeTxtContents, $applyMarkdown = false){ 120 | $readmeTxtContents = trim($readmeTxtContents, " \t\n\r"); 121 | $readme = array( 122 | 'name' => '', 123 | 'contributors' => array(), 124 | 'donate' => '', 125 | 'tags' => array(), 126 | 'requires' => '', 127 | 'tested' => '', 128 | 'stable' => '', 129 | 'short_description' => '', 130 | 'sections' => array(), 131 | ); 132 | 133 | //The readme.txt header has a fairly fixed structure, so we can parse it line-by-line 134 | $lines = explode("\n", $readmeTxtContents); 135 | //Plugin name is at the very top, e.g. === My Plugin === 136 | if ( preg_match('@===\s*(.+?)\s*===@', array_shift($lines), $matches) ){ 137 | $readme['name'] = $matches[1]; 138 | } else { 139 | return null; 140 | } 141 | 142 | //Then there's a bunch of meta fields formatted as "Field: value" 143 | $headers = array(); 144 | $headerMap = array( 145 | 'Contributors' => 'contributors', 146 | 'Donate link' => 'donate', 147 | 'Tags' => 'tags', 148 | 'Requires at least' => 'requires', 149 | 'Tested up to' => 'tested', 150 | 'Stable tag' => 'stable', 151 | ); 152 | do { //Parse each readme.txt header 153 | $pieces = explode(':', array_shift($lines), 2); 154 | if ( array_key_exists($pieces[0], $headerMap) ){ 155 | if ( isset($pieces[1]) ){ 156 | $headers[ $headerMap[$pieces[0]] ] = trim($pieces[1]); 157 | } else { 158 | $headers[ $headerMap[$pieces[0]] ] = ''; 159 | } 160 | } 161 | } while ( trim($pieces[0]) != '' ); //Until an empty line is encountered 162 | 163 | //"Contributors" is a comma-separated list. Convert it to an array. 164 | if ( !empty($headers['contributors']) ){ 165 | $headers['contributors'] = array_map('trim', explode(',', $headers['contributors'])); 166 | } 167 | 168 | //Likewise for "Tags" 169 | if ( !empty($headers['tags']) ){ 170 | $headers['tags'] = array_map('trim', explode(',', $headers['tags'])); 171 | } 172 | 173 | $readme = array_merge($readme, $headers); 174 | 175 | //After the headers comes the short description 176 | $readme['short_description'] = array_shift($lines); 177 | 178 | //Finally, a valid readme.txt also contains one or more "sections" identified by "== Section Name ==" 179 | $sections = array(); 180 | $contentBuffer = array(); 181 | $currentSection = ''; 182 | foreach($lines as $line){ 183 | //Is this a section header? 184 | if ( preg_match('@^\s*==\s+(.+?)\s+==\s*$@m', $line, $matches) ){ 185 | //Flush the content buffer for the previous section, if any 186 | if ( !empty($currentSection) ){ 187 | $sectionContent = trim(implode("\n", $contentBuffer)); 188 | $sections[$currentSection] = $sectionContent; 189 | } 190 | //Start reading a new section 191 | $currentSection = $matches[1]; 192 | $contentBuffer = array(); 193 | } else { 194 | //Buffer all section content 195 | $contentBuffer[] = $line; 196 | } 197 | } 198 | //Flush the buffer for the last section 199 | if ( !empty($currentSection) ){ 200 | $sections[$currentSection] = trim(implode("\n", $contentBuffer)); 201 | } 202 | 203 | //Apply Markdown to sections 204 | if ( $applyMarkdown ){ 205 | $sections = array_map(__CLASS__ . '::applyMarkdown', $sections); 206 | } 207 | 208 | //This is only necessary if you intend to later json_encode() the sections. 209 | //json_encode() may encode certain strings as NULL if they're not in UTF-8. 210 | $sections = array_map('utf8_encode', $sections); 211 | 212 | $readme['sections'] = $sections; 213 | 214 | return $readme; 215 | } 216 | 217 | /** @noinspection PhpUnusedPrivateMethodInspection Actually used in parseReadme(). */ 218 | /** 219 | * Transform Markdown markup to HTML. 220 | * 221 | * Tries (in vain) to emulate the transformation that WordPress.org applies to readme.txt files. 222 | * 223 | * @param string $text 224 | * @return string 225 | */ 226 | private static function applyMarkdown($text){ 227 | //The WP standard for readme files uses some custom markup, like "= H4 headers =" 228 | $text = preg_replace('@^\s*=\s*(.+?)\s*=\s*$@m', "

$1

\n", $text); 229 | return Markdown($text); 230 | } 231 | 232 | /** 233 | * Parse the plugin contents to retrieve plugin's metadata headers. 234 | * 235 | * Adapted from the get_plugin_data() function used by WordPress. 236 | * Returns an array that contains the following: 237 | * 'Name' - Name of the plugin. 238 | * 'Title' - Title of the plugin and the link to the plugin's web site. 239 | * 'Description' - Description of what the plugin does and/or notes from the author. 240 | * 'Author' - The author's name. 241 | * 'AuthorURI' - The author's web site address. 242 | * 'Version' - The plugin version number. 243 | * 'PluginURI' - Plugin web site address. 244 | * 'TextDomain' - Plugin's text domain for localization. 245 | * 'DomainPath' - Plugin's relative directory path to .mo files. 246 | * 'Network' - Boolean. Whether the plugin can only be activated network wide. 247 | * 248 | * If the input string doesn't appear to contain a valid plugin header, the function 249 | * will return NULL. 250 | * 251 | * @param string $fileContents Contents of the plugin file 252 | * @return array|null See above for description. 253 | */ 254 | public static function getPluginHeaders($fileContents) { 255 | //[Internal name => Name used in the plugin file] 256 | $pluginHeaderNames = array( 257 | 'Name' => 'Plugin Name', 258 | 'PluginURI' => 'Plugin URI', 259 | 'Version' => 'Version', 260 | 'Description' => 'Description', 261 | 'Author' => 'Author', 262 | 'AuthorURI' => 'Author URI', 263 | 'TextDomain' => 'Text Domain', 264 | 'DomainPath' => 'Domain Path', 265 | 'Network' => 'Network', 266 | //Site Wide Only is deprecated in favor of Network. 267 | '_sitewide' => 'Site Wide Only', 268 | ); 269 | 270 | $headers = self::getFileHeaders($fileContents, $pluginHeaderNames); 271 | 272 | //Site Wide Only is the old header for Network. 273 | if ( empty($headers['Network']) && !empty($headers['_sitewide']) ) { 274 | $headers['Network'] = $headers['_sitewide']; 275 | } 276 | unset($headers['_sitewide']); 277 | $headers['Network'] = (strtolower($headers['Network']) === 'true'); 278 | 279 | //For backward compatibility by default Title is the same as Name. 280 | $headers['Title'] = $headers['Name']; 281 | 282 | //If it doesn't have a name, it's probably not a plugin. 283 | if ( empty($headers['Name']) ){ 284 | return null; 285 | } else { 286 | return $headers; 287 | } 288 | } 289 | 290 | /** 291 | * Parse the theme stylesheet to retrieve its metadata headers. 292 | * 293 | * Adapted from the get_theme_data() function and the WP_Theme class in WordPress. 294 | * Returns an array that contains the following: 295 | * 'Name' - Name of the theme. 296 | * 'Description' - Theme description. 297 | * 'Author' - The author's name 298 | * 'AuthorURI' - The authors web site address. 299 | * 'Version' - The theme version number. 300 | * 'ThemeURI' - Theme web site address. 301 | * 'Template' - The slug of the parent theme. Only applies to child themes. 302 | * 'Status' - Unknown. Included for completeness. 303 | * 'Tags' - An array of tags. 304 | * 'TextDomain' - Theme's text domain for localization. 305 | * 'DomainPath' - Theme's relative directory path to .mo files. 306 | * 307 | * If the input string doesn't appear to contain a valid theme header, the function 308 | * will return NULL. 309 | * 310 | * @param string $fileContents Contents of the theme stylesheet. 311 | * @return array|null See above for description. 312 | */ 313 | public static function getThemeHeaders($fileContents) { 314 | $themeHeaderNames = array( 315 | 'Name' => 'Theme Name', 316 | 'ThemeURI' => 'Theme URI', 317 | 'Description' => 'Description', 318 | 'Author' => 'Author', 319 | 'AuthorURI' => 'Author URI', 320 | 'Version' => 'Version', 321 | 'Template' => 'Template', 322 | 'Status' => 'Status', 323 | 'Tags' => 'Tags', 324 | 'TextDomain' => 'Text Domain', 325 | 'DomainPath' => 'Domain Path', 326 | 'DetailsURI' => 'Details URI', 327 | ); 328 | $headers = self::getFileHeaders($fileContents, $themeHeaderNames); 329 | 330 | $headers['Tags'] = array_filter(array_map('trim', explode(',', strip_tags( $headers['Tags'])))); 331 | 332 | //If it doesn't have a name, it's probably not a valid theme. 333 | if ( empty($headers['Name']) ){ 334 | return null; 335 | } else { 336 | return $headers; 337 | } 338 | } 339 | 340 | /** 341 | * Parse the file contents to retrieve its metadata. 342 | * 343 | * Searches for metadata for a file, such as a plugin or theme. Each piece of 344 | * metadata must be on its own line. For a field spanning multiple lines, it 345 | * must not have any newlines or only parts of it will be displayed. 346 | * 347 | * @param string $fileContents File contents. Can be safely truncated to 8kiB as that's all WP itself scans. 348 | * @param array $headerMap The list of headers to search for in the file. 349 | * @return array 350 | */ 351 | public static function getFileHeaders($fileContents, $headerMap ) { 352 | $headers = array(); 353 | 354 | //Support systems that use CR as a line ending. 355 | $fileContents = str_replace("\r", "\n", $fileContents); 356 | 357 | foreach ($headerMap as $field => $prettyName) { 358 | $found = preg_match('/^[ \t\/*#@]*' . preg_quote($prettyName, '/') . ':(.*)$/mi', $fileContents, $matches); 359 | if ( ($found > 0) && !empty($matches[1]) ) { 360 | //Strip comment markers and closing PHP tags. 361 | $value = trim(preg_replace("/\s*(?:\*\/|\?>).*/", '', $matches[1])); 362 | $headers[$field] = $value; 363 | } else { 364 | $headers[$field] = ''; 365 | } 366 | } 367 | 368 | return $headers; 369 | } 370 | } 371 | 372 | /** 373 | * Extract plugin metadata from a plugin's ZIP file and transform it into a structure 374 | * compatible with the custom update checker. 375 | * 376 | * Deprecated. Included for backwards-compatibility. 377 | * 378 | * This is an utility function that scans the input file (assumed to be a ZIP archive) 379 | * to find and parse the plugin's main PHP file and readme.txt file. Plugin metadata from 380 | * both files is assembled into an associative array. The structure if this array is 381 | * compatible with the format of the metadata file used by the custom plugin update checker 382 | * library available at the below URL. 383 | * 384 | * @see http://w-shadow.com/blog/2010/09/02/automatic-updates-for-any-plugin/ 385 | * @see https://spreadsheets.google.com/pub?key=0AqP80E74YcUWdEdETXZLcXhjd2w0cHMwX2U1eDlWTHc&authkey=CK7h9toK&hl=en&single=true&gid=0&output=html 386 | * 387 | * Requires the ZIP extension for PHP. 388 | * @see http://php.net/manual/en/book.zip.php 389 | * 390 | * @param string|array $packageInfo Either path to a ZIP file containing a WP plugin, or the return value of analysePluginPackage(). 391 | * @return array Associative array 392 | */ 393 | function getPluginPackageMeta($packageInfo){ 394 | if ( is_string($packageInfo) && file_exists($packageInfo) ){ 395 | $packageInfo = WshWordPressPackageParser::parsePackage($packageInfo, true); 396 | } 397 | 398 | $meta = array(); 399 | 400 | if ( isset($packageInfo['header']) && !empty($packageInfo['header']) ){ 401 | $mapping = array( 402 | 'Name' => 'name', 403 | 'Version' => 'version', 404 | 'PluginURI' => 'homepage', 405 | 'Author' => 'author', 406 | 'AuthorURI' => 'author_homepage', 407 | ); 408 | foreach($mapping as $headerField => $metaField){ 409 | if ( array_key_exists($headerField, $packageInfo['header']) && !empty($packageInfo['header'][$headerField]) ){ 410 | $meta[$metaField] = $packageInfo['header'][$headerField]; 411 | } 412 | } 413 | } 414 | 415 | if ( !empty($packageInfo['readme']) ){ 416 | $mapping = array('requires', 'tested'); 417 | foreach($mapping as $readmeField){ 418 | if ( !empty($packageInfo['readme'][$readmeField]) ){ 419 | $meta[$readmeField] = $packageInfo['readme'][$readmeField]; 420 | } 421 | } 422 | if ( !empty($packageInfo['readme']['sections']) && is_array($packageInfo['readme']['sections']) ){ 423 | foreach($packageInfo['readme']['sections'] as $sectionName => $sectionContent){ 424 | $sectionName = str_replace(' ', '_', strtolower($sectionName)); 425 | $meta['sections'][$sectionName] = $sectionContent; 426 | } 427 | } 428 | 429 | //Check if we have an upgrade notice for this version 430 | if ( isset($meta['sections']['upgrade_notice']) && isset($meta['version']) ){ 431 | $regex = "@

\s*" . preg_quote($meta['version']) . "\s*

[^<>]*?

(.+?)

@i"; 432 | if ( preg_match($regex, $meta['sections']['upgrade_notice'], $matches) ){ 433 | $meta['upgrade_notice'] = trim(strip_tags($matches[1])); 434 | } 435 | } 436 | } 437 | 438 | if ( !empty($packageInfo['pluginFile']) ){ 439 | $meta['slug'] = strtolower(basename(dirname($packageInfo['pluginFile']))); 440 | } 441 | 442 | return $meta; 443 | } 444 | -------------------------------------------------------------------------------- /markdown.php: -------------------------------------------------------------------------------- 1 | 8 | # 9 | # Original Markdown 10 | # Copyright (c) 2004-2006 John Gruber 11 | # 12 | # 13 | 14 | 15 | define( 'MARKDOWN_VERSION', "1.0.1p" ); # Sun 13 Jan 2013 16 | define( 'MARKDOWNEXTRA_VERSION', "1.2.6" ); # Sun 13 Jan 2013 17 | 18 | 19 | # 20 | # Global default settings: 21 | # 22 | 23 | # Change to ">" for HTML output 24 | @define( 'MARKDOWN_EMPTY_ELEMENT_SUFFIX', " />"); 25 | 26 | # Define the width of a tab for code blocks. 27 | @define( 'MARKDOWN_TAB_WIDTH', 4 ); 28 | 29 | # Optional title attribute for footnote links and backlinks. 30 | @define( 'MARKDOWN_FN_LINK_TITLE', "" ); 31 | @define( 'MARKDOWN_FN_BACKLINK_TITLE', "" ); 32 | 33 | # Optional class attribute for footnote links and backlinks. 34 | @define( 'MARKDOWN_FN_LINK_CLASS', "" ); 35 | @define( 'MARKDOWN_FN_BACKLINK_CLASS', "" ); 36 | 37 | # Optional class prefix for fenced code block. 38 | @define( 'MARKDOWN_CODE_CLASS_PREFIX', "" ); 39 | 40 | # Class attribute for code blocks goes on the `code` tag; 41 | # setting this to true will put attributes on the `pre` tag instead. 42 | @define( 'MARKDOWN_CODE_ATTR_ON_PRE', false ); 43 | 44 | 45 | # 46 | # WordPress settings: 47 | # 48 | 49 | # Change to false to remove Markdown from posts and/or comments. 50 | @define( 'MARKDOWN_WP_POSTS', true ); 51 | @define( 'MARKDOWN_WP_COMMENTS', true ); 52 | 53 | 54 | 55 | ### Standard Function Interface ### 56 | 57 | @define( 'MARKDOWN_PARSER_CLASS', 'MarkdownExtra_Parser' ); 58 | 59 | function Markdown($text) { 60 | # 61 | # Initialize the parser and return the result of its transform method. 62 | # 63 | # Setup static parser variable. 64 | static $parser; 65 | if (!isset($parser)) { 66 | $parser_class = MARKDOWN_PARSER_CLASS; 67 | $parser = new $parser_class; 68 | } 69 | 70 | # Transform text using parser. 71 | return $parser->transform($text); 72 | } 73 | 74 | 75 | ### WordPress Plugin Interface ### 76 | 77 | /* 78 | Plugin Name: Markdown Extra 79 | Plugin Name: Markdown 80 | Plugin URI: http://michelf.ca/projects/php-markdown/ 81 | Description: Markdown syntax allows you to write using an easy-to-read, easy-to-write plain text format. Based on the original Perl version by John Gruber. More... 82 | Version: 1.2.6 83 | Author: Michel Fortin 84 | Author URI: http://michelf.ca/ 85 | */ 86 | 87 | if (isset($wp_version)) { 88 | # More details about how it works here: 89 | # 90 | 91 | # Post content and excerpts 92 | # - Remove WordPress paragraph generator. 93 | # - Run Markdown on excerpt, then remove all tags. 94 | # - Add paragraph tag around the excerpt, but remove it for the excerpt rss. 95 | if (MARKDOWN_WP_POSTS) { 96 | remove_filter('the_content', 'wpautop'); 97 | remove_filter('the_content_rss', 'wpautop'); 98 | remove_filter('the_excerpt', 'wpautop'); 99 | add_filter('the_content', 'mdwp_MarkdownPost', 6); 100 | add_filter('the_content_rss', 'mdwp_MarkdownPost', 6); 101 | add_filter('get_the_excerpt', 'mdwp_MarkdownPost', 6); 102 | add_filter('get_the_excerpt', 'trim', 7); 103 | add_filter('the_excerpt', 'mdwp_add_p'); 104 | add_filter('the_excerpt_rss', 'mdwp_strip_p'); 105 | 106 | remove_filter('content_save_pre', 'balanceTags', 50); 107 | remove_filter('excerpt_save_pre', 'balanceTags', 50); 108 | add_filter('the_content', 'balanceTags', 50); 109 | add_filter('get_the_excerpt', 'balanceTags', 9); 110 | } 111 | 112 | # Add a footnote id prefix to posts when inside a loop. 113 | function mdwp_MarkdownPost($text) { 114 | static $parser; 115 | if (!$parser) { 116 | $parser_class = MARKDOWN_PARSER_CLASS; 117 | $parser = new $parser_class; 118 | } 119 | if (is_single() || is_page() || is_feed()) { 120 | $parser->fn_id_prefix = ""; 121 | } else { 122 | $parser->fn_id_prefix = get_the_ID() . "."; 123 | } 124 | return $parser->transform($text); 125 | } 126 | 127 | # Comments 128 | # - Remove WordPress paragraph generator. 129 | # - Remove WordPress auto-link generator. 130 | # - Scramble important tags before passing them to the kses filter. 131 | # - Run Markdown on excerpt then remove paragraph tags. 132 | if (MARKDOWN_WP_COMMENTS) { 133 | remove_filter('comment_text', 'wpautop', 30); 134 | remove_filter('comment_text', 'make_clickable'); 135 | add_filter('pre_comment_content', 'Markdown', 6); 136 | add_filter('pre_comment_content', 'mdwp_hide_tags', 8); 137 | add_filter('pre_comment_content', 'mdwp_show_tags', 12); 138 | add_filter('get_comment_text', 'Markdown', 6); 139 | add_filter('get_comment_excerpt', 'Markdown', 6); 140 | add_filter('get_comment_excerpt', 'mdwp_strip_p', 7); 141 | 142 | global $mdwp_hidden_tags, $mdwp_placeholders; 143 | $mdwp_hidden_tags = explode(' ', 144 | '
    • '); 145 | $mdwp_placeholders = explode(' ', str_rot13( 146 | 'pEj07ZbbBZ U1kqgh4w4p pre2zmeN6K QTi31t9pre ol0MP1jzJR '. 147 | 'ML5IjmbRol ulANi1NsGY J7zRLJqPul liA8ctl16T K9nhooUHli')); 148 | } 149 | 150 | function mdwp_add_p($text) { 151 | if (!preg_match('{^$|^<(p|ul|ol|dl|pre|blockquote)>}i', $text)) { 152 | $text = '

    '.$text.'

    '; 153 | $text = preg_replace('{\n{2,}}', "

    \n\n

    ", $text); 154 | } 155 | return $text; 156 | } 157 | 158 | function mdwp_strip_p($t) { return preg_replace('{}i', '', $t); } 159 | 160 | function mdwp_hide_tags($text) { 161 | global $mdwp_hidden_tags, $mdwp_placeholders; 162 | return str_replace($mdwp_hidden_tags, $mdwp_placeholders, $text); 163 | } 164 | function mdwp_show_tags($text) { 165 | global $mdwp_hidden_tags, $mdwp_placeholders; 166 | return str_replace($mdwp_placeholders, $mdwp_hidden_tags, $text); 167 | } 168 | } 169 | 170 | 171 | ### bBlog Plugin Info ### 172 | 173 | function identify_modifier_markdown() { 174 | return array( 175 | 'name' => 'markdown', 176 | 'type' => 'modifier', 177 | 'nicename' => 'PHP Markdown Extra', 178 | 'description' => 'A text-to-HTML conversion tool for web writers', 179 | 'authors' => 'Michel Fortin and John Gruber', 180 | 'licence' => 'GPL', 181 | 'version' => MARKDOWNEXTRA_VERSION, 182 | 'help' => 'Markdown syntax allows you to write using an easy-to-read, easy-to-write plain text format. Based on the original Perl version by John Gruber. More...', 183 | ); 184 | } 185 | 186 | 187 | ### Smarty Modifier Interface ### 188 | 189 | function smarty_modifier_markdown($text) { 190 | return Markdown($text); 191 | } 192 | 193 | 194 | ### Textile Compatibility Mode ### 195 | 196 | # Rename this file to "classTextile.php" and it can replace Textile everywhere. 197 | 198 | if (strcasecmp(substr(__FILE__, -16), "classTextile.php") == 0) { 199 | # Try to include PHP SmartyPants. Should be in the same directory. 200 | @include_once 'smartypants.php'; 201 | # Fake Textile class. It calls Markdown instead. 202 | class Textile { 203 | function TextileThis($text, $lite='', $encode='') { 204 | if ($lite == '' && $encode == '') $text = Markdown($text); 205 | if (function_exists('SmartyPants')) $text = SmartyPants($text); 206 | return $text; 207 | } 208 | # Fake restricted version: restrictions are not supported for now. 209 | function TextileRestricted($text, $lite='', $noimage='') { 210 | return $this->TextileThis($text, $lite); 211 | } 212 | # Workaround to ensure compatibility with TextPattern 4.0.3. 213 | function blockLite($text) { return $text; } 214 | } 215 | } 216 | 217 | 218 | 219 | # 220 | # Markdown Parser Class 221 | # 222 | 223 | class Markdown_Parser { 224 | 225 | ### Configuration Variables ### 226 | 227 | # Change to ">" for HTML output. 228 | var $empty_element_suffix = MARKDOWN_EMPTY_ELEMENT_SUFFIX; 229 | var $tab_width = MARKDOWN_TAB_WIDTH; 230 | 231 | # Change to `true` to disallow markup or entities. 232 | var $no_markup = false; 233 | var $no_entities = false; 234 | 235 | # Predefined urls and titles for reference links and images. 236 | var $predef_urls = array(); 237 | var $predef_titles = array(); 238 | 239 | 240 | ### Parser Implementation ### 241 | 242 | # Regex to match balanced [brackets]. 243 | # Needed to insert a maximum bracked depth while converting to PHP. 244 | var $nested_brackets_depth = 6; 245 | var $nested_brackets_re; 246 | 247 | var $nested_url_parenthesis_depth = 4; 248 | var $nested_url_parenthesis_re; 249 | 250 | # Table of hash values for escaped characters: 251 | var $escape_chars = '\`*_{}[]()>#+-.!'; 252 | var $escape_chars_re; 253 | 254 | 255 | function Markdown_Parser() { 256 | # 257 | # Constructor function. Initialize appropriate member variables. 258 | # 259 | $this->_initDetab(); 260 | $this->prepareItalicsAndBold(); 261 | 262 | $this->nested_brackets_re = 263 | str_repeat('(?>[^\[\]]+|\[', $this->nested_brackets_depth). 264 | str_repeat('\])*', $this->nested_brackets_depth); 265 | 266 | $this->nested_url_parenthesis_re = 267 | str_repeat('(?>[^()\s]+|\(', $this->nested_url_parenthesis_depth). 268 | str_repeat('(?>\)))*', $this->nested_url_parenthesis_depth); 269 | 270 | $this->escape_chars_re = '['.preg_quote($this->escape_chars).']'; 271 | 272 | # Sort document, block, and span gamut in ascendent priority order. 273 | asort($this->document_gamut); 274 | asort($this->block_gamut); 275 | asort($this->span_gamut); 276 | } 277 | 278 | 279 | # Internal hashes used during transformation. 280 | var $urls = array(); 281 | var $titles = array(); 282 | var $html_hashes = array(); 283 | 284 | # Status flag to avoid invalid nesting. 285 | var $in_anchor = false; 286 | 287 | 288 | function setup() { 289 | # 290 | # Called before the transformation process starts to setup parser 291 | # states. 292 | # 293 | # Clear global hashes. 294 | $this->urls = $this->predef_urls; 295 | $this->titles = $this->predef_titles; 296 | $this->html_hashes = array(); 297 | 298 | $in_anchor = false; 299 | } 300 | 301 | function teardown() { 302 | # 303 | # Called after the transformation process to clear any variable 304 | # which may be taking up memory unnecessarly. 305 | # 306 | $this->urls = array(); 307 | $this->titles = array(); 308 | $this->html_hashes = array(); 309 | } 310 | 311 | 312 | function transform($text) { 313 | # 314 | # Main function. Performs some preprocessing on the input text 315 | # and pass it through the document gamut. 316 | # 317 | $this->setup(); 318 | 319 | # Remove UTF-8 BOM and marker character in input, if present. 320 | $text = preg_replace('{^\xEF\xBB\xBF|\x1A}', '', $text); 321 | 322 | # Standardize line endings: 323 | # DOS to Unix and Mac to Unix 324 | $text = preg_replace('{\r\n?}', "\n", $text); 325 | 326 | # Make sure $text ends with a couple of newlines: 327 | $text .= "\n\n"; 328 | 329 | # Convert all tabs to spaces. 330 | $text = $this->detab($text); 331 | 332 | # Turn block-level HTML blocks into hash entries 333 | $text = $this->hashHTMLBlocks($text); 334 | 335 | # Strip any lines consisting only of spaces and tabs. 336 | # This makes subsequent regexen easier to write, because we can 337 | # match consecutive blank lines with /\n+/ instead of something 338 | # contorted like /[ ]*\n+/ . 339 | $text = preg_replace('/^[ ]+$/m', '', $text); 340 | 341 | # Run document gamut methods. 342 | foreach ($this->document_gamut as $method => $priority) { 343 | $text = $this->$method($text); 344 | } 345 | 346 | $this->teardown(); 347 | 348 | return $text . "\n"; 349 | } 350 | 351 | var $document_gamut = array( 352 | # Strip link definitions, store in hashes. 353 | "stripLinkDefinitions" => 20, 354 | 355 | "runBasicBlockGamut" => 30, 356 | ); 357 | 358 | 359 | function stripLinkDefinitions($text) { 360 | # 361 | # Strips link definitions from text, stores the URLs and titles in 362 | # hash references. 363 | # 364 | $less_than_tab = $this->tab_width - 1; 365 | 366 | # Link defs are in the form: ^[id]: url "optional title" 367 | $text = preg_replace_callback('{ 368 | ^[ ]{0,'.$less_than_tab.'}\[(.+)\][ ]?: # id = $1 369 | [ ]* 370 | \n? # maybe *one* newline 371 | [ ]* 372 | (?: 373 | <(.+?)> # url = $2 374 | | 375 | (\S+?) # url = $3 376 | ) 377 | [ ]* 378 | \n? # maybe one newline 379 | [ ]* 380 | (?: 381 | (?<=\s) # lookbehind for whitespace 382 | ["(] 383 | (.*?) # title = $4 384 | [")] 385 | [ ]* 386 | )? # title is optional 387 | (?:\n+|\Z) 388 | }xm', 389 | array(&$this, '_stripLinkDefinitions_callback'), 390 | $text); 391 | return $text; 392 | } 393 | function _stripLinkDefinitions_callback($matches) { 394 | $link_id = strtolower($matches[1]); 395 | $url = $matches[2] == '' ? $matches[3] : $matches[2]; 396 | $this->urls[$link_id] = $url; 397 | $this->titles[$link_id] =& $matches[4]; 398 | return ''; # String that will replace the block 399 | } 400 | 401 | 402 | function hashHTMLBlocks($text) { 403 | if ($this->no_markup) return $text; 404 | 405 | $less_than_tab = $this->tab_width - 1; 406 | 407 | # Hashify HTML blocks: 408 | # We only want to do this for block-level HTML tags, such as headers, 409 | # lists, and tables. That's because we still want to wrap

    s around 410 | # "paragraphs" that are wrapped in non-block-level tags, such as anchors, 411 | # phrase emphasis, and spans. The list of tags we're looking for is 412 | # hard-coded: 413 | # 414 | # * List "a" is made of tags which can be both inline or block-level. 415 | # These will be treated block-level when the start tag is alone on 416 | # its line, otherwise they're not matched here and will be taken as 417 | # inline later. 418 | # * List "b" is made of tags which are always block-level; 419 | # 420 | $block_tags_a_re = 'ins|del'; 421 | $block_tags_b_re = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|address|'. 422 | 'script|noscript|form|fieldset|iframe|math|svg|'. 423 | 'article|section|nav|aside|hgroup|header|footer|'. 424 | 'figure'; 425 | 426 | # Regular expression for the content of a block tag. 427 | $nested_tags_level = 4; 428 | $attr = ' 429 | (?> # optional tag attributes 430 | \s # starts with whitespace 431 | (?> 432 | [^>"/]+ # text outside quotes 433 | | 434 | /+(?!>) # slash not followed by ">" 435 | | 436 | "[^"]*" # text inside double quotes (tolerate ">") 437 | | 438 | \'[^\']*\' # text inside single quotes (tolerate ">") 439 | )* 440 | )? 441 | '; 442 | $content = 443 | str_repeat(' 444 | (?> 445 | [^<]+ # content without tag 446 | | 447 | <\2 # nested opening tag 448 | '.$attr.' # attributes 449 | (?> 450 | /> 451 | | 452 | >', $nested_tags_level). # end of opening tag 453 | '.*?'. # last level nested tag content 454 | str_repeat(' 455 | # closing nested tag 456 | ) 457 | | 458 | <(?!/\2\s*> # other tags with a different name 459 | ) 460 | )*', 461 | $nested_tags_level); 462 | $content2 = str_replace('\2', '\3', $content); 463 | 464 | # First, look for nested blocks, e.g.: 465 | #

    466 | #
    467 | # tags for inner block must be indented. 468 | #
    469 | #
    470 | # 471 | # The outermost tags must start at the left margin for this to match, and 472 | # the inner nested divs must be indented. 473 | # We need to do this before the next, more liberal match, because the next 474 | # match will start at the first `
    ` and stop at the first `
    `. 475 | $text = preg_replace_callback('{(?> 476 | (?> 477 | (?<=\n\n) # Starting after a blank line 478 | | # or 479 | \A\n? # the beginning of the doc 480 | ) 481 | ( # save in $1 482 | 483 | # Match from `\n` to `\n`, handling nested tags 484 | # in between. 485 | 486 | [ ]{0,'.$less_than_tab.'} 487 | <('.$block_tags_b_re.')# start tag = $2 488 | '.$attr.'> # attributes followed by > and \n 489 | '.$content.' # content, support nesting 490 | # the matching end tag 491 | [ ]* # trailing spaces/tabs 492 | (?=\n+|\Z) # followed by a newline or end of document 493 | 494 | | # Special version for tags of group a. 495 | 496 | [ ]{0,'.$less_than_tab.'} 497 | <('.$block_tags_a_re.')# start tag = $3 498 | '.$attr.'>[ ]*\n # attributes followed by > 499 | '.$content2.' # content, support nesting 500 | # the matching end tag 501 | [ ]* # trailing spaces/tabs 502 | (?=\n+|\Z) # followed by a newline or end of document 503 | 504 | | # Special case just for
    . It was easier to make a special 505 | # case than to make the other regex more complicated. 506 | 507 | [ ]{0,'.$less_than_tab.'} 508 | <(hr) # start tag = $2 509 | '.$attr.' # attributes 510 | /?> # the matching end tag 511 | [ ]* 512 | (?=\n{2,}|\Z) # followed by a blank line or end of document 513 | 514 | | # Special case for standalone HTML comments: 515 | 516 | [ ]{0,'.$less_than_tab.'} 517 | (?s: 518 | 519 | ) 520 | [ ]* 521 | (?=\n{2,}|\Z) # followed by a blank line or end of document 522 | 523 | | # PHP and ASP-style processor instructions ( 530 | ) 531 | [ ]* 532 | (?=\n{2,}|\Z) # followed by a blank line or end of document 533 | 534 | ) 535 | )}Sxmi', 536 | array(&$this, '_hashHTMLBlocks_callback'), 537 | $text); 538 | 539 | return $text; 540 | } 541 | function _hashHTMLBlocks_callback($matches) { 542 | $text = $matches[1]; 543 | $key = $this->hashBlock($text); 544 | return "\n\n$key\n\n"; 545 | } 546 | 547 | 548 | function hashPart($text, $boundary = 'X') { 549 | # 550 | # Called whenever a tag must be hashed when a function insert an atomic 551 | # element in the text stream. Passing $text to through this function gives 552 | # a unique text-token which will be reverted back when calling unhash. 553 | # 554 | # The $boundary argument specify what character should be used to surround 555 | # the token. By convension, "B" is used for block elements that needs not 556 | # to be wrapped into paragraph tags at the end, ":" is used for elements 557 | # that are word separators and "X" is used in the general case. 558 | # 559 | # Swap back any tag hash found in $text so we do not have to `unhash` 560 | # multiple times at the end. 561 | $text = $this->unhash($text); 562 | 563 | # Then hash the block. 564 | static $i = 0; 565 | $key = "$boundary\x1A" . ++$i . $boundary; 566 | $this->html_hashes[$key] = $text; 567 | return $key; # String that will replace the tag. 568 | } 569 | 570 | 571 | function hashBlock($text) { 572 | # 573 | # Shortcut function for hashPart with block-level boundaries. 574 | # 575 | return $this->hashPart($text, 'B'); 576 | } 577 | 578 | 579 | var $block_gamut = array( 580 | # 581 | # These are all the transformations that form block-level 582 | # tags like paragraphs, headers, and list items. 583 | # 584 | "doHeaders" => 10, 585 | "doHorizontalRules" => 20, 586 | 587 | "doLists" => 40, 588 | "doCodeBlocks" => 50, 589 | "doBlockQuotes" => 60, 590 | ); 591 | 592 | function runBlockGamut($text) { 593 | # 594 | # Run block gamut tranformations. 595 | # 596 | # We need to escape raw HTML in Markdown source before doing anything 597 | # else. This need to be done for each block, and not only at the 598 | # begining in the Markdown function since hashed blocks can be part of 599 | # list items and could have been indented. Indented blocks would have 600 | # been seen as a code block in a previous pass of hashHTMLBlocks. 601 | $text = $this->hashHTMLBlocks($text); 602 | 603 | return $this->runBasicBlockGamut($text); 604 | } 605 | 606 | function runBasicBlockGamut($text) { 607 | # 608 | # Run block gamut tranformations, without hashing HTML blocks. This is 609 | # useful when HTML blocks are known to be already hashed, like in the first 610 | # whole-document pass. 611 | # 612 | foreach ($this->block_gamut as $method => $priority) { 613 | $text = $this->$method($text); 614 | } 615 | 616 | # Finally form paragraph and restore hashed blocks. 617 | $text = $this->formParagraphs($text); 618 | 619 | return $text; 620 | } 621 | 622 | 623 | function doHorizontalRules($text) { 624 | # Do Horizontal Rules: 625 | return preg_replace( 626 | '{ 627 | ^[ ]{0,3} # Leading space 628 | ([-*_]) # $1: First marker 629 | (?> # Repeated marker group 630 | [ ]{0,2} # Zero, one, or two spaces. 631 | \1 # Marker character 632 | ){2,} # Group repeated at least twice 633 | [ ]* # Tailing spaces 634 | $ # End of line. 635 | }mx', 636 | "\n".$this->hashBlock("empty_element_suffix")."\n", 637 | $text); 638 | } 639 | 640 | 641 | var $span_gamut = array( 642 | # 643 | # These are all the transformations that occur *within* block-level 644 | # tags like paragraphs, headers, and list items. 645 | # 646 | # Process character escapes, code spans, and inline HTML 647 | # in one shot. 648 | "parseSpan" => -30, 649 | 650 | # Process anchor and image tags. Images must come first, 651 | # because ![foo][f] looks like an anchor. 652 | "doImages" => 10, 653 | "doAnchors" => 20, 654 | 655 | # Make links out of things like `` 656 | # Must come after doAnchors, because you can use < and > 657 | # delimiters in inline links like [this](). 658 | "doAutoLinks" => 30, 659 | "encodeAmpsAndAngles" => 40, 660 | 661 | "doItalicsAndBold" => 50, 662 | "doHardBreaks" => 60, 663 | ); 664 | 665 | function runSpanGamut($text) { 666 | # 667 | # Run span gamut tranformations. 668 | # 669 | foreach ($this->span_gamut as $method => $priority) { 670 | $text = $this->$method($text); 671 | } 672 | 673 | return $text; 674 | } 675 | 676 | 677 | function doHardBreaks($text) { 678 | # Do hard breaks: 679 | return preg_replace_callback('/ {2,}\n/', 680 | array(&$this, '_doHardBreaks_callback'), $text); 681 | } 682 | function _doHardBreaks_callback($matches) { 683 | return $this->hashPart("empty_element_suffix\n"); 684 | } 685 | 686 | 687 | function doAnchors($text) { 688 | # 689 | # Turn Markdown link shortcuts into XHTML tags. 690 | # 691 | if ($this->in_anchor) return $text; 692 | $this->in_anchor = true; 693 | 694 | # 695 | # First, handle reference-style links: [link text] [id] 696 | # 697 | $text = preg_replace_callback('{ 698 | ( # wrap whole match in $1 699 | \[ 700 | ('.$this->nested_brackets_re.') # link text = $2 701 | \] 702 | 703 | [ ]? # one optional space 704 | (?:\n[ ]*)? # one optional newline followed by spaces 705 | 706 | \[ 707 | (.*?) # id = $3 708 | \] 709 | ) 710 | }xs', 711 | array(&$this, '_doAnchors_reference_callback'), $text); 712 | 713 | # 714 | # Next, inline-style links: [link text](url "optional title") 715 | # 716 | $text = preg_replace_callback('{ 717 | ( # wrap whole match in $1 718 | \[ 719 | ('.$this->nested_brackets_re.') # link text = $2 720 | \] 721 | \( # literal paren 722 | [ \n]* 723 | (?: 724 | <(.+?)> # href = $3 725 | | 726 | ('.$this->nested_url_parenthesis_re.') # href = $4 727 | ) 728 | [ \n]* 729 | ( # $5 730 | ([\'"]) # quote char = $6 731 | (.*?) # Title = $7 732 | \6 # matching quote 733 | [ \n]* # ignore any spaces/tabs between closing quote and ) 734 | )? # title is optional 735 | \) 736 | ) 737 | }xs', 738 | array(&$this, '_doAnchors_inline_callback'), $text); 739 | 740 | # 741 | # Last, handle reference-style shortcuts: [link text] 742 | # These must come last in case you've also got [link text][1] 743 | # or [link text](/foo) 744 | # 745 | $text = preg_replace_callback('{ 746 | ( # wrap whole match in $1 747 | \[ 748 | ([^\[\]]+) # link text = $2; can\'t contain [ or ] 749 | \] 750 | ) 751 | }xs', 752 | array(&$this, '_doAnchors_reference_callback'), $text); 753 | 754 | $this->in_anchor = false; 755 | return $text; 756 | } 757 | function _doAnchors_reference_callback($matches) { 758 | $whole_match = $matches[1]; 759 | $link_text = $matches[2]; 760 | $link_id =& $matches[3]; 761 | 762 | if ($link_id == "") { 763 | # for shortcut links like [this][] or [this]. 764 | $link_id = $link_text; 765 | } 766 | 767 | # lower-case and turn embedded newlines into spaces 768 | $link_id = strtolower($link_id); 769 | $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); 770 | 771 | if (isset($this->urls[$link_id])) { 772 | $url = $this->urls[$link_id]; 773 | $url = $this->encodeAttribute($url); 774 | 775 | $result = "titles[$link_id] ) ) { 777 | $title = $this->titles[$link_id]; 778 | $title = $this->encodeAttribute($title); 779 | $result .= " title=\"$title\""; 780 | } 781 | 782 | $link_text = $this->runSpanGamut($link_text); 783 | $result .= ">$link_text"; 784 | $result = $this->hashPart($result); 785 | } 786 | else { 787 | $result = $whole_match; 788 | } 789 | return $result; 790 | } 791 | function _doAnchors_inline_callback($matches) { 792 | $whole_match = $matches[1]; 793 | $link_text = $this->runSpanGamut($matches[2]); 794 | $url = $matches[3] == '' ? $matches[4] : $matches[3]; 795 | $title =& $matches[7]; 796 | 797 | $url = $this->encodeAttribute($url); 798 | 799 | $result = "encodeAttribute($title); 802 | $result .= " title=\"$title\""; 803 | } 804 | 805 | $link_text = $this->runSpanGamut($link_text); 806 | $result .= ">$link_text"; 807 | 808 | return $this->hashPart($result); 809 | } 810 | 811 | 812 | function doImages($text) { 813 | # 814 | # Turn Markdown image shortcuts into tags. 815 | # 816 | # 817 | # First, handle reference-style labeled images: ![alt text][id] 818 | # 819 | $text = preg_replace_callback('{ 820 | ( # wrap whole match in $1 821 | !\[ 822 | ('.$this->nested_brackets_re.') # alt text = $2 823 | \] 824 | 825 | [ ]? # one optional space 826 | (?:\n[ ]*)? # one optional newline followed by spaces 827 | 828 | \[ 829 | (.*?) # id = $3 830 | \] 831 | 832 | ) 833 | }xs', 834 | array(&$this, '_doImages_reference_callback'), $text); 835 | 836 | # 837 | # Next, handle inline images: ![alt text](url "optional title") 838 | # Don't forget: encode * and _ 839 | # 840 | $text = preg_replace_callback('{ 841 | ( # wrap whole match in $1 842 | !\[ 843 | ('.$this->nested_brackets_re.') # alt text = $2 844 | \] 845 | \s? # One optional whitespace character 846 | \( # literal paren 847 | [ \n]* 848 | (?: 849 | <(\S*)> # src url = $3 850 | | 851 | ('.$this->nested_url_parenthesis_re.') # src url = $4 852 | ) 853 | [ \n]* 854 | ( # $5 855 | ([\'"]) # quote char = $6 856 | (.*?) # title = $7 857 | \6 # matching quote 858 | [ \n]* 859 | )? # title is optional 860 | \) 861 | ) 862 | }xs', 863 | array(&$this, '_doImages_inline_callback'), $text); 864 | 865 | return $text; 866 | } 867 | function _doImages_reference_callback($matches) { 868 | $whole_match = $matches[1]; 869 | $alt_text = $matches[2]; 870 | $link_id = strtolower($matches[3]); 871 | 872 | if ($link_id == "") { 873 | $link_id = strtolower($alt_text); # for shortcut links like ![this][]. 874 | } 875 | 876 | $alt_text = $this->encodeAttribute($alt_text); 877 | if (isset($this->urls[$link_id])) { 878 | $url = $this->encodeAttribute($this->urls[$link_id]); 879 | $result = "\"$alt_text\"";titles[$link_id])) { 881 | $title = $this->titles[$link_id]; 882 | $title = $this->encodeAttribute($title); 883 | $result .= " title=\"$title\""; 884 | } 885 | $result .= $this->empty_element_suffix; 886 | $result = $this->hashPart($result); 887 | } 888 | else { 889 | # If there's no such link ID, leave intact: 890 | $result = $whole_match; 891 | } 892 | 893 | return $result; 894 | } 895 | function _doImages_inline_callback($matches) { 896 | $whole_match = $matches[1]; 897 | $alt_text = $matches[2]; 898 | $url = $matches[3] == '' ? $matches[4] : $matches[3]; 899 | $title =& $matches[7]; 900 | 901 | $alt_text = $this->encodeAttribute($alt_text); 902 | $url = $this->encodeAttribute($url); 903 | $result = "\"$alt_text\"";encodeAttribute($title); 906 | $result .= " title=\"$title\""; # $title already quoted 907 | } 908 | $result .= $this->empty_element_suffix; 909 | 910 | return $this->hashPart($result); 911 | } 912 | 913 | 914 | function doHeaders($text) { 915 | # Setext-style headers: 916 | # Header 1 917 | # ======== 918 | # 919 | # Header 2 920 | # -------- 921 | # 922 | $text = preg_replace_callback('{ ^(.+?)[ ]*\n(=+|-+)[ ]*\n+ }mx', 923 | array(&$this, '_doHeaders_callback_setext'), $text); 924 | 925 | # atx-style headers: 926 | # # Header 1 927 | # ## Header 2 928 | # ## Header 2 with closing hashes ## 929 | # ... 930 | # ###### Header 6 931 | # 932 | $text = preg_replace_callback('{ 933 | ^(\#{1,6}) # $1 = string of #\'s 934 | [ ]* 935 | (.+?) # $2 = Header text 936 | [ ]* 937 | \#* # optional closing #\'s (not counted) 938 | \n+ 939 | }xm', 940 | array(&$this, '_doHeaders_callback_atx'), $text); 941 | 942 | return $text; 943 | } 944 | function _doHeaders_callback_setext($matches) { 945 | # Terrible hack to check we haven't found an empty list item. 946 | if ($matches[2] == '-' && preg_match('{^-(?: |$)}', $matches[1])) 947 | return $matches[0]; 948 | 949 | $level = $matches[2]{0} == '=' ? 1 : 2; 950 | $block = "".$this->runSpanGamut($matches[1]).""; 951 | return "\n" . $this->hashBlock($block) . "\n\n"; 952 | } 953 | function _doHeaders_callback_atx($matches) { 954 | $level = strlen($matches[1]); 955 | $block = "".$this->runSpanGamut($matches[2]).""; 956 | return "\n" . $this->hashBlock($block) . "\n\n"; 957 | } 958 | 959 | 960 | function doLists($text) { 961 | # 962 | # Form HTML ordered (numbered) and unordered (bulleted) lists. 963 | # 964 | $less_than_tab = $this->tab_width - 1; 965 | 966 | # Re-usable patterns to match list item bullets and number markers: 967 | $marker_ul_re = '[*+-]'; 968 | $marker_ol_re = '\d+[\.]'; 969 | $marker_any_re = "(?:$marker_ul_re|$marker_ol_re)"; 970 | 971 | $markers_relist = array( 972 | $marker_ul_re => $marker_ol_re, 973 | $marker_ol_re => $marker_ul_re, 974 | ); 975 | 976 | foreach ($markers_relist as $marker_re => $other_marker_re) { 977 | # Re-usable pattern to match any entirel ul or ol list: 978 | $whole_list_re = ' 979 | ( # $1 = whole list 980 | ( # $2 981 | ([ ]{0,'.$less_than_tab.'}) # $3 = number of spaces 982 | ('.$marker_re.') # $4 = first list item marker 983 | [ ]+ 984 | ) 985 | (?s:.+?) 986 | ( # $5 987 | \z 988 | | 989 | \n{2,} 990 | (?=\S) 991 | (?! # Negative lookahead for another list item marker 992 | [ ]* 993 | '.$marker_re.'[ ]+ 994 | ) 995 | | 996 | (?= # Lookahead for another kind of list 997 | \n 998 | \3 # Must have the same indentation 999 | '.$other_marker_re.'[ ]+ 1000 | ) 1001 | ) 1002 | ) 1003 | '; // mx 1004 | 1005 | # We use a different prefix before nested lists than top-level lists. 1006 | # See extended comment in _ProcessListItems(). 1007 | 1008 | if ($this->list_level) { 1009 | $text = preg_replace_callback('{ 1010 | ^ 1011 | '.$whole_list_re.' 1012 | }mx', 1013 | array(&$this, '_doLists_callback'), $text); 1014 | } 1015 | else { 1016 | $text = preg_replace_callback('{ 1017 | (?:(?<=\n)\n|\A\n?) # Must eat the newline 1018 | '.$whole_list_re.' 1019 | }mx', 1020 | array(&$this, '_doLists_callback'), $text); 1021 | } 1022 | } 1023 | 1024 | return $text; 1025 | } 1026 | function _doLists_callback($matches) { 1027 | # Re-usable patterns to match list item bullets and number markers: 1028 | $marker_ul_re = '[*+-]'; 1029 | $marker_ol_re = '\d+[\.]'; 1030 | $marker_any_re = "(?:$marker_ul_re|$marker_ol_re)"; 1031 | 1032 | $list = $matches[1]; 1033 | $list_type = preg_match("/$marker_ul_re/", $matches[4]) ? "ul" : "ol"; 1034 | 1035 | $marker_any_re = ( $list_type == "ul" ? $marker_ul_re : $marker_ol_re ); 1036 | 1037 | $list .= "\n"; 1038 | $result = $this->processListItems($list, $marker_any_re); 1039 | 1040 | $result = $this->hashBlock("<$list_type>\n" . $result . ""); 1041 | return "\n". $result ."\n\n"; 1042 | } 1043 | 1044 | var $list_level = 0; 1045 | 1046 | function processListItems($list_str, $marker_any_re) { 1047 | # 1048 | # Process the contents of a single ordered or unordered list, splitting it 1049 | # into individual list items. 1050 | # 1051 | # The $this->list_level global keeps track of when we're inside a list. 1052 | # Each time we enter a list, we increment it; when we leave a list, 1053 | # we decrement. If it's zero, we're not in a list anymore. 1054 | # 1055 | # We do this because when we're not inside a list, we want to treat 1056 | # something like this: 1057 | # 1058 | # I recommend upgrading to version 1059 | # 8. Oops, now this line is treated 1060 | # as a sub-list. 1061 | # 1062 | # As a single paragraph, despite the fact that the second line starts 1063 | # with a digit-period-space sequence. 1064 | # 1065 | # Whereas when we're inside a list (or sub-list), that line will be 1066 | # treated as the start of a sub-list. What a kludge, huh? This is 1067 | # an aspect of Markdown's syntax that's hard to parse perfectly 1068 | # without resorting to mind-reading. Perhaps the solution is to 1069 | # change the syntax rules such that sub-lists must start with a 1070 | # starting cardinal number; e.g. "1." or "a.". 1071 | 1072 | $this->list_level++; 1073 | 1074 | # trim trailing blank lines: 1075 | $list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str); 1076 | 1077 | $list_str = preg_replace_callback('{ 1078 | (\n)? # leading line = $1 1079 | (^[ ]*) # leading whitespace = $2 1080 | ('.$marker_any_re.' # list marker and space = $3 1081 | (?:[ ]+|(?=\n)) # space only required if item is not empty 1082 | ) 1083 | ((?s:.*?)) # list item text = $4 1084 | (?:(\n+(?=\n))|\n) # tailing blank line = $5 1085 | (?= \n* (\z | \2 ('.$marker_any_re.') (?:[ ]+|(?=\n)))) 1086 | }xm', 1087 | array(&$this, '_processListItems_callback'), $list_str); 1088 | 1089 | $this->list_level--; 1090 | return $list_str; 1091 | } 1092 | function _processListItems_callback($matches) { 1093 | $item = $matches[4]; 1094 | $leading_line =& $matches[1]; 1095 | $leading_space =& $matches[2]; 1096 | $marker_space = $matches[3]; 1097 | $tailing_blank_line =& $matches[5]; 1098 | 1099 | if ($leading_line || $tailing_blank_line || 1100 | preg_match('/\n{2,}/', $item)) 1101 | { 1102 | # Replace marker with the appropriate whitespace indentation 1103 | $item = $leading_space . str_repeat(' ', strlen($marker_space)) . $item; 1104 | $item = $this->runBlockGamut($this->outdent($item)."\n"); 1105 | } 1106 | else { 1107 | # Recursion for sub-lists: 1108 | $item = $this->doLists($this->outdent($item)); 1109 | $item = preg_replace('/\n+$/', '', $item); 1110 | $item = $this->runSpanGamut($item); 1111 | } 1112 | 1113 | return "
  • " . $item . "
    • \n"; 1114 | } 1115 | 1116 | 1117 | function doCodeBlocks($text) { 1118 | # 1119 | # Process Markdown `
    ` blocks.
1120 | 	#
1121 | 		$text = preg_replace_callback('{
1122 | 				(?:\n\n|\A\n?)
1123 | 				(	            # $1 = the code block -- one or more lines, starting with a space/tab
1124 | 				  (?>
1125 | 					[ ]{'.$this->tab_width.'}  # Lines must start with a tab or a tab-width of spaces
1126 | 					.*\n+
1127 | 				  )+
1128 | 				)
1129 | 				((?=^[ ]{0,'.$this->tab_width.'}\S)|\Z)	# Lookahead for non-space at line-start, or end of doc
1130 | 			}xm',
1131 | 			array(&$this, '_doCodeBlocks_callback'), $text);
1132 | 
1133 | 		return $text;
1134 | 	}
1135 | 	function _doCodeBlocks_callback($matches) {
1136 | 		$codeblock = $matches[1];
1137 | 
1138 | 		$codeblock = $this->outdent($codeblock);
1139 | 		$codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES);
1140 | 
1141 | 		# trim leading newlines and trailing newlines
1142 | 		$codeblock = preg_replace('/\A\n+|\n+\z/', '', $codeblock);
1143 | 
1144 | 		$codeblock = "
    $codeblock\n
    ";
1145 | 		return "\n\n".$this->hashBlock($codeblock)."\n\n";
1146 | 	}
1147 | 
1148 | 
1149 | 	function makeCodeSpan($code) {
1150 | 	#
1151 | 	# Create a code span markup for $code. Called from handleSpanToken.
1152 | 	#
1153 | 		$code = htmlspecialchars(trim($code), ENT_NOQUOTES);
1154 | 		return $this->hashPart("$code");
1155 | 	}
1156 | 
1157 | 
1158 | 	var $em_relist = array(
1159 | 		''  => '(?:(? '(?<=\S|^)(? '(?<=\S|^)(? '(?:(? '(?<=\S|^)(? '(?<=\S|^)(? '(?:(? '(?<=\S|^)(? '(?<=\S|^)(?em_relist as $em => $em_re) {
1181 | 			foreach ($this->strong_relist as $strong => $strong_re) {
1182 | 				# Construct list of allowed token expressions.
1183 | 				$token_relist = array();
1184 | 				if (isset($this->em_strong_relist["$em$strong"])) {
1185 | 					$token_relist[] = $this->em_strong_relist["$em$strong"];
1186 | 				}
1187 | 				$token_relist[] = $em_re;
1188 | 				$token_relist[] = $strong_re;
1189 | 				
1190 | 				# Construct master expression from list.
1191 | 				$token_re = '{('. implode('|', $token_relist) .')}';
1192 | 				$this->em_strong_prepared_relist["$em$strong"] = $token_re;
1193 | 			}
1194 | 		}
1195 | 	}
1196 | 	
1197 | 	function doItalicsAndBold($text) {
1198 | 		$token_stack = array('');
1199 | 		$text_stack = array('');
1200 | 		$em = '';
1201 | 		$strong = '';
1202 | 		$tree_char_em = false;
1203 | 		
1204 | 		while (1) {
1205 | 			#
1206 | 			# Get prepared regular expression for seraching emphasis tokens
1207 | 			# in current context.
1208 | 			#
1209 | 			$token_re = $this->em_strong_prepared_relist["$em$strong"];
1210 | 			
1211 | 			#
1212 | 			# Each loop iteration search for the next emphasis token. 
1213 | 			# Each token is then passed to handleSpanToken.
1214 | 			#
1215 | 			$parts = preg_split($token_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE);
1216 | 			$text_stack[0] .= $parts[0];
1217 | 			$token =& $parts[1];
1218 | 			$text =& $parts[2];
1219 | 			
1220 | 			if (empty($token)) {
1221 | 				# Reached end of text span: empty stack without emitting.
1222 | 				# any more emphasis.
1223 | 				while ($token_stack[0]) {
1224 | 					$text_stack[1] .= array_shift($token_stack);
1225 | 					$text_stack[0] .= array_shift($text_stack);
1226 | 				}
1227 | 				break;
1228 | 			}
1229 | 			
1230 | 			$token_len = strlen($token);
1231 | 			if ($tree_char_em) {
1232 | 				# Reached closing marker while inside a three-char emphasis.
1233 | 				if ($token_len == 3) {
1234 | 					# Three-char closing marker, close em and strong.
1235 | 					array_shift($token_stack);
1236 | 					$span = array_shift($text_stack);
1237 | 					$span = $this->runSpanGamut($span);
1238 | 					$span = "$span";
1239 | 					$text_stack[0] .= $this->hashPart($span);
1240 | 					$em = '';
1241 | 					$strong = '';
1242 | 				} else {
1243 | 					# Other closing marker: close one em or strong and
1244 | 					# change current token state to match the other
1245 | 					$token_stack[0] = str_repeat($token{0}, 3-$token_len);
1246 | 					$tag = $token_len == 2 ? "strong" : "em";
1247 | 					$span = $text_stack[0];
1248 | 					$span = $this->runSpanGamut($span);
1249 | 					$span = "<$tag>$span";
1250 | 					$text_stack[0] = $this->hashPart($span);
1251 | 					$$tag = ''; # $$tag stands for $em or $strong
1252 | 				}
1253 | 				$tree_char_em = false;
1254 | 			} else if ($token_len == 3) {
1255 | 				if ($em) {
1256 | 					# Reached closing marker for both em and strong.
1257 | 					# Closing strong marker:
1258 | 					for ($i = 0; $i < 2; ++$i) {
1259 | 						$shifted_token = array_shift($token_stack);
1260 | 						$tag = strlen($shifted_token) == 2 ? "strong" : "em";
1261 | 						$span = array_shift($text_stack);
1262 | 						$span = $this->runSpanGamut($span);
1263 | 						$span = "<$tag>$span";
1264 | 						$text_stack[0] .= $this->hashPart($span);
1265 | 						$$tag = ''; # $$tag stands for $em or $strong
1266 | 					}
1267 | 				} else {
1268 | 					# Reached opening three-char emphasis marker. Push on token 
1269 | 					# stack; will be handled by the special condition above.
1270 | 					$em = $token{0};
1271 | 					$strong = "$em$em";
1272 | 					array_unshift($token_stack, $token);
1273 | 					array_unshift($text_stack, '');
1274 | 					$tree_char_em = true;
1275 | 				}
1276 | 			} else if ($token_len == 2) {
1277 | 				if ($strong) {
1278 | 					# Unwind any dangling emphasis marker:
1279 | 					if (strlen($token_stack[0]) == 1) {
1280 | 						$text_stack[1] .= array_shift($token_stack);
1281 | 						$text_stack[0] .= array_shift($text_stack);
1282 | 					}
1283 | 					# Closing strong marker:
1284 | 					array_shift($token_stack);
1285 | 					$span = array_shift($text_stack);
1286 | 					$span = $this->runSpanGamut($span);
1287 | 					$span = "$span";
1288 | 					$text_stack[0] .= $this->hashPart($span);
1289 | 					$strong = '';
1290 | 				} else {
1291 | 					array_unshift($token_stack, $token);
1292 | 					array_unshift($text_stack, '');
1293 | 					$strong = $token;
1294 | 				}
1295 | 			} else {
1296 | 				# Here $token_len == 1
1297 | 				if ($em) {
1298 | 					if (strlen($token_stack[0]) == 1) {
1299 | 						# Closing emphasis marker:
1300 | 						array_shift($token_stack);
1301 | 						$span = array_shift($text_stack);
1302 | 						$span = $this->runSpanGamut($span);
1303 | 						$span = "$span";
1304 | 						$text_stack[0] .= $this->hashPart($span);
1305 | 						$em = '';
1306 | 					} else {
1307 | 						$text_stack[0] .= $token;
1308 | 					}
1309 | 				} else {
1310 | 					array_unshift($token_stack, $token);
1311 | 					array_unshift($text_stack, '');
1312 | 					$em = $token;
1313 | 				}
1314 | 			}
1315 | 		}
1316 | 		return $text_stack[0];
1317 | 	}
1318 | 
1319 | 
1320 | 	function doBlockQuotes($text) {
1321 | 		$text = preg_replace_callback('/
1322 | 			  (								# Wrap whole match in $1
1323 | 				(?>
1324 | 				  ^[ ]*>[ ]?			# ">" at the start of a line
1325 | 					.+\n					# rest of the first line
1326 | 				  (.+\n)*					# subsequent consecutive lines
1327 | 				  \n*						# blanks
1328 | 				)+
1329 | 			  )
1330 | 			/xm',
1331 | 			array(&$this, '_doBlockQuotes_callback'), $text);
1332 | 
1333 | 		return $text;
1334 | 	}
1335 | 	function _doBlockQuotes_callback($matches) {
1336 | 		$bq = $matches[1];
1337 | 		# trim one level of quoting - trim whitespace-only lines
1338 | 		$bq = preg_replace('/^[ ]*>[ ]?|^[ ]+$/m', '', $bq);
1339 | 		$bq = $this->runBlockGamut($bq);		# recurse
1340 | 
1341 | 		$bq = preg_replace('/^/m', "  ", $bq);
1342 | 		# These leading spaces cause problem with 
     content, 
1343 | 		# so we need to fix that:
1344 | 		$bq = preg_replace_callback('{(\s*
    .+?
    )}sx', 
1345 | 			array(&$this, '_doBlockQuotes_callback2'), $bq);
1346 | 
1347 | 		return "\n". $this->hashBlock("
    \n$bq\n
    ")."\n\n";
1348 | 	}
1349 | 	function _doBlockQuotes_callback2($matches) {
1350 | 		$pre = $matches[1];
1351 | 		$pre = preg_replace('/^  /m', '', $pre);
1352 | 		return $pre;
1353 | 	}
1354 | 
1355 | 
1356 | 	function formParagraphs($text) {
1357 | 	#
1358 | 	#	Params:
1359 | 	#		$text - string to process with html 
     tags
1360 | 	#
1361 | 		# Strip leading and trailing lines:
1362 | 		$text = preg_replace('/\A\n+|\n+\z/', '', $text);
1363 | 
1364 | 		$grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY);
1365 | 
1366 | 		#
1367 | 		# Wrap 
     tags and unhashify HTML blocks
1368 | 		#
1369 | 		foreach ($grafs as $key => $value) {
1370 | 			if (!preg_match('/^B\x1A[0-9]+B$/', $value)) {
1371 | 				# Is a paragraph.
1372 | 				$value = $this->runSpanGamut($value);
1373 | 				$value = preg_replace('/^([ ]*)/', "
    ", $value);
1374 | 				$value .= "
    ";
1375 | 				$grafs[$key] = $this->unhash($value);
1376 | 			}
1377 | 			else {
1378 | 				# Is a block.
1379 | 				# Modify elements of @grafs in-place...
1380 | 				$graf = $value;
1381 | 				$block = $this->html_hashes[$graf];
1382 | 				$graf = $block;
1383 | //				if (preg_match('{
1384 | //					\A
1385 | //					(							# $1 = 
     tag
1386 | //					  
    ]*
1388 | //					  \b
1389 | //					  markdown\s*=\s*  ([\'"])	#	$2 = attr quote char
1390 | //					  1
1391 | //					  \2
1392 | //					  [^>]*
1393 | //					  >
1394 | //					)
1395 | //					(							# $3 = contents
1396 | //					.*
1397 | //					)
1398 | //					(
    )					# $4 = closing tag
1399 | //					\z
1400 | //					}xs', $block, $matches))
1401 | //				{
1402 | //					list(, $div_open, , $div_content, $div_close) = $matches;
1403 | //
1404 | //					# We can't call Markdown(), because that resets the hash;
1405 | //					# that initialization code should be pulled into its own sub, though.
1406 | //					$div_content = $this->hashHTMLBlocks($div_content);
1407 | //					
1408 | //					# Run document gamut methods on the content.
1409 | //					foreach ($this->document_gamut as $method => $priority) {
1410 | //						$div_content = $this->$method($div_content);
1411 | //					}
1412 | //
1413 | //					$div_open = preg_replace(
1414 | //						'{\smarkdown\s*=\s*([\'"]).+?\1}', '', $div_open);
1415 | //
1416 | //					$graf = $div_open . "\n" . $div_content . "\n" . $div_close;
1417 | //				}
1418 | 				$grafs[$key] = $graf;
1419 | 			}
1420 | 		}
1421 | 
1422 | 		return implode("\n\n", $grafs);
1423 | 	}
1424 | 
1425 | 
1426 | 	function encodeAttribute($text) {
1427 | 	#
1428 | 	# Encode text for a double-quoted HTML attribute. This function
1429 | 	# is *not* suitable for attributes enclosed in single quotes.
1430 | 	#
1431 | 		$text = $this->encodeAmpsAndAngles($text);
1432 | 		$text = str_replace('"', '"', $text);
1433 | 		return $text;
1434 | 	}
1435 | 	
1436 | 	
1437 | 	function encodeAmpsAndAngles($text) {
1438 | 	#
1439 | 	# Smart processing for ampersands and angle brackets that need to 
1440 | 	# be encoded. Valid character entities are left alone unless the
1441 | 	# no-entities mode is set.
1442 | 	#
1443 | 		if ($this->no_entities) {
1444 | 			$text = str_replace('&', '&', $text);
1445 | 		} else {
1446 | 			# Ampersand-encoding based entirely on Nat Irons's Amputator
1447 | 			# MT plugin: 
1448 | 			$text = preg_replace('/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/', 
1449 | 								'&', $text);;
1450 | 		}
1451 | 		# Encode remaining <'s
1452 | 		$text = str_replace('<', '<', $text);
1453 | 
1454 | 		return $text;
1455 | 	}
1456 | 
1457 | 
1458 | 	function doAutoLinks($text) {
1459 | 		$text = preg_replace_callback('{<((https?|ftp|dict):[^\'">\s]+)>}i', 
1460 | 			array(&$this, '_doAutoLinks_url_callback'), $text);
1461 | 
1462 | 		# Email addresses: 
1463 | 		$text = preg_replace_callback('{
1464 | 			<
1465 | 			(?:mailto:)?
1466 | 			(
1467 | 				(?:
1468 | 					[-!#$%&\'*+/=?^_`.{|}~\w\x80-\xFF]+
1469 | 				|
1470 | 					".*?"
1471 | 				)
1472 | 				\@
1473 | 				(?:
1474 | 					[-a-z0-9\x80-\xFF]+(\.[-a-z0-9\x80-\xFF]+)*\.[a-z]+
1475 | 				|
1476 | 					\[[\d.a-fA-F:]+\]	# IPv4 & IPv6
1477 | 				)
1478 | 			)
1479 | 			>
1480 | 			}xi',
1481 | 			array(&$this, '_doAutoLinks_email_callback'), $text);
1482 | 
1483 | 		return $text;
1484 | 	}
1485 | 	function _doAutoLinks_url_callback($matches) {
1486 | 		$url = $this->encodeAttribute($matches[1]);
1487 | 		$link = "$url";
1488 | 		return $this->hashPart($link);
1489 | 	}
1490 | 	function _doAutoLinks_email_callback($matches) {
1491 | 		$address = $matches[1];
1492 | 		$link = $this->encodeEmailAddress($address);
1493 | 		return $this->hashPart($link);
1494 | 	}
1495 | 
1496 | 
1497 | 	function encodeEmailAddress($addr) {
1498 | 	#
1499 | 	#	Input: an email address, e.g. "foo@example.com"
1500 | 	#
1501 | 	#	Output: the email address as a mailto link, with each character
1502 | 	#		of the address encoded as either a decimal or hex entity, in
1503 | 	#		the hopes of foiling most address harvesting spam bots. E.g.:
1504 | 	#
1505 | 	#	  
    foo@exampl
1508 | 	#        e.com
    
1509 | 	#
1510 | 	#	Based by a filter by Matthew Wickline, posted to BBEdit-Talk.
1511 | 	#   With some optimizations by Milian Wolff.
1512 | 	#
1513 | 		$addr = "mailto:" . $addr;
1514 | 		$chars = preg_split('/(? $char) {
1518 | 			$ord = ord($char);
1519 | 			# Ignore non-ascii chars.
1520 | 			if ($ord < 128) {
1521 | 				$r = ($seed * (1 + $key)) % 100; # Pseudo-random function.
1522 | 				# roughly 10% raw, 45% hex, 45% dec
1523 | 				# '@' *must* be encoded. I insist.
1524 | 				if ($r > 90 && $char != '@') /* do nothing */;
1525 | 				else if ($r < 45) $chars[$key] = '&#x'.dechex($ord).';';
1526 | 				else              $chars[$key] = '&#'.$ord.';';
1527 | 			}
1528 | 		}
1529 | 		
1530 | 		$addr = implode('', $chars);
1531 | 		$text = implode('', array_slice($chars, 7)); # text without `mailto:`
1532 | 		$addr = "$text";
1533 | 
1534 | 		return $addr;
1535 | 	}
1536 | 
1537 | 
1538 | 	function parseSpan($str) {
1539 | 	#
1540 | 	# Take the string $str and parse it into tokens, hashing embeded HTML,
1541 | 	# escaped characters and handling code spans.
1542 | 	#
1543 | 		$output = '';
1544 | 		
1545 | 		$span_re = '{
1546 | 				(
1547 | 					\\\\'.$this->escape_chars_re.'
1548 | 				|
1549 | 					(?no_markup ? '' : '
1552 | 				|
1553 | 							# comment
1554 | 				|
1555 | 					<\?.*?\?> | <%.*?%>		# processing instruction
1556 | 				|
1557 | 					<[!$]?[-a-zA-Z0-9:_]+	# regular tags
1558 | 					(?>
1559 | 						\s
1560 | 						(?>[^"\'>]+|"[^"]*"|\'[^\']*\')*
1561 | 					)?
1562 | 					>
1563 | 				|
1564 | 					<[-a-zA-Z0-9:_]+\s*/> # xml-style empty tag
1565 | 				|
1566 | 					 # closing tag
1567 | 			').'
1568 | 				)
1569 | 				}xs';
1570 | 
1571 | 		while (1) {
1572 | 			#
1573 | 			# Each loop iteration seach for either the next tag, the next 
1574 | 			# openning code span marker, or the next escaped character. 
1575 | 			# Each token is then passed to handleSpanToken.
1576 | 			#
1577 | 			$parts = preg_split($span_re, $str, 2, PREG_SPLIT_DELIM_CAPTURE);
1578 | 			
1579 | 			# Create token from text preceding tag.
1580 | 			if ($parts[0] != "") {
1581 | 				$output .= $parts[0];
1582 | 			}
1583 | 			
1584 | 			# Check if we reach the end.
1585 | 			if (isset($parts[1])) {
1586 | 				$output .= $this->handleSpanToken($parts[1], $parts[2]);
1587 | 				$str = $parts[2];
1588 | 			}
1589 | 			else {
1590 | 				break;
1591 | 			}
1592 | 		}
1593 | 		
1594 | 		return $output;
1595 | 	}
1596 | 	
1597 | 	
1598 | 	function handleSpanToken($token, &$str) {
1599 | 	#
1600 | 	# Handle $token provided by parseSpan by determining its nature and 
1601 | 	# returning the corresponding value that should replace it.
1602 | 	#
1603 | 		switch ($token{0}) {
1604 | 			case "\\":
1605 | 				return $this->hashPart("&#". ord($token{1}). ";");
1606 | 			case "`":
1607 | 				# Search for end marker in remaining text.
1608 | 				if (preg_match('/^(.*?[^`])'.preg_quote($token).'(?!`)(.*)$/sm', 
1609 | 					$str, $matches))
1610 | 				{
1611 | 					$str = $matches[2];
1612 | 					$codespan = $this->makeCodeSpan($matches[1]);
1613 | 					return $this->hashPart($codespan);
1614 | 				}
1615 | 				return $token; // return as text since no ending marker found.
1616 | 			default:
1617 | 				return $this->hashPart($token);
1618 | 		}
1619 | 	}
1620 | 
1621 | 
1622 | 	function outdent($text) {
1623 | 	#
1624 | 	# Remove one level of line-leading tabs or spaces
1625 | 	#
1626 | 		return preg_replace('/^(\t|[ ]{1,'.$this->tab_width.'})/m', '', $text);
1627 | 	}
1628 | 
1629 | 
1630 | 	# String length function for detab. `_initDetab` will create a function to 
1631 | 	# hanlde UTF-8 if the default function does not exist.
1632 | 	var $utf8_strlen = 'mb_strlen';
1633 | 	
1634 | 	function detab($text) {
1635 | 	#
1636 | 	# Replace tabs with the appropriate amount of space.
1637 | 	#
1638 | 		# For each line we separate the line in blocks delemited by
1639 | 		# tab characters. Then we reconstruct every line by adding the 
1640 | 		# appropriate number of space between each blocks.
1641 | 		
1642 | 		$text = preg_replace_callback('/^.*\t.*$/m',
1643 | 			array(&$this, '_detab_callback'), $text);
1644 | 
1645 | 		return $text;
1646 | 	}
1647 | 	function _detab_callback($matches) {
1648 | 		$line = $matches[0];
1649 | 		$strlen = $this->utf8_strlen; # strlen function for UTF-8.
1650 | 		
1651 | 		# Split in blocks.
1652 | 		$blocks = explode("\t", $line);
1653 | 		# Add each blocks to the line.
1654 | 		$line = $blocks[0];
1655 | 		unset($blocks[0]); # Do not add first block twice.
1656 | 		foreach ($blocks as $block) {
1657 | 			# Calculate amount of space, insert spaces, insert block.
1658 | 			$amount = $this->tab_width - 
1659 | 				$strlen($line, 'UTF-8') % $this->tab_width;
1660 | 			$line .= str_repeat(" ", $amount) . $block;
1661 | 		}
1662 | 		return $line;
1663 | 	}
1664 | 	function _initDetab() {
1665 | 	#
1666 | 	# Check for the availability of the function in the `utf8_strlen` property
1667 | 	# (initially `mb_strlen`). If the function is not available, create a 
1668 | 	# function that will loosely count the number of UTF-8 characters with a
1669 | 	# regular expression.
1670 | 	#
1671 | 		if (function_exists($this->utf8_strlen)) return;
1672 | 		$this->utf8_strlen = create_function('$text', 'return preg_match_all(
1673 | 			"/[\\\\x00-\\\\xBF]|[\\\\xC0-\\\\xFF][\\\\x80-\\\\xBF]*/", 
1674 | 			$text, $m);');
1675 | 	}
1676 | 
1677 | 
1678 | 	function unhash($text) {
1679 | 	#
1680 | 	# Swap back in all the tags hashed by _HashHTMLBlocks.
1681 | 	#
1682 | 		return preg_replace_callback('/(.)\x1A[0-9]+\1/', 
1683 | 			array(&$this, '_unhash_callback'), $text);
1684 | 	}
1685 | 	function _unhash_callback($matches) {
1686 | 		return $this->html_hashes[$matches[0]];
1687 | 	}
1688 | 
1689 | }
1690 | 
1691 | 
1692 | #
1693 | # Markdown Extra Parser Class
1694 | #
1695 | 
1696 | class MarkdownExtra_Parser extends Markdown_Parser {
1697 | 
1698 | 	### Configuration Variables ###
1699 | 
1700 | 	# Prefix for footnote ids.
1701 | 	var $fn_id_prefix = "";
1702 | 	
1703 | 	# Optional title attribute for footnote links and backlinks.
1704 | 	var $fn_link_title = MARKDOWN_FN_LINK_TITLE;
1705 | 	var $fn_backlink_title = MARKDOWN_FN_BACKLINK_TITLE;
1706 | 	
1707 | 	# Optional class attribute for footnote links and backlinks.
1708 | 	var $fn_link_class = MARKDOWN_FN_LINK_CLASS;
1709 | 	var $fn_backlink_class = MARKDOWN_FN_BACKLINK_CLASS;
1710 | 
1711 | 	# Optional class prefix for fenced code block.
1712 | 	var $code_class_prefix = MARKDOWN_CODE_CLASS_PREFIX;
1713 | 	# Class attribute for code blocks goes on the `code` tag;
1714 | 	# setting this to true will put attributes on the `pre` tag instead.
1715 | 	var $code_attr_on_pre = MARKDOWN_CODE_ATTR_ON_PRE;
1716 | 	
1717 | 	# Predefined abbreviations.
1718 | 	var $predef_abbr = array();
1719 | 
1720 | 
1721 | 	### Parser Implementation ###
1722 | 
1723 | 	function MarkdownExtra_Parser() {
1724 | 	#
1725 | 	# Constructor function. Initialize the parser object.
1726 | 	#
1727 | 		# Add extra escapable characters before parent constructor 
1728 | 		# initialize the table.
1729 | 		$this->escape_chars .= ':|';
1730 | 		
1731 | 		# Insert extra document, block, and span transformations. 
1732 | 		# Parent constructor will do the sorting.
1733 | 		$this->document_gamut += array(
1734 | 			"doFencedCodeBlocks" => 5,
1735 | 			"stripFootnotes"     => 15,
1736 | 			"stripAbbreviations" => 25,
1737 | 			"appendFootnotes"    => 50,
1738 | 			);
1739 | 		$this->block_gamut += array(
1740 | 			"doFencedCodeBlocks" => 5,
1741 | 			"doTables"           => 15,
1742 | 			"doDefLists"         => 45,
1743 | 			);
1744 | 		$this->span_gamut += array(
1745 | 			"doFootnotes"        => 5,
1746 | 			"doAbbreviations"    => 70,
1747 | 			);
1748 | 		
1749 | 		parent::Markdown_Parser();
1750 | 	}
1751 | 	
1752 | 	
1753 | 	# Extra variables used during extra transformations.
1754 | 	var $footnotes = array();
1755 | 	var $footnotes_ordered = array();
1756 | 	var $footnotes_ref_count = array();
1757 | 	var $footnotes_numbers = array();
1758 | 	var $abbr_desciptions = array();
1759 | 	var $abbr_word_re = '';
1760 | 	
1761 | 	# Give the current footnote number.
1762 | 	var $footnote_counter = 1;
1763 | 	
1764 | 	
1765 | 	function setup() {
1766 | 	#
1767 | 	# Setting up Extra-specific variables.
1768 | 	#
1769 | 		parent::setup();
1770 | 		
1771 | 		$this->footnotes = array();
1772 | 		$this->footnotes_ordered = array();
1773 | 		$this->footnotes_ref_count = array();
1774 | 		$this->footnotes_numbers = array();
1775 | 		$this->abbr_desciptions = array();
1776 | 		$this->abbr_word_re = '';
1777 | 		$this->footnote_counter = 1;
1778 | 		
1779 | 		foreach ($this->predef_abbr as $abbr_word => $abbr_desc) {
1780 | 			if ($this->abbr_word_re)
1781 | 				$this->abbr_word_re .= '|';
1782 | 			$this->abbr_word_re .= preg_quote($abbr_word);
1783 | 			$this->abbr_desciptions[$abbr_word] = trim($abbr_desc);
1784 | 		}
1785 | 	}
1786 | 	
1787 | 	function teardown() {
1788 | 	#
1789 | 	# Clearing Extra-specific variables.
1790 | 	#
1791 | 		$this->footnotes = array();
1792 | 		$this->footnotes_ordered = array();
1793 | 		$this->footnotes_ref_count = array();
1794 | 		$this->footnotes_numbers = array();
1795 | 		$this->abbr_desciptions = array();
1796 | 		$this->abbr_word_re = '';
1797 | 		
1798 | 		parent::teardown();
1799 | 	}
1800 | 	
1801 | 	
1802 | 	### Extra Attribute Parser ###
1803 | 
1804 | 	# Expression to use to catch attributes (includes the braces)
1805 | 	var $id_class_attr_catch_re = '\{((?:[ ]*[#.][-_:a-zA-Z0-9]+){1,})[ ]*\}';
1806 | 	# Expression to use when parsing in a context when no capture is desired
1807 | 	var $id_class_attr_nocatch_re = '\{(?:[ ]*[#.][-_:a-zA-Z0-9]+){1,}[ ]*\}';
1808 | 
1809 | 	function doExtraAttributes($tag_name, $attr) {
1810 | 	#
1811 | 	# Parse attributes caught by the $this->id_class_attr_catch_re expression
1812 | 	# and return the HTML-formatted list of attributes.
1813 | 	#
1814 | 	# Currently supported attributes are .class and #id.
1815 | 	#
1816 | 		if (empty($attr)) return "";
1817 | 		
1818 | 		# Split on components
1819 | 		preg_match_all("/[.#][-_:a-zA-Z0-9]+/", $attr, $matches);
1820 | 		$elements = $matches[0];
1821 | 
1822 | 		# handle classes and ids (only first id taken into account)
1823 | 		$classes = array();
1824 | 		$id = false;
1825 | 		foreach ($elements as $element) {
1826 | 			if ($element{0} == '.') {
1827 | 				$classes[] = substr($element, 1);
1828 | 			} else if ($element{0} == '#') {
1829 | 				if ($id === false) $id = substr($element, 1);
1830 | 			}
1831 | 		}
1832 | 
1833 | 		# compose attributes as string
1834 | 		$attr_str = "";
1835 | 		if (!empty($id)) {
1836 | 			$attr_str .= ' id="'.$id.'"';
1837 | 		}
1838 | 		if (!empty($classes)) {
1839 | 			$attr_str .= ' class="'.implode(" ", $classes).'"';
1840 | 		}
1841 | 		return $attr_str;
1842 | 	}
1843 | 
1844 | 
1845 | 	### HTML Block Parser ###
1846 | 	
1847 | 	# Tags that are always treated as block tags:
1848 | 	var $block_tags_re = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|address|form|fieldset|iframe|hr|legend|article|section|nav|aside|hgroup|header|footer|figcaption';
1849 | 						   
1850 | 	# Tags treated as block tags only if the opening tag is alone on its line:
1851 | 	var $context_block_tags_re = 'script|noscript|ins|del|iframe|object|source|track|param|math|svg|canvas|audio|video';
1852 | 	
1853 | 	# Tags where markdown="1" default to span mode:
1854 | 	var $contain_span_tags_re = 'p|h[1-6]|li|dd|dt|td|th|legend|address';
1855 | 	
1856 | 	# Tags which must not have their contents modified, no matter where 
1857 | 	# they appear:
1858 | 	var $clean_tags_re = 'script|math|svg';
1859 | 	
1860 | 	# Tags that do not need to be closed.
1861 | 	var $auto_close_tags_re = 'hr|img|param|source|track';
1862 | 	
1863 | 
1864 | 	function hashHTMLBlocks($text) {
1865 | 	#
1866 | 	# Hashify HTML Blocks and "clean tags".
1867 | 	#
1868 | 	# We only want to do this for block-level HTML tags, such as headers,
1869 | 	# lists, and tables. That's because we still want to wrap 
    s around
1870 | 	# "paragraphs" that are wrapped in non-block-level tags, such as anchors,
1871 | 	# phrase emphasis, and spans. The list of tags we're looking for is
1872 | 	# hard-coded.
1873 | 	#
1874 | 	# This works by calling _HashHTMLBlocks_InMarkdown, which then calls
1875 | 	# _HashHTMLBlocks_InHTML when it encounter block tags. When the markdown="1" 
1876 | 	# attribute is found within a tag, _HashHTMLBlocks_InHTML calls back
1877 | 	#  _HashHTMLBlocks_InMarkdown to handle the Markdown syntax within the tag.
1878 | 	# These two functions are calling each other. It's recursive!
1879 | 	#
1880 | 		if ($this->no_markup)  return $text;
1881 | 
1882 | 		#
1883 | 		# Call the HTML-in-Markdown hasher.
1884 | 		#
1885 | 		list($text, ) = $this->_hashHTMLBlocks_inMarkdown($text);
1886 | 		
1887 | 		return $text;
1888 | 	}
1889 | 	function _hashHTMLBlocks_inMarkdown($text, $indent = 0, 
1890 | 										$enclosing_tag_re = '', $span = false)
1891 | 	{
1892 | 	#
1893 | 	# Parse markdown text, calling _HashHTMLBlocks_InHTML for block tags.
1894 | 	#
1895 | 	# *   $indent is the number of space to be ignored when checking for code 
1896 | 	#     blocks. This is important because if we don't take the indent into 
1897 | 	#     account, something like this (which looks right) won't work as expected:
1898 | 	#
1899 | 	#     
    
1900 | 	#         
    
1901 | 	#         Hello World.  <-- Is this a Markdown code block or text?
1902 | 	#         
      <-- Is this a Markdown code block or a real tag?
1903 | 	#     
    
1904 | 	#
1905 | 	#     If you don't like this, just don't indent the tag on which
1906 | 	#     you apply the markdown="1" attribute.
1907 | 	#
1908 | 	# *   If $enclosing_tag_re is not empty, stops at the first unmatched closing 
1909 | 	#     tag with that name. Nested tags supported.
1910 | 	#
1911 | 	# *   If $span is true, text inside must treated as span. So any double 
1912 | 	#     newline will be replaced by a single newline so that it does not create 
1913 | 	#     paragraphs.
1914 | 	#
1915 | 	# Returns an array of that form: ( processed text , remaining text )
1916 | 	#
1917 | 		if ($text === '') return array('', '');
1918 | 
1919 | 		# Regex to check for the presense of newlines around a block tag.
1920 | 		$newline_before_re = '/(?:^\n?|\n\n)*$/';
1921 | 		$newline_after_re = 
1922 | 			'{
1923 | 				^						# Start of text following the tag.
1924 | 				(?>[ ]*)?		# Optional comment.
1925 | 				[ ]*\n					# Must be followed by newline.
1926 | 			}xs';
1927 | 		
1928 | 		# Regex to match any tag.
1929 | 		$block_tag_re =
1930 | 			'{
1931 | 				(					# $2: Capture whole tag.
1932 | 									# Tag name.
1934 | 							'.$this->block_tags_re.'			|
1935 | 							'.$this->context_block_tags_re.'	|
1936 | 							'.$this->clean_tags_re.'        	|
1937 | 							(?!\s)'.$enclosing_tag_re.'
1938 | 						)
1939 | 						(?:
1940 | 							(?=[\s"\'/a-zA-Z0-9])	# Allowed characters after tag name.
1941 | 							(?>
1942 | 								".*?"		|	# Double quotes (can contain `>`)
1943 | 								\'.*?\'   	|	# Single quotes (can contain `>`)
1944 | 								.+?				# Anything but quotes and `>`.
1945 | 							)*?
1946 | 						)?
1947 | 					>					# End of tag.
1948 | 				|
1949 | 						# HTML Comment
1950 | 				|
1951 | 					<\?.*?\?> | <%.*?%>	# Processing instruction
1952 | 				|
1953 | 						# CData Block
1954 | 				|
1955 | 					# Code span marker
1956 | 					`+
1957 | 				'. ( !$span ? ' # If not in span.
1958 | 				|
1959 | 					# Indented code block
1960 | 					(?: ^[ ]*\n | ^ | \n[ ]*\n )
1961 | 					[ ]{'.($indent+4).'}[^\n]* \n
1962 | 					(?>
1963 | 						(?: [ ]{'.($indent+4).'}[^\n]* | [ ]* ) \n
1964 | 					)*
1965 | 				|
1966 | 					# Fenced code block marker
1967 | 					(?<= ^ | \n )
1968 | 					[ ]{0,'.($indent+3).'}~{3,}
1969 | 									[ ]*
1970 | 					(?:
1971 | 						[.]?[-_:a-zA-Z0-9]+ # standalone class name
1972 | 					|
1973 | 						'.$this->id_class_attr_nocatch_re.' # extra attributes
1974 | 					)?
1975 | 					[ ]*
1976 | 					\n
1977 | 				' : '' ). ' # End (if not is span).
1978 | 				)
1979 | 			}xs';
1980 | 
1981 | 		
1982 | 		$depth = 0;		# Current depth inside the tag tree.
1983 | 		$parsed = "";	# Parsed text that will be returned.
1984 | 
1985 | 		#
1986 | 		# Loop through every tag until we find the closing tag of the parent
1987 | 		# or loop until reaching the end of text if no parent tag specified.
1988 | 		#
1989 | 		do {
1990 | 			#
1991 | 			# Split the text using the first $tag_match pattern found.
1992 | 			# Text before  pattern will be first in the array, text after
1993 | 			# pattern will be at the end, and between will be any catches made 
1994 | 			# by the pattern.
1995 | 			#
1996 | 			$parts = preg_split($block_tag_re, $text, 2, 
1997 | 								PREG_SPLIT_DELIM_CAPTURE);
1998 | 			
1999 | 			# If in Markdown span mode, add a empty-string span-level hash 
2000 | 			# after each newline to prevent triggering any block element.
2001 | 			if ($span) {
2002 | 				$void = $this->hashPart("", ':');
2003 | 				$newline = "$void\n";
2004 | 				$parts[0] = $void . str_replace("\n", $newline, $parts[0]) . $void;
2005 | 			}
2006 | 			
2007 | 			$parsed .= $parts[0]; # Text before current tag.
2008 | 			
2009 | 			# If end of $text has been reached. Stop loop.
2010 | 			if (count($parts) < 3) {
2011 | 				$text = "";
2012 | 				break;
2013 | 			}
2014 | 			
2015 | 			$tag  = $parts[1]; # Tag to handle.
2016 | 			$text = $parts[2]; # Remaining text after current tag.
2017 | 			$tag_re = preg_quote($tag); # For use in a regular expression.
2018 | 			
2019 | 			#
2020 | 			# Check for: Code span marker
2021 | 			#
2022 | 			if ($tag{0} == "`") {
2023 | 				# Find corresponding end marker.
2024 | 				$tag_re = preg_quote($tag);
2025 | 				if (preg_match('{^(?>.+?|\n(?!\n))*?(?.*\n)*?[ ]{'.($fence_indent).'}'.$fence_re.'[ ]*(?:\n|$)}', $text,
2045 | 					$matches)) 
2046 | 				{
2047 | 					# End marker found: pass text unchanged until marker.
2048 | 					$parsed .= $tag . $matches[0];
2049 | 					$text = substr($text, strlen($matches[0]));
2050 | 				}
2051 | 				else {
2052 | 					# No end marker: just skip it.
2053 | 					$parsed .= $tag;
2054 | 				}
2055 | 			}
2056 | 			#
2057 | 			# Check for: Indented code block.
2058 | 			#
2059 | 			else if ($tag{0} == "\n" || $tag{0} == " ") {
2060 | 				# Indented code block: pass it unchanged, will be handled 
2061 | 				# later.
2062 | 				$parsed .= $tag;
2063 | 			}
2064 | 			#
2065 | 			# Check for: Opening Block level tag or
2066 | 			#            Opening Context Block tag (like ins and del) 
2067 | 			#               used as a block tag (tag is alone on it's line).
2068 | 			#
2069 | 			else if (preg_match('{^<(?:'.$this->block_tags_re.')\b}', $tag) ||
2070 | 				(	preg_match('{^<(?:'.$this->context_block_tags_re.')\b}', $tag) &&
2071 | 					preg_match($newline_before_re, $parsed) &&
2072 | 					preg_match($newline_after_re, $text)	)
2073 | 				)
2074 | 			{
2075 | 				# Need to parse tag and following text using the HTML parser.
2076 | 				list($block_text, $text) = 
2077 | 					$this->_hashHTMLBlocks_inHTML($tag . $text, "hashBlock", true);
2078 | 				
2079 | 				# Make sure it stays outside of any paragraph by adding newlines.
2080 | 				$parsed .= "\n\n$block_text\n\n";
2081 | 			}
2082 | 			#
2083 | 			# Check for: Clean tag (like script, math)
2084 | 			#            HTML Comments, processing instructions.
2085 | 			#
2086 | 			else if (preg_match('{^<(?:'.$this->clean_tags_re.')\b}', $tag) ||
2087 | 				$tag{1} == '!' || $tag{1} == '?')
2088 | 			{
2089 | 				# Need to parse tag and following text using the HTML parser.
2090 | 				# (don't check for markdown attribute)
2091 | 				list($block_text, $text) = 
2092 | 					$this->_hashHTMLBlocks_inHTML($tag . $text, "hashClean", false);
2093 | 				
2094 | 				$parsed .= $block_text;
2095 | 			}
2096 | 			#
2097 | 			# Check for: Tag with same name as enclosing tag.
2098 | 			#
2099 | 			else if ($enclosing_tag_re !== '' &&
2100 | 				# Same name as enclosing tag.
2101 | 				preg_match('{^= 0);
2124 | 		
2125 | 		return array($parsed, $text);
2126 | 	}
2127 | 	function _hashHTMLBlocks_inHTML($text, $hash_method, $md_attr) {
2128 | 	#
2129 | 	# Parse HTML, calling _HashHTMLBlocks_InMarkdown for block tags.
2130 | 	#
2131 | 	# *   Calls $hash_method to convert any blocks.
2132 | 	# *   Stops when the first opening tag closes.
2133 | 	# *   $md_attr indicate if the use of the `markdown="1"` attribute is allowed.
2134 | 	#     (it is not inside clean tags)
2135 | 	#
2136 | 	# Returns an array of that form: ( processed text , remaining text )
2137 | 	#
2138 | 		if ($text === '') return array('', '');
2139 | 		
2140 | 		# Regex to match `markdown` attribute inside of a tag.
2141 | 		$markdown_attr_re = '
2142 | 			{
2143 | 				\s*			# Eat whitespace before the `markdown` attribute
2144 | 				markdown
2145 | 				\s*=\s*
2146 | 				(?>
2147 | 					(["\'])		# $1: quote delimiter		
2148 | 					(.*?)		# $2: attribute value
2149 | 					\1			# matching delimiter	
2150 | 				|
2151 | 					([^\s>]*)	# $3: unquoted attribute value
2152 | 				)
2153 | 				()				# $4: make $3 always defined (avoid warnings)
2154 | 			}xs';
2155 | 		
2156 | 		# Regex to match any tag.
2157 | 		$tag_re = '{
2158 | 				(					# $2: Capture whole tag.
2159 | 					
2164 | 								".*?"		|	# Double quotes (can contain `>`)
2165 | 								\'.*?\'   	|	# Single quotes (can contain `>`)
2166 | 								.+?				# Anything but quotes and `>`.
2167 | 							)*?
2168 | 						)?
2169 | 					>					# End of tag.
2170 | 				|
2171 | 						# HTML Comment
2172 | 				|
2173 | 					<\?.*?\?> | <%.*?%>	# Processing instruction
2174 | 				|
2175 | 						# CData Block
2176 | 				)
2177 | 			}xs';
2178 | 		
2179 | 		$original_text = $text;		# Save original text in case of faliure.
2180 | 		
2181 | 		$depth		= 0;	# Current depth inside the tag tree.
2182 | 		$block_text	= "";	# Temporary text holder for current text.
2183 | 		$parsed		= "";	# Parsed text that will be returned.
2184 | 
2185 | 		#
2186 | 		# Get the name of the starting tag.
2187 | 		# (This pattern makes $base_tag_name_re safe without quoting.)
2188 | 		#
2189 | 		if (preg_match('/^<([\w:$]*)\b/', $text, $matches))
2190 | 			$base_tag_name_re = $matches[1];
2191 | 
2192 | 		#
2193 | 		# Loop through every tag until we find the corresponding closing tag.
2194 | 		#
2195 | 		do {
2196 | 			#
2197 | 			# Split the text using the first $tag_match pattern found.
2198 | 			# Text before  pattern will be first in the array, text after
2199 | 			# pattern will be at the end, and between will be any catches made 
2200 | 			# by the pattern.
2201 | 			#
2202 | 			$parts = preg_split($tag_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE);
2203 | 			
2204 | 			if (count($parts) < 3) {
2205 | 				#
2206 | 				# End of $text reached with unbalenced tag(s).
2207 | 				# In that case, we return original text unchanged and pass the
2208 | 				# first character as filtered to prevent an infinite loop in the 
2209 | 				# parent function.
2210 | 				#
2211 | 				return array($original_text{0}, substr($original_text, 1));
2212 | 			}
2213 | 			
2214 | 			$block_text .= $parts[0]; # Text before current tag.
2215 | 			$tag         = $parts[1]; # Tag to handle.
2216 | 			$text        = $parts[2]; # Remaining text after current tag.
2217 | 			
2218 | 			#
2219 | 			# Check for: Auto-close tag (like 
    )
2220 | 			#			 Comments and Processing Instructions.
2221 | 			#
2222 | 			if (preg_match('{^auto_close_tags_re.')\b}', $tag) ||
2223 | 				$tag{1} == '!' || $tag{1} == '?')
2224 | 			{
2225 | 				# Just add the tag to the block as if it was text.
2226 | 				$block_text .= $tag;
2227 | 			}
2228 | 			else {
2229 | 				#
2230 | 				# Increase/decrease nested tag count. Only do so if
2231 | 				# the tag's name match base tag's.
2232 | 				#
2233 | 				if (preg_match('{^mode = $attr_m[2] . $attr_m[3];
2250 | 					$span_mode = $this->mode == 'span' || $this->mode != 'block' &&
2251 | 						preg_match('{^<(?:'.$this->contain_span_tags_re.')\b}', $tag);
2252 | 					
2253 | 					# Calculate indent before tag.
2254 | 					if (preg_match('/(?:^|\n)( *?)(?! ).*?$/', $block_text, $matches)) {
2255 | 						$strlen = $this->utf8_strlen;
2256 | 						$indent = $strlen($matches[1], 'UTF-8');
2257 | 					} else {
2258 | 						$indent = 0;
2259 | 					}
2260 | 					
2261 | 					# End preceding block with this tag.
2262 | 					$block_text .= $tag;
2263 | 					$parsed .= $this->$hash_method($block_text);
2264 | 					
2265 | 					# Get enclosing tag name for the ParseMarkdown function.
2266 | 					# (This pattern makes $tag_name_re safe without quoting.)
2267 | 					preg_match('/^<([\w:$]*)\b/', $tag, $matches);
2268 | 					$tag_name_re = $matches[1];
2269 | 					
2270 | 					# Parse the content using the HTML-in-Markdown parser.
2271 | 					list ($block_text, $text)
2272 | 						= $this->_hashHTMLBlocks_inMarkdown($text, $indent, 
2273 | 							$tag_name_re, $span_mode);
2274 | 					
2275 | 					# Outdent markdown text.
2276 | 					if ($indent > 0) {
2277 | 						$block_text = preg_replace("/^[ ]{1,$indent}/m", "", 
2278 | 													$block_text);
2279 | 					}
2280 | 					
2281 | 					# Append tag content to parsed text.
2282 | 					if (!$span_mode)	$parsed .= "\n\n$block_text\n\n";
2283 | 					else				$parsed .= "$block_text";
2284 | 					
2285 | 					# Start over with a new block.
2286 | 					$block_text = "";
2287 | 				}
2288 | 				else $block_text .= $tag;
2289 | 			}
2290 | 			
2291 | 		} while ($depth > 0);
2292 | 		
2293 | 		#
2294 | 		# Hash last block text that wasn't processed inside the loop.
2295 | 		#
2296 | 		$parsed .= $this->$hash_method($block_text);
2297 | 		
2298 | 		return array($parsed, $text);
2299 | 	}
2300 | 
2301 | 
2302 | 	function hashClean($text) {
2303 | 	#
2304 | 	# Called whenever a tag must be hashed when a function inserts a "clean" tag
2305 | 	# in $text, it passes through this function and is automaticaly escaped, 
2306 | 	# blocking invalid nested overlap.
2307 | 	#
2308 | 		return $this->hashPart($text, 'C');
2309 | 	}
2310 | 
2311 | 
2312 | 	function doHeaders($text) {
2313 | 	#
2314 | 	# Redefined to add id and class attribute support.
2315 | 	#
2316 | 		# Setext-style headers:
2317 | 		#	  Header 1  {#header1}
2318 | 		#	  ========
2319 | 		#  
2320 | 		#	  Header 2  {#header2 .class1 .class2}
2321 | 		#	  --------
2322 | 		#
2323 | 		$text = preg_replace_callback(
2324 | 			'{
2325 | 				(^.+?)								# $1: Header text
2326 | 				(?:[ ]+ '.$this->id_class_attr_catch_re.' )?	 # $3 = id/class attributes
2327 | 				[ ]*\n(=+|-+)[ ]*\n+				# $3: Header footer
2328 | 			}mx',
2329 | 			array(&$this, '_doHeaders_callback_setext'), $text);
2330 | 
2331 | 		# atx-style headers:
2332 | 		#	# Header 1        {#header1}
2333 | 		#	## Header 2       {#header2}
2334 | 		#	## Header 2 with closing hashes ##  {#header3.class1.class2}
2335 | 		#	...
2336 | 		#	###### Header 6   {.class2}
2337 | 		#
2338 | 		$text = preg_replace_callback('{
2339 | 				^(\#{1,6})	# $1 = string of #\'s
2340 | 				[ ]*
2341 | 				(.+?)		# $2 = Header text
2342 | 				[ ]*
2343 | 				\#*			# optional closing #\'s (not counted)
2344 | 				(?:[ ]+ '.$this->id_class_attr_catch_re.' )?	 # $3 = id/class attributes
2345 | 				[ ]*
2346 | 				\n+
2347 | 			}xm',
2348 | 			array(&$this, '_doHeaders_callback_atx'), $text);
2349 | 
2350 | 		return $text;
2351 | 	}
2352 | 	function _doHeaders_callback_setext($matches) {
2353 | 		if ($matches[3] == '-' && preg_match('{^- }', $matches[1]))
2354 | 			return $matches[0];
2355 | 		$level = $matches[3]{0} == '=' ? 1 : 2;
2356 | 		$attr  = $this->doExtraAttributes("h$level", $dummy =& $matches[2]);
2357 | 		$block = "".$this->runSpanGamut($matches[1])."";
2358 | 		return "\n" . $this->hashBlock($block) . "\n\n";
2359 | 	}
2360 | 	function _doHeaders_callback_atx($matches) {
2361 | 		$level = strlen($matches[1]);
2362 | 		$attr  = $this->doExtraAttributes("h$level", $dummy =& $matches[3]);
2363 | 		$block = "".$this->runSpanGamut($matches[2])."";
2364 | 		return "\n" . $this->hashBlock($block) . "\n\n";
2365 | 	}
2366 | 
2367 | 
2368 | 	function doTables($text) {
2369 | 	#
2370 | 	# Form HTML tables.
2371 | 	#
2372 | 		$less_than_tab = $this->tab_width - 1;
2373 | 		#
2374 | 		# Find tables with leading pipe.
2375 | 		#
2376 | 		#	| Header 1 | Header 2
2377 | 		#	| -------- | --------
2378 | 		#	| Cell 1   | Cell 2
2379 | 		#	| Cell 3   | Cell 4
2380 | 		#
2381 | 		$text = preg_replace_callback('
2382 | 			{
2383 | 				^							# Start of a line
2384 | 				[ ]{0,'.$less_than_tab.'}	# Allowed whitespace.
2385 | 				[|]							# Optional leading pipe (present)
2386 | 				(.+) \n						# $1: Header row (at least one pipe)
2387 | 				
2388 | 				[ ]{0,'.$less_than_tab.'}	# Allowed whitespace.
2389 | 				[|] ([ ]*[-:]+[-| :]*) \n	# $2: Header underline
2390 | 				
2391 | 				(							# $3: Cells
2392 | 					(?>
2393 | 						[ ]*				# Allowed whitespace.
2394 | 						[|] .* \n			# Row content.
2395 | 					)*
2396 | 				)
2397 | 				(?=\n|\Z)					# Stop at final double newline.
2398 | 			}xm',
2399 | 			array(&$this, '_doTable_leadingPipe_callback'), $text);
2400 | 		
2401 | 		#
2402 | 		# Find tables without leading pipe.
2403 | 		#
2404 | 		#	Header 1 | Header 2
2405 | 		#	-------- | --------
2406 | 		#	Cell 1   | Cell 2
2407 | 		#	Cell 3   | Cell 4
2408 | 		#
2409 | 		$text = preg_replace_callback('
2410 | 			{
2411 | 				^							# Start of a line
2412 | 				[ ]{0,'.$less_than_tab.'}	# Allowed whitespace.
2413 | 				(\S.*[|].*) \n				# $1: Header row (at least one pipe)
2414 | 				
2415 | 				[ ]{0,'.$less_than_tab.'}	# Allowed whitespace.
2416 | 				([-:]+[ ]*[|][-| :]*) \n	# $2: Header underline
2417 | 				
2418 | 				(							# $3: Cells
2419 | 					(?>
2420 | 						.* [|] .* \n		# Row content
2421 | 					)*
2422 | 				)
2423 | 				(?=\n|\Z)					# Stop at final double newline.
2424 | 			}xm',
2425 | 			array(&$this, '_DoTable_callback'), $text);
2426 | 
2427 | 		return $text;
2428 | 	}
2429 | 	function _doTable_leadingPipe_callback($matches) {
2430 | 		$head		= $matches[1];
2431 | 		$underline	= $matches[2];
2432 | 		$content	= $matches[3];
2433 | 		
2434 | 		# Remove leading pipe for each row.
2435 | 		$content	= preg_replace('/^ *[|]/m', '', $content);
2436 | 		
2437 | 		return $this->_doTable_callback(array($matches[0], $head, $underline, $content));
2438 | 	}
2439 | 	function _doTable_callback($matches) {
2440 | 		$head		= $matches[1];
2441 | 		$underline	= $matches[2];
2442 | 		$content	= $matches[3];
2443 | 
2444 | 		# Remove any tailing pipes for each line.
2445 | 		$head		= preg_replace('/[|] *$/m', '', $head);
2446 | 		$underline	= preg_replace('/[|] *$/m', '', $underline);
2447 | 		$content	= preg_replace('/[|] *$/m', '', $content);
2448 | 		
2449 | 		# Reading alignement from header underline.
2450 | 		$separators	= preg_split('/ *[|] */', $underline);
2451 | 		foreach ($separators as $n => $s) {
2452 | 			if (preg_match('/^ *-+: *$/', $s))		$attr[$n] = ' align="right"';
2453 | 			else if (preg_match('/^ *:-+: *$/', $s))$attr[$n] = ' align="center"';
2454 | 			else if (preg_match('/^ *:-+ *$/', $s))	$attr[$n] = ' align="left"';
2455 | 			else									$attr[$n] = '';
2456 | 		}
2457 | 		
2458 | 		# Parsing span elements, including code spans, character escapes, 
2459 | 		# and inline HTML tags, so that pipes inside those gets ignored.
2460 | 		$head		= $this->parseSpan($head);
2461 | 		$headers	= preg_split('/ *[|] */', $head);
2462 | 		$col_count	= count($headers);
2463 | 		
2464 | 		# Write column headers.
2465 | 		$text = "\n";
2466 | 		$text .= "\n";
2467 | 		$text .= "\n";
2468 | 		foreach ($headers as $n => $header)
2469 | 			$text .= "  ".$this->runSpanGamut(trim($header))."\n";
2470 | 		$text .= "\n";
2471 | 		$text .= "\n";
2472 | 		
2473 | 		# Split content by row.
2474 | 		$rows = explode("\n", trim($content, "\n"));
2475 | 		
2476 | 		$text .= "\n";
2477 | 		foreach ($rows as $row) {
2478 | 			# Parsing span elements, including code spans, character escapes, 
2479 | 			# and inline HTML tags, so that pipes inside those gets ignored.
2480 | 			$row = $this->parseSpan($row);
2481 | 			
2482 | 			# Split row by cell.
2483 | 			$row_cells = preg_split('/ *[|] */', $row, $col_count);
2484 | 			$row_cells = array_pad($row_cells, $col_count, '');
2485 | 			
2486 | 			$text .= "\n";
2487 | 			foreach ($row_cells as $n => $cell)
2488 | 				$text .= "  ".$this->runSpanGamut(trim($cell))."\n";
2489 | 			$text .= "\n";
2490 | 		}
2491 | 		$text .= "\n";
2492 | 		$text .= "
    ";
2493 | 		
2494 | 		return $this->hashBlock($text) . "\n";
2495 | 	}
2496 | 
2497 | 	
2498 | 	function doDefLists($text) {
2499 | 	#
2500 | 	# Form HTML definition lists.
2501 | 	#
2502 | 		$less_than_tab = $this->tab_width - 1;
2503 | 
2504 | 		# Re-usable pattern to match any entire dl list:
2505 | 		$whole_list_re = '(?>
2506 | 			(								# $1 = whole list
2507 | 			  (								# $2
2508 | 				[ ]{0,'.$less_than_tab.'}
2509 | 				((?>.*\S.*\n)+)				# $3 = defined term
2510 | 				\n?
2511 | 				[ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition
2512 | 			  )
2513 | 			  (?s:.+?)
2514 | 			  (								# $4
2515 | 				  \z
2516 | 				|
2517 | 				  \n{2,}
2518 | 				  (?=\S)
2519 | 				  (?!						# Negative lookahead for another term
2520 | 					[ ]{0,'.$less_than_tab.'}
2521 | 					(?: \S.*\n )+?			# defined term
2522 | 					\n?
2523 | 					[ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition
2524 | 				  )
2525 | 				  (?!						# Negative lookahead for another definition
2526 | 					[ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition
2527 | 				  )
2528 | 			  )
2529 | 			)
2530 | 		)'; // mx
2531 | 
2532 | 		$text = preg_replace_callback('{
2533 | 				(?>\A\n?|(?<=\n\n))
2534 | 				'.$whole_list_re.'
2535 | 			}mx',
2536 | 			array(&$this, '_doDefLists_callback'), $text);
2537 | 
2538 | 		return $text;
2539 | 	}
2540 | 	function _doDefLists_callback($matches) {
2541 | 		# Re-usable patterns to match list item bullets and number markers:
2542 | 		$list = $matches[1];
2543 | 		
2544 | 		# Turn double returns into triple returns, so that we can make a
2545 | 		# paragraph for the last item in a list, if necessary:
2546 | 		$result = trim($this->processDefListItems($list));
2547 | 		$result = "
    \n" . $result . "\n
    ";
2548 | 		return $this->hashBlock($result) . "\n\n";
2549 | 	}
2550 | 
2551 | 
2552 | 	function processDefListItems($list_str) {
2553 | 	#
2554 | 	#	Process the contents of a single definition list, splitting it
2555 | 	#	into individual term and definition list items.
2556 | 	#
2557 | 		$less_than_tab = $this->tab_width - 1;
2558 | 		
2559 | 		# trim trailing blank lines:
2560 | 		$list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str);
2561 | 
2562 | 		# Process definition terms.
2563 | 		$list_str = preg_replace_callback('{
2564 | 			(?>\A\n?|\n\n+)					# leading line
2565 | 			(								# definition terms = $1
2566 | 				[ ]{0,'.$less_than_tab.'}	# leading whitespace
2567 | 				(?![:][ ]|[ ])				# negative lookahead for a definition 
2568 | 											#   mark (colon) or more whitespace.
2569 | 				(?> \S.* \n)+?				# actual term (not whitespace).	
2570 | 			)			
2571 | 			(?=\n?[ ]{0,3}:[ ])				# lookahead for following line feed 
2572 | 											#   with a definition mark.
2573 | 			}xm',
2574 | 			array(&$this, '_processDefListItems_callback_dt'), $list_str);
2575 | 
2576 | 		# Process actual definitions.
2577 | 		$list_str = preg_replace_callback('{
2578 | 			\n(\n+)?						# leading line = $1
2579 | 			(								# marker space = $2
2580 | 				[ ]{0,'.$less_than_tab.'}	# whitespace before colon
2581 | 				[:][ ]+						# definition mark (colon)
2582 | 			)
2583 | 			((?s:.+?))						# definition text = $3
2584 | 			(?= \n+ 						# stop at next definition mark,
2585 | 				(?:							# next term or end of text
2586 | 					[ ]{0,'.$less_than_tab.'} [:][ ]	|
2587 | 					
     | \z
2588 | 				)						
2589 | 			)					
2590 | 			}xm',
2591 | 			array(&$this, '_processDefListItems_callback_dd'), $list_str);
2592 | 
2593 | 		return $list_str;
2594 | 	}
2595 | 	function _processDefListItems_callback_dt($matches) {
2596 | 		$terms = explode("\n", trim($matches[1]));
2597 | 		$text = '';
2598 | 		foreach ($terms as $term) {
2599 | 			$term = $this->runSpanGamut(trim($term));
2600 | 			$text .= "\n
    " . $term . "
    ";
2601 | 		}
2602 | 		return $text . "\n";
2603 | 	}
2604 | 	function _processDefListItems_callback_dd($matches) {
2605 | 		$leading_line	= $matches[1];
2606 | 		$marker_space	= $matches[2];
2607 | 		$def			= $matches[3];
2608 | 
2609 | 		if ($leading_line || preg_match('/\n{2,}/', $def)) {
2610 | 			# Replace marker with the appropriate whitespace indentation
2611 | 			$def = str_repeat(' ', strlen($marker_space)) . $def;
2612 | 			$def = $this->runBlockGamut($this->outdent($def . "\n\n"));
2613 | 			$def = "\n". $def ."\n";
2614 | 		}
2615 | 		else {
2616 | 			$def = rtrim($def);
2617 | 			$def = $this->runSpanGamut($this->outdent($def));
2618 | 		}
2619 | 
2620 | 		return "\n
    " . $def . "
    \n";
2621 | 	}
2622 | 
2623 | 
2624 | 	function doFencedCodeBlocks($text) {
2625 | 	#
2626 | 	# Adding the fenced code block syntax to regular Markdown:
2627 | 	#
2628 | 	# ~~~
2629 | 	# Code block
2630 | 	# ~~~
2631 | 	#
2632 | 		$less_than_tab = $this->tab_width;
2633 | 		
2634 | 		$text = preg_replace_callback('{
2635 | 				(?:\n|\A)
2636 | 				# 1: Opening marker
2637 | 				(
2638 | 					~{3,} # Marker: three tilde or more.
2639 | 				)
2640 | 				[ ]*
2641 | 				(?:
2642 | 					[.]?([-_:a-zA-Z0-9]+) # 2: standalone class name
2643 | 				|
2644 | 					'.$this->id_class_attr_catch_re.' # 3: Extra attributes
2645 | 				)?
2646 | 				[ ]* \n # Whitespace and newline following marker.
2647 | 				
2648 | 				# 4: Content
2649 | 				(
2650 | 					(?>
2651 | 						(?!\1 [ ]* \n)	# Not a closing marker.
2652 | 						.*\n+
2653 | 					)+
2654 | 				)
2655 | 				
2656 | 				# Closing marker.
2657 | 				\1 [ ]* \n
2658 | 			}xm',
2659 | 			array(&$this, '_doFencedCodeBlocks_callback'), $text);
2660 | 
2661 | 		return $text;
2662 | 	}
2663 | 	function _doFencedCodeBlocks_callback($matches) {
2664 | 		$classname =& $matches[2];
2665 | 		$attrs     =& $matches[3];
2666 | 		$codeblock = $matches[4];
2667 | 		$codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES);
2668 | 		$codeblock = preg_replace_callback('/^\n+/',
2669 | 			array(&$this, '_doFencedCodeBlocks_newlines'), $codeblock);
2670 | 
2671 | 		if ($classname != "") {
2672 | 			if ($classname{0} == '.')
2673 | 				$classname = substr($classname, 1);
2674 | 			$attr_str = ' class="'.$this->code_class_prefix.$classname.'"';
2675 | 		} else {
2676 | 			$attr_str = $this->doExtraAttributes($this->code_attr_on_pre ? "pre" : "code", $attrs);
2677 | 		}
2678 | 		$pre_attr_str  = $this->code_attr_on_pre ? $attr_str : '';
2679 | 		$code_attr_str = $this->code_attr_on_pre ? '' : $attr_str;
2680 | 		$codeblock  = "$codeblock
    ";
2681 | 		
2682 | 		return "\n\n".$this->hashBlock($codeblock)."\n\n";
2683 | 	}
2684 | 	function _doFencedCodeBlocks_newlines($matches) {
2685 | 		return str_repeat("empty_element_suffix", 
2686 | 			strlen($matches[0]));
2687 | 	}
2688 | 
2689 | 
2690 | 	#
2691 | 	# Redefining emphasis markers so that emphasis by underscore does not
2692 | 	# work in the middle of a word.
2693 | 	#
2694 | 	var $em_relist = array(
2695 | 		''  => '(?:(? '(?<=\S|^)(? '(?<=\S|^)(? '(?:(? '(?<=\S|^)(? '(?<=\S|^)(? '(?:(? '(?<=\S|^)(? '(?<=\S|^)(? tags
2715 | 	#
2716 | 		# Strip leading and trailing lines:
2717 | 		$text = preg_replace('/\A\n+|\n+\z/', '', $text);
2718 | 		
2719 | 		$grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY);
2720 | 
2721 | 		#
2722 | 		# Wrap 
     tags and unhashify HTML blocks
2723 | 		#
2724 | 		foreach ($grafs as $key => $value) {
2725 | 			$value = trim($this->runSpanGamut($value));
2726 | 			
2727 | 			# Check if this should be enclosed in a paragraph.
2728 | 			# Clean tag hashes & block tag hashes are left alone.
2729 | 			$is_p = !preg_match('/^B\x1A[0-9]+B|^C\x1A[0-9]+C$/', $value);
2730 | 			
2731 | 			if ($is_p) {
2732 | 				$value = "
    $value
    ";
2733 | 			}
2734 | 			$grafs[$key] = $value;
2735 | 		}
2736 | 		
2737 | 		# Join grafs in one text, then unhash HTML tags. 
2738 | 		$text = implode("\n\n", $grafs);
2739 | 		
2740 | 		# Finish by removing any tag hashes still present in $text.
2741 | 		$text = $this->unhash($text);
2742 | 		
2743 | 		return $text;
2744 | 	}
2745 | 	
2746 | 	
2747 | 	### Footnotes
2748 | 	
2749 | 	function stripFootnotes($text) {
2750 | 	#
2751 | 	# Strips link definitions from text, stores the URLs and titles in
2752 | 	# hash references.
2753 | 	#
2754 | 		$less_than_tab = $this->tab_width - 1;
2755 | 
2756 | 		# Link defs are in the form: [^id]: url "optional title"
2757 | 		$text = preg_replace_callback('{
2758 | 			^[ ]{0,'.$less_than_tab.'}\[\^(.+?)\][ ]?:	# note_id = $1
2759 | 			  [ ]*
2760 | 			  \n?					# maybe *one* newline
2761 | 			(						# text = $2 (no blank lines allowed)
2762 | 				(?:					
2763 | 					.+				# actual text
2764 | 				|
2765 | 					\n				# newlines but 
2766 | 					(?!\[\^.+?\]:\s)# negative lookahead for footnote marker.
2767 | 					(?!\n+[ ]{0,3}\S)# ensure line is not blank and followed 
2768 | 									# by non-indented content
2769 | 				)*
2770 | 			)		
2771 | 			}xm',
2772 | 			array(&$this, '_stripFootnotes_callback'),
2773 | 			$text);
2774 | 		return $text;
2775 | 	}
2776 | 	function _stripFootnotes_callback($matches) {
2777 | 		$note_id = $this->fn_id_prefix . $matches[1];
2778 | 		$this->footnotes[$note_id] = $this->outdent($matches[2]);
2779 | 		return ''; # String that will replace the block
2780 | 	}
2781 | 
2782 | 
2783 | 	function doFootnotes($text) {
2784 | 	#
2785 | 	# Replace footnote references in $text [^id] with a special text-token 
2786 | 	# which will be replaced by the actual footnote marker in appendFootnotes.
2787 | 	#
2788 | 		if (!$this->in_anchor) {
2789 | 			$text = preg_replace('{\[\^(.+?)\]}', "F\x1Afn:\\1\x1A:", $text);
2790 | 		}
2791 | 		return $text;
2792 | 	}
2793 | 
2794 | 	
2795 | 	function appendFootnotes($text) {
2796 | 	#
2797 | 	# Append footnote list to text.
2798 | 	#
2799 | 		$text = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', 
2800 | 			array(&$this, '_appendFootnotes_callback'), $text);
2801 | 	
2802 | 		if (!empty($this->footnotes_ordered)) {
2803 | 			$text .= "\n\n";
2804 | 			$text .= "
    \n";
2805 | 			$text .= "empty_element_suffix ."\n";
2806 | 			$text .= "
      \n\n";
2807 | 			
2808 | 			$attr = " rev=\"footnote\"";
2809 | 			if ($this->fn_backlink_class != "") {
2810 | 				$class = $this->fn_backlink_class;
2811 | 				$class = $this->encodeAttribute($class);
2812 | 				$attr .= " class=\"$class\"";
2813 | 			}
2814 | 			if ($this->fn_backlink_title != "") {
2815 | 				$title = $this->fn_backlink_title;
2816 | 				$title = $this->encodeAttribute($title);
2817 | 				$attr .= " title=\"$title\"";
2818 | 			}
2819 | 			$num = 0;
2820 | 			
2821 | 			while (!empty($this->footnotes_ordered)) {
2822 | 				$footnote = reset($this->footnotes_ordered);
2823 | 				$note_id = key($this->footnotes_ordered);
2824 | 				unset($this->footnotes_ordered[$note_id]);
2825 | 				$ref_count = $this->footnotes_ref_count[$note_id];
2826 | 				unset($this->footnotes_ref_count[$note_id]);
2827 | 				unset($this->footnotes[$note_id]);
2828 | 				
2829 | 				$footnote .= "\n"; # Need to append newline before parsing.
2830 | 				$footnote = $this->runBlockGamut("$footnote\n");				
2831 | 				$footnote = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', 
2832 | 					array(&$this, '_appendFootnotes_callback'), $footnote);
2833 | 				
2834 | 				$attr = str_replace("%%", ++$num, $attr);
2835 | 				$note_id = $this->encodeAttribute($note_id);
2836 | 
2837 | 				# Prepare backlink, multiple backlinks if multiple references
2838 | 				$backlink = "↩";
2839 | 				for ($ref_num = 2; $ref_num <= $ref_count; ++$ref_num) {
2840 | 					$backlink .= " ↩";
2841 | 				}
2842 | 				# Add backlink to last paragraph; create new paragraph if needed.
2843 | 				if (preg_match('{
      $}', $footnote)) {
2844 | 					$footnote = substr($footnote, 0, -4) . " $backlink
      ";
2845 | 				} else {
2846 | 					$footnote .= "\n\n
      $backlink
      ";
2847 | 				}
2848 | 				
2849 | 				$text .= "
    1. \n";
2850 | 				$text .= $footnote . "\n";
2851 | 				$text .= "
      2. \n\n";
2852 | 			}
2853 | 			
2854 | 			$text .= "
    \n";
2855 | 			$text .= "
    ";
2856 | 		}
2857 | 		return $text;
2858 | 	}
2859 | 	function _appendFootnotes_callback($matches) {
2860 | 		$node_id = $this->fn_id_prefix . $matches[1];
2861 | 		
2862 | 		# Create footnote marker only if it has a corresponding footnote *and*
2863 | 		# the footnote hasn't been used by another marker.
2864 | 		if (isset($this->footnotes[$node_id])) {
2865 | 			$num =& $this->footnotes_numbers[$node_id];
2866 | 			if (!isset($num)) {
2867 | 				# Transfer footnote content to the ordered list and give it its
2868 | 				# number
2869 | 				$this->footnotes_ordered[$node_id] = $this->footnotes[$node_id];
2870 | 				$this->footnotes_ref_count[$node_id] = 1;
2871 | 				$num = $this->footnote_counter++;
2872 | 				$ref_count_mark = '';
2873 | 			} else {
2874 | 				$ref_count_mark = $this->footnotes_ref_count[$node_id] += 1;
2875 | 			}
2876 | 			
2877 | 			$attr = " rel=\"footnote\"";
2878 | 			if ($this->fn_link_class != "") {
2879 | 				$class = $this->fn_link_class;
2880 | 				$class = $this->encodeAttribute($class);
2881 | 				$attr .= " class=\"$class\"";
2882 | 			}
2883 | 			if ($this->fn_link_title != "") {
2884 | 				$title = $this->fn_link_title;
2885 | 				$title = $this->encodeAttribute($title);
2886 | 				$attr .= " title=\"$title\"";
2887 | 			}
2888 | 			
2889 | 			$attr = str_replace("%%", $num, $attr);
2890 | 			$node_id = $this->encodeAttribute($node_id);
2891 | 			
2892 | 			return
2893 | 				"".
2894 | 				"$num".
2895 | 				"";
2896 | 		}
2897 | 		
2898 | 		return "[^".$matches[1]."]";
2899 | 	}
2900 | 		
2901 | 	
2902 | 	### Abbreviations ###
2903 | 	
2904 | 	function stripAbbreviations($text) {
2905 | 	#
2906 | 	# Strips abbreviations from text, stores titles in hash references.
2907 | 	#
2908 | 		$less_than_tab = $this->tab_width - 1;
2909 | 
2910 | 		# Link defs are in the form: [id]*: url "optional title"
2911 | 		$text = preg_replace_callback('{
2912 | 			^[ ]{0,'.$less_than_tab.'}\*\[(.+?)\][ ]?:	# abbr_id = $1
2913 | 			(.*)					# text = $2 (no blank lines allowed)	
2914 | 			}xm',
2915 | 			array(&$this, '_stripAbbreviations_callback'),
2916 | 			$text);
2917 | 		return $text;
2918 | 	}
2919 | 	function _stripAbbreviations_callback($matches) {
2920 | 		$abbr_word = $matches[1];
2921 | 		$abbr_desc = $matches[2];
2922 | 		if ($this->abbr_word_re)
2923 | 			$this->abbr_word_re .= '|';
2924 | 		$this->abbr_word_re .= preg_quote($abbr_word);
2925 | 		$this->abbr_desciptions[$abbr_word] = trim($abbr_desc);
2926 | 		return ''; # String that will replace the block
2927 | 	}
2928 | 	
2929 | 	
2930 | 	function doAbbreviations($text) {
2931 | 	#
2932 | 	# Find defined abbreviations in text and wrap them in  elements.
2933 | 	#
2934 | 		if ($this->abbr_word_re) {
2935 | 			// cannot use the /x modifier because abbr_word_re may 
2936 | 			// contain significant spaces:
2937 | 			$text = preg_replace_callback('{'.
2938 | 				'(?abbr_word_re.')'.
2940 | 				'(?![\w\x1A])'.
2941 | 				'}', 
2942 | 				array(&$this, '_doAbbreviations_callback'), $text);
2943 | 		}
2944 | 		return $text;
2945 | 	}
2946 | 	function _doAbbreviations_callback($matches) {
2947 | 		$abbr = $matches[0];
2948 | 		if (isset($this->abbr_desciptions[$abbr])) {
2949 | 			$desc = $this->abbr_desciptions[$abbr];
2950 | 			if (empty($desc)) {
2951 | 				return $this->hashPart("$abbr");
2952 | 			} else {
2953 | 				$desc = $this->encodeAttribute($desc);
2954 | 				return $this->hashPart("$abbr");
2955 | 			}
2956 | 		} else {
2957 | 			return $matches[0];
2958 | 		}
2959 | 	}
2960 | 
2961 | }
2962 | 
2963 | 
2964 | /*
2965 | 
2966 | PHP Markdown Extra
2967 | ==================
2968 | 
2969 | Description
2970 | -----------
2971 | 
2972 | This is a PHP port of the original Markdown formatter written in Perl 
2973 | by John Gruber. This special "Extra" version of PHP Markdown features 
2974 | further enhancements to the syntax for making additional constructs 
2975 | such as tables and definition list.
2976 | 
2977 | Markdown is a text-to-HTML filter; it translates an easy-to-read /
2978 | easy-to-write structured text format into HTML. Markdown's text format
2979 | is mostly similar to that of plain text email, and supports features such
2980 | as headers, *emphasis*, code blocks, blockquotes, and links.
2981 | 
2982 | Markdown's syntax is designed not as a generic markup language, but
2983 | specifically to serve as a front-end to (X)HTML. You can use span-level
2984 | HTML tags anywhere in a Markdown document, and you can use block level
2985 | HTML tags (like 
     and  as well).
2986 | 
2987 | For more information about Markdown's syntax, see:
2988 | 
2989 | 
2990 | 
2991 | 
2992 | Bugs
2993 | ----
2994 | 
2995 | To file bug reports please send email to:
2996 | 
2997 | 
2998 | 
2999 | Please include with your report: (1) the example input; (2) the output you
3000 | expected; (3) the output Markdown actually produced.
3001 | 
3002 | 
3003 | Version History
3004 | --------------- 
3005 | 
3006 | See the readme file for detailed release notes for this version.
3007 | 
3008 | 
3009 | Copyright and License
3010 | ---------------------
3011 | 
3012 | PHP Markdown & Extra
3013 | Copyright (c) 2004-2013 Michel Fortin
3014 |   
3015 | All rights reserved.
3016 | 
3017 | Based on Markdown  
3018 | Copyright (c) 2003-2006 John Gruber   
3019 |    
3020 | All rights reserved.
3021 | 
3022 | Redistribution and use in source and binary forms, with or without
3023 | modification, are permitted provided that the following conditions are
3024 | met:
3025 | 
3026 | *	Redistributions of source code must retain the above copyright notice,
3027 | 	this list of conditions and the following disclaimer.
3028 | 
3029 | *	Redistributions in binary form must reproduce the above copyright
3030 | 	notice, this list of conditions and the following disclaimer in the
3031 | 	documentation and/or other materials provided with the distribution.
3032 | 
3033 | *	Neither the name "Markdown" nor the names of its contributors may
3034 | 	be used to endorse or promote products derived from this software
3035 | 	without specific prior written permission.
3036 | 
3037 | This software is provided by the copyright holders and contributors "as
3038 | is" and any express or implied warranties, including, but not limited
3039 | to, the implied warranties of merchantability and fitness for a
3040 | particular purpose are disclaimed. In no event shall the copyright owner
3041 | or contributors be liable for any direct, indirect, incidental, special,
3042 | exemplary, or consequential damages (including, but not limited to,
3043 | procurement of substitute goods or services; loss of use, data, or
3044 | profits; or business interruption) however caused and on any theory of
3045 | liability, whether in contract, strict liability, or tort (including
3046 | negligence or otherwise) arising in any way out of the use of this
3047 | software, even if advised of the possibility of such damage.
3048 | 
3049 | */
3050 | 


--------------------------------------------------------------------------------