extra.php
3230 lines
| 1 | <?php |
| 2 | # |
| 3 | # Markdown Extra - A text-to-HTML conversion tool for web writers |
| 4 | # |
| 5 | # PHP Markdown & Extra |
| 6 | # Copyright (c) 2004-2013 Michel Fortin |
| 7 | # <http://michelf.ca/projects/php-markdown/> |
| 8 | # |
| 9 | # Original Markdown |
| 10 | # Copyright (c) 2004-2006 John Gruber |
| 11 | # <http://daringfireball.net/projects/markdown/> |
| 12 | # |
| 13 | # Tweaked to remove WordPress interface |
| 14 | |
| 15 | if ( ! defined( 'ABSPATH' ) ) { |
| 16 | exit( 0 ); |
| 17 | } |
| 18 | |
| 19 | define( 'MARKDOWN_VERSION', "1.0.2" ); # 29 Nov 2013 |
| 20 | define( 'MARKDOWNEXTRA_VERSION', "1.2.8" ); # 29 Nov 2013 |
| 21 | |
| 22 | |
| 23 | # |
| 24 | # Global default settings: |
| 25 | # |
| 26 | |
| 27 | // phpcs:disable WordPress.PHP.NoSilencedErrors.Discouraged |
| 28 | |
| 29 | # Change to ">" for HTML output |
| 30 | @define( 'MARKDOWN_EMPTY_ELEMENT_SUFFIX', " />"); |
| 31 | |
| 32 | # Define the width of a tab for code blocks. |
| 33 | @define( 'MARKDOWN_TAB_WIDTH', 4 ); |
| 34 | |
| 35 | // Optional title attribute for footnote links and backlinks. |
| 36 | @define( 'MARKDOWN_FN_LINK_TITLE', esc_attr__( 'Read footnote.', 'jetpack' ) ); |
| 37 | @define( 'MARKDOWN_FN_BACKLINK_TITLE', esc_attr__( 'Return to main content.', 'jetpack' ) ); |
| 38 | |
| 39 | # Optional class attribute for footnote links and backlinks. |
| 40 | @define( 'MARKDOWN_FN_LINK_CLASS', "jetpack-footnote" ); |
| 41 | @define( 'MARKDOWN_FN_BACKLINK_CLASS', "" ); |
| 42 | |
| 43 | # Optional class prefix for fenced code block. |
| 44 | @define( 'MARKDOWN_CODE_CLASS_PREFIX', "language-" ); |
| 45 | |
| 46 | # Class attribute for code blocks goes on the `code` tag; |
| 47 | # setting this to true will put attributes on the `pre` tag instead. |
| 48 | @define( 'MARKDOWN_CODE_ATTR_ON_PRE', false ); |
| 49 | |
| 50 | |
| 51 | |
| 52 | ### Standard Function Interface ### |
| 53 | |
| 54 | @define( 'MARKDOWN_PARSER_CLASS', 'MarkdownExtra_Parser' ); |
| 55 | |
| 56 | // phpcs:enable WordPress.PHP.NoSilencedErrors.Discouraged |
| 57 | |
| 58 | function Markdown($text) { |
| 59 | # |
| 60 | # Initialize the parser and return the result of its transform method. |
| 61 | # |
| 62 | # Setup static parser variable. |
| 63 | static $parser; |
| 64 | if (!isset($parser)) { |
| 65 | $parser_class = MARKDOWN_PARSER_CLASS; |
| 66 | $parser = new $parser_class; |
| 67 | } |
| 68 | |
| 69 | # Transform text using parser. |
| 70 | return $parser->transform($text); |
| 71 | } |
| 72 | |
| 73 | /** |
| 74 | * Returns the length of $text loosely counting the number of UTF-8 characters with regular expression. |
| 75 | * Used by the Markdown_Parser class when mb_strlen is not available. |
| 76 | * |
| 77 | * @since 5.9 |
| 78 | * |
| 79 | * @return string Length of the multibyte string |
| 80 | * |
| 81 | */ |
| 82 | function jetpack_utf8_strlen( $text ) { |
| 83 | return preg_match_all( "/[\\x00-\\xBF]|[\\xC0-\\xFF][\\x80-\\xBF]*/", $text, $m ); |
| 84 | } |
| 85 | |
| 86 | # |
| 87 | # Markdown Parser Class |
| 88 | # |
| 89 | |
| 90 | class Markdown_Parser { |
| 91 | |
| 92 | ### Configuration Variables ### |
| 93 | |
| 94 | # Change to ">" for HTML output. |
| 95 | public $empty_element_suffix = MARKDOWN_EMPTY_ELEMENT_SUFFIX; |
| 96 | public $tab_width = MARKDOWN_TAB_WIDTH; |
| 97 | |
| 98 | # Change to `true` to disallow markup or entities. |
| 99 | public $no_markup = false; |
| 100 | public $no_entities = false; |
| 101 | |
| 102 | # Predefined urls and titles for reference links and images. |
| 103 | public $predef_urls = array(); |
| 104 | public $predef_titles = array(); |
| 105 | |
| 106 | |
| 107 | ### Parser Implementation ### |
| 108 | |
| 109 | # Regex to match balanced [brackets]. |
| 110 | # Needed to insert a maximum bracked depth while converting to PHP. |
| 111 | public $nested_brackets_depth = 6; |
| 112 | public $nested_brackets_re; |
| 113 | |
| 114 | public $nested_url_parenthesis_depth = 4; |
| 115 | public $nested_url_parenthesis_re; |
| 116 | |
| 117 | # Table of hash values for escaped characters: |
| 118 | public $escape_chars = '\`*_{}[]()>#+-.!'; |
| 119 | public $escape_chars_re; |
| 120 | |
| 121 | |
| 122 | function __construct() { |
| 123 | # |
| 124 | # Constructor function. Initialize appropriate member variables. |
| 125 | # |
| 126 | $this->_initDetab(); |
| 127 | $this->prepareItalicsAndBold(); |
| 128 | |
| 129 | $this->nested_brackets_re = |
| 130 | str_repeat('(?>[^\[\]]+|\[', $this->nested_brackets_depth). |
| 131 | str_repeat('\])*', $this->nested_brackets_depth); |
| 132 | |
| 133 | $this->nested_url_parenthesis_re = |
| 134 | str_repeat('(?>[^()\s]+|\(', $this->nested_url_parenthesis_depth). |
| 135 | str_repeat('(?>\)))*', $this->nested_url_parenthesis_depth); |
| 136 | |
| 137 | $this->escape_chars_re = '['.preg_quote($this->escape_chars).']'; |
| 138 | |
| 139 | # Sort document, block, and span gamut in ascendent priority order. |
| 140 | asort($this->document_gamut); |
| 141 | asort($this->block_gamut); |
| 142 | asort($this->span_gamut); |
| 143 | } |
| 144 | |
| 145 | |
| 146 | # Internal hashes used during transformation. |
| 147 | public $urls = array(); |
| 148 | public $titles = array(); |
| 149 | public $html_hashes = array(); |
| 150 | |
| 151 | # Status flag to avoid invalid nesting. |
| 152 | public $in_anchor = false; |
| 153 | |
| 154 | |
| 155 | function setup() { |
| 156 | # |
| 157 | # Called before the transformation process starts to setup parser |
| 158 | # states. |
| 159 | # |
| 160 | # Clear global hashes. |
| 161 | $this->urls = $this->predef_urls; |
| 162 | $this->titles = $this->predef_titles; |
| 163 | $this->html_hashes = array(); |
| 164 | |
| 165 | $this->in_anchor = false; |
| 166 | } |
| 167 | |
| 168 | function teardown() { |
| 169 | # |
| 170 | # Called after the transformation process to clear any variable |
| 171 | # which may be taking up memory unnecessarly. |
| 172 | # |
| 173 | $this->urls = array(); |
| 174 | $this->titles = array(); |
| 175 | $this->html_hashes = array(); |
| 176 | } |
| 177 | |
| 178 | |
| 179 | function transform($text) { |
| 180 | # |
| 181 | # Main function. Performs some preprocessing on the input text |
| 182 | # and pass it through the document gamut. |
| 183 | # |
| 184 | $this->setup(); |
| 185 | |
| 186 | # Remove UTF-8 BOM and marker character in input, if present. |
| 187 | $text = preg_replace('{^\xEF\xBB\xBF|\x1A}', '', $text); |
| 188 | |
| 189 | # Standardize line endings: |
| 190 | # DOS to Unix and Mac to Unix |
| 191 | $text = preg_replace('{\r\n?}', "\n", $text); |
| 192 | |
| 193 | # Make sure $text ends with a couple of newlines: |
| 194 | $text .= "\n\n"; |
| 195 | |
| 196 | # Convert all tabs to spaces. |
| 197 | $text = $this->detab($text); |
| 198 | |
| 199 | # Turn block-level HTML blocks into hash entries |
| 200 | $text = $this->hashHTMLBlocks($text); |
| 201 | |
| 202 | # Strip any lines consisting only of spaces and tabs. |
| 203 | # This makes subsequent regexen easier to write, because we can |
| 204 | # match consecutive blank lines with /\n+/ instead of something |
| 205 | # contorted like /[ ]*\n+/ . |
| 206 | $text = preg_replace('/^[ ]+$/m', '', $text); |
| 207 | |
| 208 | # Run document gamut methods. |
| 209 | foreach ($this->document_gamut as $method => $priority) { |
| 210 | $text = $this->$method($text); |
| 211 | } |
| 212 | |
| 213 | $this->teardown(); |
| 214 | |
| 215 | return $text . "\n"; |
| 216 | } |
| 217 | |
| 218 | public $document_gamut = array( |
| 219 | # Strip link definitions, store in hashes. |
| 220 | "stripLinkDefinitions" => 20, |
| 221 | |
| 222 | "runBasicBlockGamut" => 30, |
| 223 | ); |
| 224 | |
| 225 | |
| 226 | function stripLinkDefinitions($text) { |
| 227 | # |
| 228 | # Strips link definitions from text, stores the URLs and titles in |
| 229 | # hash references. |
| 230 | # |
| 231 | $less_than_tab = $this->tab_width - 1; |
| 232 | |
| 233 | # Link defs are in the form: ^[id]: url "optional title" |
| 234 | $text = preg_replace_callback('{ |
| 235 | ^[ ]{0,'.$less_than_tab.'}\[(.+)\][ ]?: # id = $1 |
| 236 | [ ]* |
| 237 | \n? # maybe *one* newline |
| 238 | [ ]* |
| 239 | (?: |
| 240 | <(.+?)> # url = $2 |
| 241 | | |
| 242 | (\S+?) # url = $3 |
| 243 | ) |
| 244 | [ ]* |
| 245 | \n? # maybe one newline |
| 246 | [ ]* |
| 247 | (?: |
| 248 | (?<=\s) # lookbehind for whitespace |
| 249 | ["(] |
| 250 | (.*?) # title = $4 |
| 251 | [")] |
| 252 | [ ]* |
| 253 | )? # title is optional |
| 254 | (?:\n+|\Z) |
| 255 | }xm', |
| 256 | array(&$this, '_stripLinkDefinitions_callback'), |
| 257 | $text); |
| 258 | return $text; |
| 259 | } |
| 260 | function _stripLinkDefinitions_callback($matches) { |
| 261 | $link_id = strtolower($matches[1]); |
| 262 | $url = $matches[2] == '' ? $matches[3] : $matches[2]; |
| 263 | $this->urls[$link_id] = $url; |
| 264 | $this->titles[$link_id] =& $matches[4]; |
| 265 | return ''; # String that will replace the block |
| 266 | } |
| 267 | |
| 268 | |
| 269 | function hashHTMLBlocks($text) { |
| 270 | if ($this->no_markup) return $text; |
| 271 | |
| 272 | $less_than_tab = $this->tab_width - 1; |
| 273 | |
| 274 | # Hashify HTML blocks: |
| 275 | # We only want to do this for block-level HTML tags, such as headers, |
| 276 | # lists, and tables. That's because we still want to wrap <p>s around |
| 277 | # "paragraphs" that are wrapped in non-block-level tags, such as anchors, |
| 278 | # phrase emphasis, and spans. The list of tags we're looking for is |
| 279 | # hard-coded: |
| 280 | # |
| 281 | # * List "a" is made of tags which can be both inline or block-level. |
| 282 | # These will be treated block-level when the start tag is alone on |
| 283 | # its line, otherwise they're not matched here and will be taken as |
| 284 | # inline later. |
| 285 | # * List "b" is made of tags which are always block-level; |
| 286 | # |
| 287 | $block_tags_a_re = 'ins|del'; |
| 288 | $block_tags_b_re = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|address|'. |
| 289 | 'script|noscript|form|fieldset|iframe|math|svg|'. |
| 290 | 'article|section|nav|aside|hgroup|header|footer|'. |
| 291 | 'figure'; |
| 292 | |
| 293 | # Regular expression for the content of a block tag. |
| 294 | $nested_tags_level = 4; |
| 295 | $attr = ' |
| 296 | (?> # optional tag attributes |
| 297 | \s # starts with whitespace |
| 298 | (?> |
| 299 | [^>"/]+ # text outside quotes |
| 300 | | |
| 301 | /+(?!>) # slash not followed by ">" |
| 302 | | |
| 303 | "[^"]*" # text inside double quotes (tolerate ">") |
| 304 | | |
| 305 | \'[^\']*\' # text inside single quotes (tolerate ">") |
| 306 | )* |
| 307 | )? |
| 308 | '; |
| 309 | $content = |
| 310 | str_repeat(' |
| 311 | (?> |
| 312 | [^<]+ # content without tag |
| 313 | | |
| 314 | <\2 # nested opening tag |
| 315 | '.$attr.' # attributes |
| 316 | (?> |
| 317 | /> |
| 318 | | |
| 319 | >', $nested_tags_level). # end of opening tag |
| 320 | '.*?'. # last level nested tag content |
| 321 | str_repeat(' |
| 322 | </\2\s*> # closing nested tag |
| 323 | ) |
| 324 | | |
| 325 | <(?!/\2\s*> # other tags with a different name |
| 326 | ) |
| 327 | )*', |
| 328 | $nested_tags_level); |
| 329 | $content2 = str_replace('\2', '\3', $content); |
| 330 | |
| 331 | # First, look for nested blocks, e.g.: |
| 332 | # <div> |
| 333 | # <div> |
| 334 | # tags for inner block must be indented. |
| 335 | # </div> |
| 336 | # </div> |
| 337 | # |
| 338 | # The outermost tags must start at the left margin for this to match, and |
| 339 | # the inner nested divs must be indented. |
| 340 | # We need to do this before the next, more liberal match, because the next |
| 341 | # match will start at the first `<div>` and stop at the first `</div>`. |
| 342 | $text = preg_replace_callback('{(?> |
| 343 | (?> |
| 344 | (?<=\n\n) # Starting after a blank line |
| 345 | | # or |
| 346 | \A\n? # the beginning of the doc |
| 347 | ) |
| 348 | ( # save in $1 |
| 349 | |
| 350 | # Match from `\n<tag>` to `</tag>\n`, handling nested tags |
| 351 | # in between. |
| 352 | |
| 353 | [ ]{0,'.$less_than_tab.'} |
| 354 | <('.$block_tags_b_re.')# start tag = $2 |
| 355 | '.$attr.'> # attributes followed by > and \n |
| 356 | '.$content.' # content, support nesting |
| 357 | </\2> # the matching end tag |
| 358 | [ ]* # trailing spaces/tabs |
| 359 | (?=\n+|\Z) # followed by a newline or end of document |
| 360 | |
| 361 | | # Special version for tags of group a. |
| 362 | |
| 363 | [ ]{0,'.$less_than_tab.'} |
| 364 | <('.$block_tags_a_re.')# start tag = $3 |
| 365 | '.$attr.'>[ ]*\n # attributes followed by > |
| 366 | '.$content2.' # content, support nesting |
| 367 | </\3> # the matching end tag |
| 368 | [ ]* # trailing spaces/tabs |
| 369 | (?=\n+|\Z) # followed by a newline or end of document |
| 370 | |
| 371 | | # Special case just for <hr />. It was easier to make a special |
| 372 | # case than to make the other regex more complicated. |
| 373 | |
| 374 | [ ]{0,'.$less_than_tab.'} |
| 375 | <(hr) # start tag = $2 |
| 376 | '.$attr.' # attributes |
| 377 | /?> # the matching end tag |
| 378 | [ ]* |
| 379 | (?=\n{2,}|\Z) # followed by a blank line or end of document |
| 380 | |
| 381 | | # Special case for standalone HTML comments: |
| 382 | |
| 383 | [ ]{0,'.$less_than_tab.'} |
| 384 | (?s: |
| 385 | <!-- .*? --> |
| 386 | ) |
| 387 | [ ]* |
| 388 | (?=\n{2,}|\Z) # followed by a blank line or end of document |
| 389 | |
| 390 | | # PHP and ASP-style processor instructions (<? and <%) |
| 391 | |
| 392 | [ ]{0,'.$less_than_tab.'} |
| 393 | (?s: |
| 394 | <([?%]) # $2 |
| 395 | .*? |
| 396 | \2> |
| 397 | ) |
| 398 | [ ]* |
| 399 | (?=\n{2,}|\Z) # followed by a blank line or end of document |
| 400 | |
| 401 | ) |
| 402 | )}Sxmi', |
| 403 | array(&$this, '_hashHTMLBlocks_callback'), |
| 404 | $text); |
| 405 | |
| 406 | return $text; |
| 407 | } |
| 408 | function _hashHTMLBlocks_callback($matches) { |
| 409 | $text = $matches[1]; |
| 410 | $key = $this->hashBlock($text); |
| 411 | return "\n\n$key\n\n"; |
| 412 | } |
| 413 | |
| 414 | |
| 415 | function hashPart($text, $boundary = 'X') { |
| 416 | # |
| 417 | # Called whenever a tag must be hashed when a function insert an atomic |
| 418 | # element in the text stream. Passing $text to through this function gives |
| 419 | # a unique text-token which will be reverted back when calling unhash. |
| 420 | # |
| 421 | # The $boundary argument specify what character should be used to surround |
| 422 | # the token. By convension, "B" is used for block elements that needs not |
| 423 | # to be wrapped into paragraph tags at the end, ":" is used for elements |
| 424 | # that are word separators and "X" is used in the general case. |
| 425 | # |
| 426 | # Swap back any tag hash found in $text so we do not have to `unhash` |
| 427 | # multiple times at the end. |
| 428 | $text = $this->unhash($text); |
| 429 | |
| 430 | # Then hash the block. |
| 431 | static $i = 0; |
| 432 | $key = "$boundary\x1A" . ++$i . $boundary; |
| 433 | $this->html_hashes[$key] = $text; |
| 434 | return $key; # String that will replace the tag. |
| 435 | } |
| 436 | |
| 437 | |
| 438 | function hashBlock($text) { |
| 439 | # |
| 440 | # Shortcut function for hashPart with block-level boundaries. |
| 441 | # |
| 442 | return $this->hashPart($text, 'B'); |
| 443 | } |
| 444 | |
| 445 | |
| 446 | public $block_gamut = array( |
| 447 | # |
| 448 | # These are all the transformations that form block-level |
| 449 | # tags like paragraphs, headers, and list items. |
| 450 | # |
| 451 | "doHeaders" => 10, |
| 452 | "doHorizontalRules" => 20, |
| 453 | |
| 454 | "doLists" => 40, |
| 455 | "doCodeBlocks" => 50, |
| 456 | "doBlockQuotes" => 60, |
| 457 | ); |
| 458 | |
| 459 | function runBlockGamut($text) { |
| 460 | # |
| 461 | # Run block gamut tranformations. |
| 462 | # |
| 463 | # We need to escape raw HTML in Markdown source before doing anything |
| 464 | # else. This need to be done for each block, and not only at the |
| 465 | # beginning in the Markdown function since hashed blocks can be part of |
| 466 | # list items and could have been indented. Indented blocks would have |
| 467 | # been seen as a code block in a previous pass of hashHTMLBlocks. |
| 468 | $text = $this->hashHTMLBlocks($text); |
| 469 | |
| 470 | return $this->runBasicBlockGamut($text); |
| 471 | } |
| 472 | |
| 473 | function runBasicBlockGamut($text) { |
| 474 | # |
| 475 | # Run block gamut tranformations, without hashing HTML blocks. This is |
| 476 | # useful when HTML blocks are known to be already hashed, like in the first |
| 477 | # whole-document pass. |
| 478 | # |
| 479 | foreach ($this->block_gamut as $method => $priority) { |
| 480 | $text = $this->$method($text); |
| 481 | } |
| 482 | |
| 483 | # Finally form paragraph and restore hashed blocks. |
| 484 | $text = $this->formParagraphs($text); |
| 485 | |
| 486 | return $text; |
| 487 | } |
| 488 | |
| 489 | |
| 490 | function doHorizontalRules($text) { |
| 491 | # Do Horizontal Rules: |
| 492 | return preg_replace( |
| 493 | '{ |
| 494 | ^[ ]{0,3} # Leading space |
| 495 | ([-*_]) # $1: First marker |
| 496 | (?> # Repeated marker group |
| 497 | [ ]{0,2} # Zero, one, or two spaces. |
| 498 | \1 # Marker character |
| 499 | ){2,} # Group repeated at least twice |
| 500 | [ ]* # Tailing spaces |
| 501 | $ # End of line. |
| 502 | }mx', |
| 503 | "\n".$this->hashBlock("<hr$this->empty_element_suffix")."\n", |
| 504 | $text); |
| 505 | } |
| 506 | |
| 507 | |
| 508 | public $span_gamut = array( |
| 509 | # |
| 510 | # These are all the transformations that occur *within* block-level |
| 511 | # tags like paragraphs, headers, and list items. |
| 512 | # |
| 513 | # Process character escapes, code spans, and inline HTML |
| 514 | # in one shot. |
| 515 | "parseSpan" => -30, |
| 516 | |
| 517 | # Process anchor and image tags. Images must come first, |
| 518 | # because ![foo][f] looks like an anchor. |
| 519 | "doImages" => 10, |
| 520 | "doAnchors" => 20, |
| 521 | |
| 522 | # Make links out of things like `<http://example.com/>` |
| 523 | # Must come after doAnchors, because you can use < and > |
| 524 | # delimiters in inline links like [this](<url>). |
| 525 | "doAutoLinks" => 30, |
| 526 | "encodeAmpsAndAngles" => 40, |
| 527 | |
| 528 | "doItalicsAndBold" => 50, |
| 529 | "doHardBreaks" => 60, |
| 530 | ); |
| 531 | |
| 532 | function runSpanGamut($text) { |
| 533 | # |
| 534 | # Run span gamut tranformations. |
| 535 | # |
| 536 | foreach ($this->span_gamut as $method => $priority) { |
| 537 | $text = $this->$method($text); |
| 538 | } |
| 539 | |
| 540 | return $text; |
| 541 | } |
| 542 | |
| 543 | |
| 544 | function doHardBreaks($text) { |
| 545 | # Do hard breaks: |
| 546 | return preg_replace_callback('/ {2,}\n/', |
| 547 | array(&$this, '_doHardBreaks_callback'), $text); |
| 548 | } |
| 549 | function _doHardBreaks_callback($matches) { |
| 550 | return $this->hashPart("<br$this->empty_element_suffix\n"); |
| 551 | } |
| 552 | |
| 553 | |
| 554 | function doAnchors($text) { |
| 555 | # |
| 556 | # Turn Markdown link shortcuts into XHTML <a> tags. |
| 557 | # |
| 558 | if ($this->in_anchor) return $text; |
| 559 | $this->in_anchor = true; |
| 560 | |
| 561 | # |
| 562 | # First, handle reference-style links: [link text] [id] |
| 563 | # |
| 564 | $text = preg_replace_callback('{ |
| 565 | ( # wrap whole match in $1 |
| 566 | \[ |
| 567 | ('.$this->nested_brackets_re.') # link text = $2 |
| 568 | \] |
| 569 | |
| 570 | [ ]? # one optional space |
| 571 | (?:\n[ ]*)? # one optional newline followed by spaces |
| 572 | |
| 573 | \[ |
| 574 | (.*?) # id = $3 |
| 575 | \] |
| 576 | ) |
| 577 | }xs', |
| 578 | array(&$this, '_doAnchors_reference_callback'), $text); |
| 579 | |
| 580 | # |
| 581 | # Next, inline-style links: [link text](url "optional title") |
| 582 | # |
| 583 | $text = preg_replace_callback('{ |
| 584 | ( # wrap whole match in $1 |
| 585 | \[ |
| 586 | ('.$this->nested_brackets_re.') # link text = $2 |
| 587 | \] |
| 588 | \( # literal paren |
| 589 | [ \n]* |
| 590 | (?: |
| 591 | <(.+?)> # href = $3 |
| 592 | | |
| 593 | ('.$this->nested_url_parenthesis_re.') # href = $4 |
| 594 | ) |
| 595 | [ \n]* |
| 596 | ( # $5 |
| 597 | ([\'"]) # quote char = $6 |
| 598 | (.*?) # Title = $7 |
| 599 | \6 # matching quote |
| 600 | [ \n]* # ignore any spaces/tabs between closing quote and ) |
| 601 | )? # title is optional |
| 602 | \) |
| 603 | ) |
| 604 | }xs', |
| 605 | array(&$this, '_doAnchors_inline_callback'), $text); |
| 606 | |
| 607 | # |
| 608 | # Last, handle reference-style shortcuts: [link text] |
| 609 | # These must come last in case you've also got [link text][1] |
| 610 | # or [link text](/foo) |
| 611 | # |
| 612 | $text = preg_replace_callback('{ |
| 613 | ( # wrap whole match in $1 |
| 614 | \[ |
| 615 | ([^\[\]]+) # link text = $2; can\'t contain [ or ] |
| 616 | \] |
| 617 | ) |
| 618 | }xs', |
| 619 | array(&$this, '_doAnchors_reference_callback'), $text); |
| 620 | |
| 621 | $this->in_anchor = false; |
| 622 | return $text; |
| 623 | } |
| 624 | function _doAnchors_reference_callback($matches) { |
| 625 | $whole_match = $matches[1]; |
| 626 | $link_text = $matches[2]; |
| 627 | $link_id =& $matches[3]; |
| 628 | |
| 629 | if ($link_id == "") { |
| 630 | # for shortcut links like [this][] or [this]. |
| 631 | $link_id = $link_text; |
| 632 | } |
| 633 | |
| 634 | # lower-case and turn embedded newlines into spaces |
| 635 | $link_id = strtolower($link_id); |
| 636 | $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); |
| 637 | |
| 638 | if (isset($this->urls[$link_id])) { |
| 639 | $url = $this->urls[$link_id]; |
| 640 | $url = $this->encodeAttribute($url); |
| 641 | |
| 642 | $result = "<a href=\"$url\""; |
| 643 | if ( isset( $this->titles[$link_id] ) ) { |
| 644 | $title = $this->titles[$link_id]; |
| 645 | $title = $this->encodeAttribute($title); |
| 646 | $result .= " title=\"$title\""; |
| 647 | } |
| 648 | |
| 649 | $link_text = $this->runSpanGamut($link_text); |
| 650 | $result .= ">$link_text</a>"; |
| 651 | $result = $this->hashPart($result); |
| 652 | } |
| 653 | else { |
| 654 | $result = $whole_match; |
| 655 | } |
| 656 | return $result; |
| 657 | } |
| 658 | function _doAnchors_inline_callback($matches) { |
| 659 | $whole_match = $matches[1]; |
| 660 | $link_text = $this->runSpanGamut($matches[2]); |
| 661 | $url = $matches[3] == '' ? $matches[4] : $matches[3]; |
| 662 | $title =& $matches[7]; |
| 663 | |
| 664 | $url = $this->encodeAttribute($url); |
| 665 | |
| 666 | $result = "<a href=\"$url\""; |
| 667 | if (isset($title)) { |
| 668 | $title = $this->encodeAttribute($title); |
| 669 | $result .= " title=\"$title\""; |
| 670 | } |
| 671 | |
| 672 | $link_text = $this->runSpanGamut($link_text); |
| 673 | $result .= ">$link_text</a>"; |
| 674 | |
| 675 | return $this->hashPart($result); |
| 676 | } |
| 677 | |
| 678 | |
| 679 | function doImages($text) { |
| 680 | # |
| 681 | # Turn Markdown image shortcuts into <img> tags. |
| 682 | # |
| 683 | # |
| 684 | # First, handle reference-style labeled images: ![alt text][id] |
| 685 | # |
| 686 | $text = preg_replace_callback('{ |
| 687 | ( # wrap whole match in $1 |
| 688 | !\[ |
| 689 | ('.$this->nested_brackets_re.') # alt text = $2 |
| 690 | \] |
| 691 | |
| 692 | [ ]? # one optional space |
| 693 | (?:\n[ ]*)? # one optional newline followed by spaces |
| 694 | |
| 695 | \[ |
| 696 | (.*?) # id = $3 |
| 697 | \] |
| 698 | |
| 699 | ) |
| 700 | }xs', |
| 701 | array(&$this, '_doImages_reference_callback'), $text); |
| 702 | |
| 703 | # |
| 704 | # Next, handle inline images:  |
| 705 | # Don't forget: encode * and _ |
| 706 | # |
| 707 | $text = preg_replace_callback('{ |
| 708 | ( # wrap whole match in $1 |
| 709 | !\[ |
| 710 | ('.$this->nested_brackets_re.') # alt text = $2 |
| 711 | \] |
| 712 | \s? # One optional whitespace character |
| 713 | \( # literal paren |
| 714 | [ \n]* |
| 715 | (?: |
| 716 | <(\S*)> # src url = $3 |
| 717 | | |
| 718 | ('.$this->nested_url_parenthesis_re.') # src url = $4 |
| 719 | ) |
| 720 | [ \n]* |
| 721 | ( # $5 |
| 722 | ([\'"]) # quote char = $6 |
| 723 | (.*?) # title = $7 |
| 724 | \6 # matching quote |
| 725 | [ \n]* |
| 726 | )? # title is optional |
| 727 | \) |
| 728 | ) |
| 729 | }xs', |
| 730 | array(&$this, '_doImages_inline_callback'), $text); |
| 731 | |
| 732 | return $text; |
| 733 | } |
| 734 | function _doImages_reference_callback($matches) { |
| 735 | $whole_match = $matches[1]; |
| 736 | $alt_text = $matches[2]; |
| 737 | $link_id = strtolower($matches[3]); |
| 738 | |
| 739 | if ($link_id == "") { |
| 740 | $link_id = strtolower($alt_text); # for shortcut links like ![this][]. |
| 741 | } |
| 742 | |
| 743 | $alt_text = $this->encodeAttribute($alt_text); |
| 744 | if (isset($this->urls[$link_id])) { |
| 745 | $url = $this->encodeAttribute($this->urls[$link_id]); |
| 746 | $result = "<img src=\"$url\" alt=\"$alt_text\""; |
| 747 | if (isset($this->titles[$link_id])) { |
| 748 | $title = $this->titles[$link_id]; |
| 749 | $title = $this->encodeAttribute($title); |
| 750 | $result .= " title=\"$title\""; |
| 751 | } |
| 752 | $result .= $this->empty_element_suffix; |
| 753 | $result = $this->hashPart($result); |
| 754 | } |
| 755 | else { |
| 756 | # If there's no such link ID, leave intact: |
| 757 | $result = $whole_match; |
| 758 | } |
| 759 | |
| 760 | return $result; |
| 761 | } |
| 762 | function _doImages_inline_callback($matches) { |
| 763 | $whole_match = $matches[1]; |
| 764 | $alt_text = $matches[2]; |
| 765 | $url = $matches[3] == '' ? $matches[4] : $matches[3]; |
| 766 | $title =& $matches[7]; |
| 767 | |
| 768 | $alt_text = $this->encodeAttribute($alt_text); |
| 769 | $url = $this->encodeAttribute($url); |
| 770 | $result = "<img src=\"$url\" alt=\"$alt_text\""; |
| 771 | if (isset($title)) { |
| 772 | $title = $this->encodeAttribute($title); |
| 773 | $result .= " title=\"$title\""; # $title already quoted |
| 774 | } |
| 775 | $result .= $this->empty_element_suffix; |
| 776 | |
| 777 | return $this->hashPart($result); |
| 778 | } |
| 779 | |
| 780 | |
| 781 | function doHeaders($text) { |
| 782 | # Setext-style headers: |
| 783 | # Header 1 |
| 784 | # ======== |
| 785 | # |
| 786 | # Header 2 |
| 787 | # -------- |
| 788 | # |
| 789 | $text = preg_replace_callback('{ ^(.+?)[ ]*\n(=+|-+)[ ]*\n+ }mx', |
| 790 | array(&$this, '_doHeaders_callback_setext'), $text); |
| 791 | |
| 792 | # atx-style headers: |
| 793 | # # Header 1 |
| 794 | # ## Header 2 |
| 795 | # ## Header 2 with closing hashes ## |
| 796 | # ... |
| 797 | # ###### Header 6 |
| 798 | # |
| 799 | $text = preg_replace_callback('{ |
| 800 | ^(\#{1,6}) # $1 = string of #\'s |
| 801 | [ ]* |
| 802 | (.+?) # $2 = Header text |
| 803 | [ ]* |
| 804 | \#* # optional closing #\'s (not counted) |
| 805 | \n+ |
| 806 | }xm', |
| 807 | array(&$this, '_doHeaders_callback_atx'), $text); |
| 808 | |
| 809 | return $text; |
| 810 | } |
| 811 | function _doHeaders_callback_setext($matches) { |
| 812 | # Terrible hack to check we haven't found an empty list item. |
| 813 | if ($matches[2] == '-' && preg_match('{^-(?: |$)}', $matches[1])) |
| 814 | return $matches[0]; |
| 815 | |
| 816 | $level = $matches[2][0] == '=' ? 1 : 2; |
| 817 | $block = "<h$level>".$this->runSpanGamut($matches[1])."</h$level>"; |
| 818 | return "\n" . $this->hashBlock($block) . "\n\n"; |
| 819 | } |
| 820 | function _doHeaders_callback_atx($matches) { |
| 821 | $level = strlen($matches[1]); |
| 822 | $block = "<h$level>".$this->runSpanGamut($matches[2])."</h$level>"; |
| 823 | return "\n" . $this->hashBlock($block) . "\n\n"; |
| 824 | } |
| 825 | |
| 826 | |
| 827 | function doLists($text) { |
| 828 | # |
| 829 | # Form HTML ordered (numbered) and unordered (bulleted) lists. |
| 830 | # |
| 831 | $less_than_tab = $this->tab_width - 1; |
| 832 | |
| 833 | # Re-usable patterns to match list item bullets and number markers: |
| 834 | $marker_ul_re = '[*+-]'; |
| 835 | $marker_ol_re = '\d+[\.]'; |
| 836 | $marker_any_re = "(?:$marker_ul_re|$marker_ol_re)"; |
| 837 | |
| 838 | $markers_relist = array( |
| 839 | $marker_ul_re => $marker_ol_re, |
| 840 | $marker_ol_re => $marker_ul_re, |
| 841 | ); |
| 842 | |
| 843 | foreach ($markers_relist as $marker_re => $other_marker_re) { |
| 844 | # Re-usable pattern to match any entirel ul or ol list: |
| 845 | $whole_list_re = ' |
| 846 | ( # $1 = whole list |
| 847 | ( # $2 |
| 848 | ([ ]{0,'.$less_than_tab.'}) # $3 = number of spaces |
| 849 | ('.$marker_re.') # $4 = first list item marker |
| 850 | [ ]+ |
| 851 | ) |
| 852 | (?s:.+?) |
| 853 | ( # $5 |
| 854 | \z |
| 855 | | |
| 856 | \n{2,} |
| 857 | (?=\S) |
| 858 | (?! # Negative lookahead for another list item marker |
| 859 | [ ]* |
| 860 | '.$marker_re.'[ ]+ |
| 861 | ) |
| 862 | | |
| 863 | (?= # Lookahead for another kind of list |
| 864 | \n |
| 865 | \3 # Must have the same indentation |
| 866 | '.$other_marker_re.'[ ]+ |
| 867 | ) |
| 868 | ) |
| 869 | ) |
| 870 | '; // mx |
| 871 | |
| 872 | # We use a different prefix before nested lists than top-level lists. |
| 873 | # See extended comment in _ProcessListItems(). |
| 874 | |
| 875 | if ($this->list_level) { |
| 876 | $text = preg_replace_callback('{ |
| 877 | ^ |
| 878 | '.$whole_list_re.' |
| 879 | }mx', |
| 880 | array(&$this, '_doLists_callback'), $text); |
| 881 | } |
| 882 | else { |
| 883 | $text = preg_replace_callback('{ |
| 884 | (?:(?<=\n)\n|\A\n?) # Must eat the newline |
| 885 | '.$whole_list_re.' |
| 886 | }mx', |
| 887 | array(&$this, '_doLists_callback'), $text); |
| 888 | } |
| 889 | } |
| 890 | |
| 891 | return $text; |
| 892 | } |
| 893 | function _doLists_callback($matches) { |
| 894 | # Re-usable patterns to match list item bullets and number markers: |
| 895 | $marker_ul_re = '[*+-]'; |
| 896 | $marker_ol_re = '\d+[\.]'; |
| 897 | $marker_any_re = "(?:$marker_ul_re|$marker_ol_re)"; |
| 898 | |
| 899 | $list = $matches[1]; |
| 900 | $list_type = preg_match("/$marker_ul_re/", $matches[4]) ? "ul" : "ol"; |
| 901 | |
| 902 | $marker_any_re = ( $list_type == "ul" ? $marker_ul_re : $marker_ol_re ); |
| 903 | |
| 904 | $list .= "\n"; |
| 905 | $result = $this->processListItems($list, $marker_any_re); |
| 906 | |
| 907 | $result = $this->hashBlock("<$list_type>\n" . $result . "</$list_type>"); |
| 908 | return "\n". $result ."\n\n"; |
| 909 | } |
| 910 | |
| 911 | public $list_level = 0; |
| 912 | |
| 913 | function processListItems($list_str, $marker_any_re) { |
| 914 | # |
| 915 | # Process the contents of a single ordered or unordered list, splitting it |
| 916 | # into individual list items. |
| 917 | # |
| 918 | # The $this->list_level global keeps track of when we're inside a list. |
| 919 | # Each time we enter a list, we increment it; when we leave a list, |
| 920 | # we decrement. If it's zero, we're not in a list anymore. |
| 921 | # |
| 922 | # We do this because when we're not inside a list, we want to treat |
| 923 | # something like this: |
| 924 | # |
| 925 | # I recommend upgrading to version |
| 926 | # 8. Oops, now this line is treated |
| 927 | # as a sub-list. |
| 928 | # |
| 929 | # As a single paragraph, despite the fact that the second line starts |
| 930 | # with a digit-period-space sequence. |
| 931 | # |
| 932 | # Whereas when we're inside a list (or sub-list), that line will be |
| 933 | # treated as the start of a sub-list. What a kludge, huh? This is |
| 934 | # an aspect of Markdown's syntax that's hard to parse perfectly |
| 935 | # without resorting to mind-reading. Perhaps the solution is to |
| 936 | # change the syntax rules such that sub-lists must start with a |
| 937 | # starting cardinal number; e.g. "1." or "a.". |
| 938 | |
| 939 | $this->list_level++; |
| 940 | |
| 941 | # trim trailing blank lines: |
| 942 | $list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str); |
| 943 | |
| 944 | $list_str = preg_replace_callback('{ |
| 945 | (\n)? # leading line = $1 |
| 946 | (^[ ]*) # leading whitespace = $2 |
| 947 | ('.$marker_any_re.' # list marker and space = $3 |
| 948 | (?:[ ]+|(?=\n)) # space only required if item is not empty |
| 949 | ) |
| 950 | ((?s:.*?)) # list item text = $4 |
| 951 | (?:(\n+(?=\n))|\n) # tailing blank line = $5 |
| 952 | (?= \n* (\z | \2 ('.$marker_any_re.') (?:[ ]+|(?=\n)))) |
| 953 | }xm', |
| 954 | array(&$this, '_processListItems_callback'), $list_str); |
| 955 | |
| 956 | $this->list_level--; |
| 957 | return $list_str; |
| 958 | } |
| 959 | function _processListItems_callback($matches) { |
| 960 | $item = $matches[4]; |
| 961 | $leading_line =& $matches[1]; |
| 962 | $leading_space =& $matches[2]; |
| 963 | $marker_space = $matches[3]; |
| 964 | $tailing_blank_line =& $matches[5]; |
| 965 | |
| 966 | if ($leading_line || $tailing_blank_line || |
| 967 | preg_match('/\n{2,}/', $item)) |
| 968 | { |
| 969 | # Replace marker with the appropriate whitespace indentation |
| 970 | $item = $leading_space . str_repeat(' ', strlen($marker_space)) . $item; |
| 971 | $item = $this->runBlockGamut($this->outdent($item)."\n"); |
| 972 | } |
| 973 | else { |
| 974 | # Recursion for sub-lists: |
| 975 | $item = $this->doLists($this->outdent($item)); |
| 976 | $item = preg_replace('/\n+$/', '', $item); |
| 977 | $item = $this->runSpanGamut($item); |
| 978 | } |
| 979 | |
| 980 | return "<li>" . $item . "</li>\n"; |
| 981 | } |
| 982 | |
| 983 | |
| 984 | function doCodeBlocks($text) { |
| 985 | # |
| 986 | # Process Markdown `<pre><code>` blocks. |
| 987 | # |
| 988 | $text = preg_replace_callback('{ |
| 989 | (?:\n\n|\A\n?) |
| 990 | ( # $1 = the code block -- one or more lines, starting with a space/tab |
| 991 | (?> |
| 992 | [ ]{'.$this->tab_width.'} # Lines must start with a tab or a tab-width of spaces |
| 993 | .*\n+ |
| 994 | )+ |
| 995 | ) |
| 996 | ((?=^[ ]{0,'.$this->tab_width.'}\S)|\Z) # Lookahead for non-space at line-start, or end of doc |
| 997 | }xm', |
| 998 | array(&$this, '_doCodeBlocks_callback'), $text); |
| 999 | |
| 1000 | return $text; |
| 1001 | } |
| 1002 | function _doCodeBlocks_callback($matches) { |
| 1003 | $codeblock = $matches[1]; |
| 1004 | |
| 1005 | $codeblock = $this->outdent($codeblock); |
| 1006 | $codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES); |
| 1007 | |
| 1008 | # trim leading newlines and trailing newlines |
| 1009 | $codeblock = preg_replace('/\A\n+|\n+\z/', '', $codeblock); |
| 1010 | |
| 1011 | $codeblock = "<pre><code>$codeblock\n</code></pre>"; |
| 1012 | return "\n\n".$this->hashBlock($codeblock)."\n\n"; |
| 1013 | } |
| 1014 | |
| 1015 | |
| 1016 | function makeCodeSpan($code) { |
| 1017 | # |
| 1018 | # Create a code span markup for $code. Called from handleSpanToken. |
| 1019 | # |
| 1020 | $code = htmlspecialchars(trim($code), ENT_NOQUOTES); |
| 1021 | return $this->hashPart("<code>$code</code>"); |
| 1022 | } |
| 1023 | |
| 1024 | |
| 1025 | public $em_relist = array( |
| 1026 | '' => '(?:(?<!\*)\*(?!\*)|(?<!_)_(?!_))(?=\S|$)(?![\.,:;]\s)', |
| 1027 | '*' => '(?<=\S|^)(?<!\*)\*(?!\*)', |
| 1028 | '_' => '(?<=\S|^)(?<!_)_(?!_)', |
| 1029 | ); |
| 1030 | public $strong_relist = array( |
| 1031 | '' => '(?:(?<!\*)\*\*(?!\*)|(?<!_)__(?!_))(?=\S|$)(?![\.,:;]\s)', |
| 1032 | '**' => '(?<=\S|^)(?<!\*)\*\*(?!\*)', |
| 1033 | '__' => '(?<=\S|^)(?<!_)__(?!_)', |
| 1034 | ); |
| 1035 | public $em_strong_relist = array( |
| 1036 | '' => '(?:(?<!\*)\*\*\*(?!\*)|(?<!_)___(?!_))(?=\S|$)(?![\.,:;]\s)', |
| 1037 | '***' => '(?<=\S|^)(?<!\*)\*\*\*(?!\*)', |
| 1038 | '___' => '(?<=\S|^)(?<!_)___(?!_)', |
| 1039 | ); |
| 1040 | public $em_strong_prepared_relist; |
| 1041 | |
| 1042 | function prepareItalicsAndBold() { |
| 1043 | # |
| 1044 | # Prepare regular expressions for searching emphasis tokens in any |
| 1045 | # context. |
| 1046 | # |
| 1047 | foreach ($this->em_relist as $em => $em_re) { |
| 1048 | foreach ($this->strong_relist as $strong => $strong_re) { |
| 1049 | # Construct list of allowed token expressions. |
| 1050 | $token_relist = array(); |
| 1051 | if (isset($this->em_strong_relist["$em$strong"])) { |
| 1052 | $token_relist[] = $this->em_strong_relist["$em$strong"]; |
| 1053 | } |
| 1054 | $token_relist[] = $em_re; |
| 1055 | $token_relist[] = $strong_re; |
| 1056 | |
| 1057 | # Construct master expression from list. |
| 1058 | $token_re = '{('. implode('|', $token_relist) .')}'; |
| 1059 | $this->em_strong_prepared_relist["$em$strong"] = $token_re; |
| 1060 | } |
| 1061 | } |
| 1062 | } |
| 1063 | |
| 1064 | function doItalicsAndBold($text) { |
| 1065 | $token_stack = array(''); |
| 1066 | $text_stack = array(''); |
| 1067 | $em = ''; |
| 1068 | $strong = ''; |
| 1069 | $tree_char_em = false; |
| 1070 | |
| 1071 | while (1) { |
| 1072 | # |
| 1073 | # Get prepared regular expression for seraching emphasis tokens |
| 1074 | # in current context. |
| 1075 | # |
| 1076 | $token_re = $this->em_strong_prepared_relist["$em$strong"]; |
| 1077 | |
| 1078 | # |
| 1079 | # Each loop iteration search for the next emphasis token. |
| 1080 | # Each token is then passed to handleSpanToken. |
| 1081 | # |
| 1082 | $parts = preg_split($token_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE); |
| 1083 | $text_stack[0] .= $parts[0]; |
| 1084 | $token =& $parts[1]; |
| 1085 | $text =& $parts[2]; |
| 1086 | |
| 1087 | if (empty($token)) { |
| 1088 | # Reached end of text span: empty stack without emitting. |
| 1089 | # any more emphasis. |
| 1090 | while ($token_stack[0]) { |
| 1091 | $text_stack[1] .= array_shift($token_stack); |
| 1092 | $text_stack[0] .= array_shift($text_stack); |
| 1093 | } |
| 1094 | break; |
| 1095 | } |
| 1096 | |
| 1097 | $token_len = strlen($token); |
| 1098 | if ($tree_char_em) { |
| 1099 | # Reached closing marker while inside a three-char emphasis. |
| 1100 | if ($token_len == 3) { |
| 1101 | # Three-char closing marker, close em and strong. |
| 1102 | array_shift($token_stack); |
| 1103 | $span = array_shift($text_stack); |
| 1104 | $span = $this->runSpanGamut($span); |
| 1105 | $span = "<strong><em>$span</em></strong>"; |
| 1106 | $text_stack[0] .= $this->hashPart($span); |
| 1107 | $em = ''; |
| 1108 | $strong = ''; |
| 1109 | } else { |
| 1110 | # Other closing marker: close one em or strong and |
| 1111 | # change current token state to match the other |
| 1112 | $token_stack[0] = str_repeat($token[0], 3-$token_len); |
| 1113 | $tag = $token_len == 2 ? "strong" : "em"; |
| 1114 | $span = $text_stack[0]; |
| 1115 | $span = $this->runSpanGamut($span); |
| 1116 | $span = "<$tag>$span</$tag>"; |
| 1117 | $text_stack[0] = $this->hashPart($span); |
| 1118 | $$tag = ''; # $$tag stands for $em or $strong |
| 1119 | } |
| 1120 | $tree_char_em = false; |
| 1121 | } else if ($token_len == 3) { |
| 1122 | if ($em) { |
| 1123 | # Reached closing marker for both em and strong. |
| 1124 | # Closing strong marker: |
| 1125 | for ($i = 0; $i < 2; ++$i) { |
| 1126 | $shifted_token = array_shift($token_stack); |
| 1127 | $tag = strlen($shifted_token) == 2 ? "strong" : "em"; |
| 1128 | $span = array_shift($text_stack); |
| 1129 | $span = $this->runSpanGamut($span); |
| 1130 | $span = "<$tag>$span</$tag>"; |
| 1131 | $text_stack[0] .= $this->hashPart($span); |
| 1132 | $$tag = ''; # $$tag stands for $em or $strong |
| 1133 | } |
| 1134 | } else { |
| 1135 | # Reached opening three-char emphasis marker. Push on token |
| 1136 | # stack; will be handled by the special condition above. |
| 1137 | $em = $token[0]; |
| 1138 | $strong = "$em$em"; |
| 1139 | array_unshift($token_stack, $token); |
| 1140 | array_unshift($text_stack, ''); |
| 1141 | $tree_char_em = true; |
| 1142 | } |
| 1143 | } else if ($token_len == 2) { |
| 1144 | if ($strong) { |
| 1145 | # Unwind any dangling emphasis marker: |
| 1146 | if (strlen($token_stack[0]) == 1) { |
| 1147 | $text_stack[1] .= array_shift($token_stack); |
| 1148 | $text_stack[0] .= array_shift($text_stack); |
| 1149 | } |
| 1150 | # Closing strong marker: |
| 1151 | array_shift($token_stack); |
| 1152 | $span = array_shift($text_stack); |
| 1153 | $span = $this->runSpanGamut($span); |
| 1154 | $span = "<strong>$span</strong>"; |
| 1155 | $text_stack[0] .= $this->hashPart($span); |
| 1156 | $strong = ''; |
| 1157 | } else { |
| 1158 | array_unshift($token_stack, $token); |
| 1159 | array_unshift($text_stack, ''); |
| 1160 | $strong = $token; |
| 1161 | } |
| 1162 | } else { |
| 1163 | # Here $token_len == 1 |
| 1164 | if ($em) { |
| 1165 | if (strlen($token_stack[0]) == 1) { |
| 1166 | # Closing emphasis marker: |
| 1167 | array_shift($token_stack); |
| 1168 | $span = array_shift($text_stack); |
| 1169 | $span = $this->runSpanGamut($span); |
| 1170 | $span = "<em>$span</em>"; |
| 1171 | $text_stack[0] .= $this->hashPart($span); |
| 1172 | $em = ''; |
| 1173 | } else { |
| 1174 | $text_stack[0] .= $token; |
| 1175 | } |
| 1176 | } else { |
| 1177 | array_unshift($token_stack, $token); |
| 1178 | array_unshift($text_stack, ''); |
| 1179 | $em = $token; |
| 1180 | } |
| 1181 | } |
| 1182 | } |
| 1183 | return $text_stack[0]; |
| 1184 | } |
| 1185 | |
| 1186 | |
| 1187 | function doBlockQuotes($text) { |
| 1188 | $text = preg_replace_callback('/ |
| 1189 | ( # Wrap whole match in $1 |
| 1190 | (?> |
| 1191 | ^[ ]*>[ ]? # ">" at the start of a line |
| 1192 | .+\n # rest of the first line |
| 1193 | (.+\n)* # subsequent consecutive lines |
| 1194 | \n* # blanks |
| 1195 | )+ |
| 1196 | ) |
| 1197 | /xm', |
| 1198 | array(&$this, '_doBlockQuotes_callback'), $text); |
| 1199 | |
| 1200 | return $text; |
| 1201 | } |
| 1202 | function _doBlockQuotes_callback($matches) { |
| 1203 | $bq = $matches[1]; |
| 1204 | # trim one level of quoting - trim whitespace-only lines |
| 1205 | $bq = preg_replace('/^[ ]*>[ ]?|^[ ]+$/m', '', $bq); |
| 1206 | $bq = $this->runBlockGamut($bq); # recurse |
| 1207 | |
| 1208 | $bq = preg_replace('/^/m', " ", $bq); |
| 1209 | # These leading spaces cause problem with <pre> content, |
| 1210 | # so we need to fix that: |
| 1211 | $bq = preg_replace_callback('{(\s*<pre>.+?</pre>)}sx', |
| 1212 | array(&$this, '_doBlockQuotes_callback2'), $bq); |
| 1213 | |
| 1214 | return "\n". $this->hashBlock("<blockquote>\n$bq\n</blockquote>")."\n\n"; |
| 1215 | } |
| 1216 | function _doBlockQuotes_callback2($matches) { |
| 1217 | $pre = $matches[1]; |
| 1218 | $pre = preg_replace('/^ /m', '', $pre); |
| 1219 | return $pre; |
| 1220 | } |
| 1221 | |
| 1222 | |
| 1223 | function formParagraphs($text) { |
| 1224 | # |
| 1225 | # Params: |
| 1226 | # $text - string to process with html <p> tags |
| 1227 | # |
| 1228 | # Strip leading and trailing lines: |
| 1229 | $text = preg_replace('/\A\n+|\n+\z/', '', $text); |
| 1230 | |
| 1231 | $grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY); |
| 1232 | |
| 1233 | # |
| 1234 | # Wrap <p> tags and unhashify HTML blocks |
| 1235 | # |
| 1236 | foreach ($grafs as $key => $value) { |
| 1237 | if (!preg_match('/^B\x1A[0-9]+B$/', $value)) { |
| 1238 | # Is a paragraph. |
| 1239 | $value = $this->runSpanGamut($value); |
| 1240 | $value = preg_replace('/^([ ]*)/', "<p>", $value); |
| 1241 | $value .= "</p>"; |
| 1242 | $grafs[$key] = $this->unhash($value); |
| 1243 | } |
| 1244 | else { |
| 1245 | # Is a block. |
| 1246 | # Modify elements of @grafs in-place... |
| 1247 | $graf = $value; |
| 1248 | $block = $this->html_hashes[$graf]; |
| 1249 | $graf = $block; |
| 1250 | // if (preg_match('{ |
| 1251 | // \A |
| 1252 | // ( # $1 = <div> tag |
| 1253 | // <div \s+ |
| 1254 | // [^>]* |
| 1255 | // \b |
| 1256 | // markdown\s*=\s* ([\'"]) # $2 = attr quote char |
| 1257 | // 1 |
| 1258 | // \2 |
| 1259 | // [^>]* |
| 1260 | // > |
| 1261 | // ) |
| 1262 | // ( # $3 = contents |
| 1263 | // .* |
| 1264 | // ) |
| 1265 | // (</div>) # $4 = closing tag |
| 1266 | // \z |
| 1267 | // }xs', $block, $matches)) |
| 1268 | // { |
| 1269 | // list(, $div_open, , $div_content, $div_close) = $matches; |
| 1270 | // |
| 1271 | // # We can't call Markdown(), because that resets the hash; |
| 1272 | // # that initialization code should be pulled into its own sub, though. |
| 1273 | // $div_content = $this->hashHTMLBlocks($div_content); |
| 1274 | // |
| 1275 | // # Run document gamut methods on the content. |
| 1276 | // foreach ($this->document_gamut as $method => $priority) { |
| 1277 | // $div_content = $this->$method($div_content); |
| 1278 | // } |
| 1279 | // |
| 1280 | // $div_open = preg_replace( |
| 1281 | // '{\smarkdown\s*=\s*([\'"]).+?\1}', '', $div_open); |
| 1282 | // |
| 1283 | // $graf = $div_open . "\n" . $div_content . "\n" . $div_close; |
| 1284 | // } |
| 1285 | $grafs[$key] = $graf; |
| 1286 | } |
| 1287 | } |
| 1288 | |
| 1289 | return implode("\n\n", $grafs); |
| 1290 | } |
| 1291 | |
| 1292 | |
| 1293 | function encodeAttribute($text) { |
| 1294 | # |
| 1295 | # Encode text for a double-quoted HTML attribute. This function |
| 1296 | # is *not* suitable for attributes enclosed in single quotes. |
| 1297 | # |
| 1298 | $text = $this->encodeAmpsAndAngles($text); |
| 1299 | $text = str_replace('"', '"', $text); |
| 1300 | return $text; |
| 1301 | } |
| 1302 | |
| 1303 | |
| 1304 | function encodeAmpsAndAngles($text) { |
| 1305 | # |
| 1306 | # Smart processing for ampersands and angle brackets that need to |
| 1307 | # be encoded. Valid character entities are left alone unless the |
| 1308 | # no-entities mode is set. |
| 1309 | # |
| 1310 | if ($this->no_entities) { |
| 1311 | $text = str_replace('&', '&', $text); |
| 1312 | } else { |
| 1313 | # Ampersand-encoding based entirely on Nat Irons's Amputator |
| 1314 | # MT plugin: <http://bumppo.net/projects/amputator/> |
| 1315 | $text = preg_replace('/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/', |
| 1316 | '&', $text);; |
| 1317 | } |
| 1318 | # Encode remaining <'s |
| 1319 | $text = str_replace('<', '<', $text); |
| 1320 | |
| 1321 | return $text; |
| 1322 | } |
| 1323 | |
| 1324 | |
| 1325 | function doAutoLinks($text) { |
| 1326 | $text = preg_replace_callback('{<((https?|ftp|dict):[^\'">\s]+)>}i', |
| 1327 | array(&$this, '_doAutoLinks_url_callback'), $text); |
| 1328 | |
| 1329 | # Email addresses: <address@domain.foo> |
| 1330 | $text = preg_replace_callback('{ |
| 1331 | < |
| 1332 | (?:mailto:)? |
| 1333 | ( |
| 1334 | (?: |
| 1335 | [-!#$%&\'*+/=?^_`.{|}~\w\x80-\xFF]+ |
| 1336 | | |
| 1337 | ".*?" |
| 1338 | ) |
| 1339 | \@ |
| 1340 | (?: |
| 1341 | [-a-z0-9\x80-\xFF]+(\.[-a-z0-9\x80-\xFF]+)*\.[a-z]+ |
| 1342 | | |
| 1343 | \[[\d.a-fA-F:]+\] # IPv4 & IPv6 |
| 1344 | ) |
| 1345 | ) |
| 1346 | > |
| 1347 | }xi', |
| 1348 | array(&$this, '_doAutoLinks_email_callback'), $text); |
| 1349 | $text = preg_replace_callback('{<(tel:([^\'">\s]+))>}i',array(&$this, '_doAutoLinks_tel_callback'), $text); |
| 1350 | |
| 1351 | return $text; |
| 1352 | } |
| 1353 | function _doAutoLinks_tel_callback($matches) { |
| 1354 | $url = $this->encodeAttribute($matches[1]); |
| 1355 | $tel = $this->encodeAttribute($matches[2]); |
| 1356 | $link = "<a href=\"$url\">$tel</a>"; |
| 1357 | return $this->hashPart($link); |
| 1358 | } |
| 1359 | function _doAutoLinks_url_callback($matches) { |
| 1360 | $url = $this->encodeAttribute($matches[1]); |
| 1361 | $link = "<a href=\"$url\">$url</a>"; |
| 1362 | return $this->hashPart($link); |
| 1363 | } |
| 1364 | function _doAutoLinks_email_callback($matches) { |
| 1365 | $address = $matches[1]; |
| 1366 | $link = $this->encodeEmailAddress($address); |
| 1367 | return $this->hashPart($link); |
| 1368 | } |
| 1369 | |
| 1370 | |
| 1371 | function encodeEmailAddress($addr) { |
| 1372 | # |
| 1373 | # Input: an email address, e.g. "foo@example.com" |
| 1374 | # |
| 1375 | # Output: the email address as a mailto link, with each character |
| 1376 | # of the address encoded as either a decimal or hex entity, in |
| 1377 | # the hopes of foiling most address harvesting spam bots. E.g.: |
| 1378 | # |
| 1379 | # <p><a href="mailto:foo |
| 1380 | # @example.co |
| 1381 | # m">foo@exampl |
| 1382 | # e.com</a></p> |
| 1383 | # |
| 1384 | # Based by a filter by Matthew Wickline, posted to BBEdit-Talk. |
| 1385 | # With some optimizations by Milian Wolff. |
| 1386 | # |
| 1387 | $addr = "mailto:" . $addr; |
| 1388 | $chars = preg_split('/(?<!^)(?!$)/', $addr); |
| 1389 | $seed = (int)abs(crc32($addr) / strlen($addr)); # Deterministic seed. |
| 1390 | |
| 1391 | foreach ($chars as $key => $char) { |
| 1392 | $ord = ord($char); |
| 1393 | # Ignore non-ascii chars. |
| 1394 | if ($ord < 128) { |
| 1395 | $r = ($seed * (1 + $key)) % 100; # Pseudo-random function. |
| 1396 | # roughly 10% raw, 45% hex, 45% dec |
| 1397 | # '@' *must* be encoded. I insist. |
| 1398 | if ($r > 90 && $char != '@') /* do nothing */; |
| 1399 | else if ($r < 45) $chars[$key] = '&#x'.dechex($ord).';'; |
| 1400 | else $chars[$key] = '&#'.$ord.';'; |
| 1401 | } |
| 1402 | } |
| 1403 | |
| 1404 | $addr = implode('', $chars); |
| 1405 | $text = implode('', array_slice($chars, 7)); # text without `mailto:` |
| 1406 | $addr = "<a href=\"$addr\">$text</a>"; |
| 1407 | |
| 1408 | return $addr; |
| 1409 | } |
| 1410 | |
| 1411 | |
| 1412 | function parseSpan($str) { |
| 1413 | # |
| 1414 | # Take the string $str and parse it into tokens, hashing embedded HTML, |
| 1415 | # escaped characters and handling code spans. |
| 1416 | # |
| 1417 | $output = ''; |
| 1418 | |
| 1419 | $span_re = '{ |
| 1420 | ( |
| 1421 | \\\\'.$this->escape_chars_re.' |
| 1422 | | |
| 1423 | (?<![`\\\\]) |
| 1424 | `+ # code span marker |
| 1425 | '.( $this->no_markup ? '' : ' |
| 1426 | | |
| 1427 | <!-- .*? --> # comment |
| 1428 | | |
| 1429 | <\?.*?\?> | <%.*?%> # processing instruction |
| 1430 | | |
| 1431 | <[!$]?[-a-zA-Z0-9:_]+ # regular tags |
| 1432 | (?> |
| 1433 | \s |
| 1434 | (?>[^"\'>]+|"[^"]*"|\'[^\']*\')* |
| 1435 | )? |
| 1436 | > |
| 1437 | | |
| 1438 | <[-a-zA-Z0-9:_]+\s*/> # xml-style empty tag |
| 1439 | | |
| 1440 | </[-a-zA-Z0-9:_]+\s*> # closing tag |
| 1441 | ').' |
| 1442 | ) |
| 1443 | }xs'; |
| 1444 | |
| 1445 | while (1) { |
| 1446 | # |
| 1447 | # Each loop iteration search for either the next tag, the next |
| 1448 | # openning code span marker, or the next escaped character. |
| 1449 | # Each token is then passed to handleSpanToken. |
| 1450 | # |
| 1451 | $parts = preg_split($span_re, $str, 2, PREG_SPLIT_DELIM_CAPTURE); |
| 1452 | |
| 1453 | # Create token from text preceding tag. |
| 1454 | if ($parts[0] != "") { |
| 1455 | $output .= $parts[0]; |
| 1456 | } |
| 1457 | |
| 1458 | # Check if we reach the end. |
| 1459 | if (isset($parts[1])) { |
| 1460 | $output .= $this->handleSpanToken($parts[1], $parts[2]); |
| 1461 | $str = $parts[2]; |
| 1462 | } |
| 1463 | else { |
| 1464 | break; |
| 1465 | } |
| 1466 | } |
| 1467 | |
| 1468 | return $output; |
| 1469 | } |
| 1470 | |
| 1471 | |
| 1472 | function handleSpanToken($token, &$str) { |
| 1473 | # |
| 1474 | # Handle $token provided by parseSpan by determining its nature and |
| 1475 | # returning the corresponding value that should replace it. |
| 1476 | # |
| 1477 | switch ($token[0]) { |
| 1478 | case "\\": |
| 1479 | return $this->hashPart("&#". ord($token[1]). ";"); |
| 1480 | case "`": |
| 1481 | # Search for end marker in remaining text. |
| 1482 | if (preg_match('/^(.*?[^`])'.preg_quote($token).'(?!`)(.*)$/sm', |
| 1483 | $str, $matches)) |
| 1484 | { |
| 1485 | $str = $matches[2]; |
| 1486 | $codespan = $this->makeCodeSpan($matches[1]); |
| 1487 | return $this->hashPart($codespan); |
| 1488 | } |
| 1489 | return $token; // return as text since no ending marker found. |
| 1490 | default: |
| 1491 | return $this->hashPart($token); |
| 1492 | } |
| 1493 | } |
| 1494 | |
| 1495 | |
| 1496 | function outdent($text) { |
| 1497 | # |
| 1498 | # Remove one level of line-leading tabs or spaces |
| 1499 | # |
| 1500 | return preg_replace('/^(\t|[ ]{1,'.$this->tab_width.'})/m', '', $text); |
| 1501 | } |
| 1502 | |
| 1503 | |
| 1504 | # String length function for detab. `_initDetab` will create a function to |
| 1505 | # hanlde UTF-8 if the default function does not exist. |
| 1506 | public $utf8_strlen = 'mb_strlen'; |
| 1507 | |
| 1508 | function detab($text) { |
| 1509 | # |
| 1510 | # Replace tabs with the appropriate amount of space. |
| 1511 | # |
| 1512 | # For each line we separate the line in blocks delemited by |
| 1513 | # tab characters. Then we reconstruct every line by adding the |
| 1514 | # appropriate number of space between each blocks. |
| 1515 | |
| 1516 | $text = preg_replace_callback('/^.*\t.*$/m', |
| 1517 | array(&$this, '_detab_callback'), $text); |
| 1518 | |
| 1519 | return $text; |
| 1520 | } |
| 1521 | function _detab_callback($matches) { |
| 1522 | $line = $matches[0]; |
| 1523 | $strlen = $this->utf8_strlen; # strlen function for UTF-8. |
| 1524 | |
| 1525 | # Split in blocks. |
| 1526 | $blocks = explode("\t", $line); |
| 1527 | # Add each blocks to the line. |
| 1528 | $line = $blocks[0]; |
| 1529 | unset($blocks[0]); # Do not add first block twice. |
| 1530 | foreach ($blocks as $block) { |
| 1531 | # Calculate amount of space, insert spaces, insert block. |
| 1532 | $amount = $this->tab_width - |
| 1533 | $strlen($line, 'UTF-8') % $this->tab_width; |
| 1534 | $line .= str_repeat(" ", $amount) . $block; |
| 1535 | } |
| 1536 | return $line; |
| 1537 | } |
| 1538 | |
| 1539 | /** |
| 1540 | * Check for the availability of the function in the `utf8_strlen` property |
| 1541 | * (initially `mb_strlen`). If the function is not available, use jetpack_utf8_strlen |
| 1542 | * that will loosely count the number of UTF-8 characters with a regular expression. |
| 1543 | */ |
| 1544 | function _initDetab() { |
| 1545 | if ( function_exists( $this->utf8_strlen ) ) { |
| 1546 | return; |
| 1547 | } |
| 1548 | $this->utf8_strlen = 'jetpack_utf8_strlen'; |
| 1549 | } |
| 1550 | |
| 1551 | |
| 1552 | function unhash($text) { |
| 1553 | # |
| 1554 | # Swap back in all the tags hashed by _HashHTMLBlocks. |
| 1555 | # |
| 1556 | return preg_replace_callback('/(.)\x1A[0-9]+\1/', |
| 1557 | array(&$this, '_unhash_callback'), $text); |
| 1558 | } |
| 1559 | function _unhash_callback($matches) { |
| 1560 | return $this->html_hashes[$matches[0]]; |
| 1561 | } |
| 1562 | |
| 1563 | } |
| 1564 | |
| 1565 | |
| 1566 | # |
| 1567 | # Markdown Extra Parser Class |
| 1568 | # |
| 1569 | |
| 1570 | class MarkdownExtra_Parser extends Markdown_Parser { |
| 1571 | |
| 1572 | ### Configuration Variables ### |
| 1573 | |
| 1574 | # Prefix for footnote ids. |
| 1575 | public $fn_id_prefix = ""; |
| 1576 | |
| 1577 | # Optional title attribute for footnote links and backlinks. |
| 1578 | public $fn_link_title = MARKDOWN_FN_LINK_TITLE; |
| 1579 | public $fn_backlink_title = MARKDOWN_FN_BACKLINK_TITLE; |
| 1580 | |
| 1581 | # Optional class attribute for footnote links and backlinks. |
| 1582 | public $fn_link_class = MARKDOWN_FN_LINK_CLASS; |
| 1583 | public $fn_backlink_class = MARKDOWN_FN_BACKLINK_CLASS; |
| 1584 | |
| 1585 | # Optional class prefix for fenced code block. |
| 1586 | public $code_class_prefix = MARKDOWN_CODE_CLASS_PREFIX; |
| 1587 | # Class attribute for code blocks goes on the `code` tag; |
| 1588 | # setting this to true will put attributes on the `pre` tag instead. |
| 1589 | public $code_attr_on_pre = MARKDOWN_CODE_ATTR_ON_PRE; |
| 1590 | |
| 1591 | # Predefined abbreviations. |
| 1592 | public $predef_abbr = array(); |
| 1593 | |
| 1594 | /** |
| 1595 | * Reference attribute. |
| 1596 | * |
| 1597 | * @var array |
| 1598 | */ |
| 1599 | public $ref_attr; |
| 1600 | |
| 1601 | /** |
| 1602 | * Parsing mode. |
| 1603 | * |
| 1604 | * @var string |
| 1605 | */ |
| 1606 | public $mode; |
| 1607 | |
| 1608 | |
| 1609 | ### Parser Implementation ### |
| 1610 | |
| 1611 | function __construct() { |
| 1612 | # |
| 1613 | # Constructor function. Initialize the parser object. |
| 1614 | # |
| 1615 | # Add extra escapable characters before parent constructor |
| 1616 | # initialize the table. |
| 1617 | $this->escape_chars .= ':|'; |
| 1618 | |
| 1619 | # Insert extra document, block, and span transformations. |
| 1620 | # Parent constructor will do the sorting. |
| 1621 | $this->document_gamut += array( |
| 1622 | "doFencedCodeBlocks" => 5, |
| 1623 | "stripFootnotes" => 15, |
| 1624 | "stripAbbreviations" => 25, |
| 1625 | "appendFootnotes" => 50, |
| 1626 | ); |
| 1627 | $this->block_gamut += array( |
| 1628 | "doFencedCodeBlocks" => 5, |
| 1629 | "doTables" => 15, |
| 1630 | "doDefLists" => 45, |
| 1631 | ); |
| 1632 | $this->span_gamut += array( |
| 1633 | "doFootnotes" => 5, |
| 1634 | "doAbbreviations" => 70, |
| 1635 | ); |
| 1636 | |
| 1637 | parent::__construct(); |
| 1638 | } |
| 1639 | |
| 1640 | |
| 1641 | # Extra variables used during extra transformations. |
| 1642 | public $footnotes = array(); |
| 1643 | public $footnotes_ordered = array(); |
| 1644 | public $footnotes_ref_count = array(); |
| 1645 | public $footnotes_numbers = array(); |
| 1646 | public $abbr_desciptions = array(); |
| 1647 | public $abbr_word_re = ''; |
| 1648 | |
| 1649 | # Give the current footnote number. |
| 1650 | public $footnote_counter = 1; |
| 1651 | |
| 1652 | |
| 1653 | function setup() { |
| 1654 | # |
| 1655 | # Setting up Extra-specific variables. |
| 1656 | # |
| 1657 | parent::setup(); |
| 1658 | |
| 1659 | $this->footnotes = array(); |
| 1660 | $this->footnotes_ordered = array(); |
| 1661 | $this->footnotes_ref_count = array(); |
| 1662 | $this->footnotes_numbers = array(); |
| 1663 | $this->abbr_desciptions = array(); |
| 1664 | $this->abbr_word_re = ''; |
| 1665 | $this->footnote_counter = 1; |
| 1666 | |
| 1667 | foreach ($this->predef_abbr as $abbr_word => $abbr_desc) { |
| 1668 | if ($this->abbr_word_re) |
| 1669 | $this->abbr_word_re .= '|'; |
| 1670 | $this->abbr_word_re .= preg_quote($abbr_word); |
| 1671 | $this->abbr_desciptions[$abbr_word] = trim($abbr_desc); |
| 1672 | } |
| 1673 | } |
| 1674 | |
| 1675 | function teardown() { |
| 1676 | # |
| 1677 | # Clearing Extra-specific variables. |
| 1678 | # |
| 1679 | $this->footnotes = array(); |
| 1680 | $this->footnotes_ordered = array(); |
| 1681 | $this->footnotes_ref_count = array(); |
| 1682 | $this->footnotes_numbers = array(); |
| 1683 | $this->abbr_desciptions = array(); |
| 1684 | $this->abbr_word_re = ''; |
| 1685 | |
| 1686 | parent::teardown(); |
| 1687 | } |
| 1688 | |
| 1689 | |
| 1690 | ### Extra Attribute Parser ### |
| 1691 | |
| 1692 | # Expression to use to catch attributes (includes the braces) |
| 1693 | public $id_class_attr_catch_re = '\{((?:[ ]*[#.][-_:a-zA-Z0-9]+){1,})[ ]*\}'; |
| 1694 | # Expression to use when parsing in a context when no capture is desired |
| 1695 | public $id_class_attr_nocatch_re = '\{(?:[ ]*[#.][-_:a-zA-Z0-9]+){1,}[ ]*\}'; |
| 1696 | |
| 1697 | function doExtraAttributes($tag_name, $attr) { |
| 1698 | # |
| 1699 | # Parse attributes caught by the $this->id_class_attr_catch_re expression |
| 1700 | # and return the HTML-formatted list of attributes. |
| 1701 | # |
| 1702 | # Currently supported attributes are .class and #id. |
| 1703 | # |
| 1704 | if (empty($attr)) return ""; |
| 1705 | |
| 1706 | # Split on components |
| 1707 | preg_match_all('/[#.][-_:a-zA-Z0-9]+/', $attr, $matches); |
| 1708 | $elements = $matches[0]; |
| 1709 | |
| 1710 | # handle classes and ids (only first id taken into account) |
| 1711 | $classes = array(); |
| 1712 | $id = false; |
| 1713 | foreach ($elements as $element) { |
| 1714 | if ($element[0] == '.') { |
| 1715 | $classes[] = substr($element, 1); |
| 1716 | } else if ($element[0] == '#') { |
| 1717 | if ($id === false) $id = substr($element, 1); |
| 1718 | } |
| 1719 | } |
| 1720 | |
| 1721 | # compose attributes as string |
| 1722 | $attr_str = ""; |
| 1723 | if (!empty($id)) { |
| 1724 | $attr_str .= ' id="'.$id.'"'; |
| 1725 | } |
| 1726 | if (!empty($classes)) { |
| 1727 | $attr_str .= ' class="'.implode(" ", $classes).'"'; |
| 1728 | } |
| 1729 | return $attr_str; |
| 1730 | } |
| 1731 | |
| 1732 | |
| 1733 | function stripLinkDefinitions($text) { |
| 1734 | # |
| 1735 | # Strips link definitions from text, stores the URLs and titles in |
| 1736 | # hash references. |
| 1737 | # |
| 1738 | $less_than_tab = $this->tab_width - 1; |
| 1739 | |
| 1740 | # Link defs are in the form: ^[id]: url "optional title" |
| 1741 | $text = preg_replace_callback('{ |
| 1742 | ^[ ]{0,'.$less_than_tab.'}\[(.+)\][ ]?: # id = $1 |
| 1743 | [ ]* |
| 1744 | \n? # maybe *one* newline |
| 1745 | [ ]* |
| 1746 | (?: |
| 1747 | <(.+?)> # url = $2 |
| 1748 | | |
| 1749 | (\S+?) # url = $3 |
| 1750 | ) |
| 1751 | [ ]* |
| 1752 | \n? # maybe one newline |
| 1753 | [ ]* |
| 1754 | (?: |
| 1755 | (?<=\s) # lookbehind for whitespace |
| 1756 | ["(] |
| 1757 | (.*?) # title = $4 |
| 1758 | [")] |
| 1759 | [ ]* |
| 1760 | )? # title is optional |
| 1761 | (?:[ ]* '.$this->id_class_attr_catch_re.' )? # $5 = extra id & class attr |
| 1762 | (?:\n+|\Z) |
| 1763 | }xm', |
| 1764 | array(&$this, '_stripLinkDefinitions_callback'), |
| 1765 | $text); |
| 1766 | return $text; |
| 1767 | } |
| 1768 | function _stripLinkDefinitions_callback($matches) { |
| 1769 | $link_id = strtolower($matches[1]); |
| 1770 | $url = $matches[2] == '' ? $matches[3] : $matches[2]; |
| 1771 | $this->urls[$link_id] = $url; |
| 1772 | $this->titles[$link_id] =& $matches[4]; |
| 1773 | $this->ref_attr[$link_id] = $this->doExtraAttributes("", $dummy =& $matches[5]); |
| 1774 | return ''; # String that will replace the block |
| 1775 | } |
| 1776 | |
| 1777 | |
| 1778 | ### HTML Block Parser ### |
| 1779 | |
| 1780 | # Tags that are always treated as block tags: |
| 1781 | public $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'; |
| 1782 | |
| 1783 | # Tags treated as block tags only if the opening tag is alone on its line: |
| 1784 | public $context_block_tags_re = 'script|noscript|ins|del|iframe|object|source|track|param|math|svg|canvas|audio|video'; |
| 1785 | |
| 1786 | # Tags where markdown="1" default to span mode: |
| 1787 | public $contain_span_tags_re = 'p|h[1-6]|li|dd|dt|td|th|legend|address'; |
| 1788 | |
| 1789 | # Tags which must not have their contents modified, no matter where |
| 1790 | # they appear: |
| 1791 | public $clean_tags_re = 'script|math|svg'; |
| 1792 | |
| 1793 | # Tags that do not need to be closed. |
| 1794 | public $auto_close_tags_re = 'hr|img|param|source|track'; |
| 1795 | |
| 1796 | |
| 1797 | function hashHTMLBlocks($text) { |
| 1798 | # |
| 1799 | # Hashify HTML Blocks and "clean tags". |
| 1800 | # |
| 1801 | # We only want to do this for block-level HTML tags, such as headers, |
| 1802 | # lists, and tables. That's because we still want to wrap <p>s around |
| 1803 | # "paragraphs" that are wrapped in non-block-level tags, such as anchors, |
| 1804 | # phrase emphasis, and spans. The list of tags we're looking for is |
| 1805 | # hard-coded. |
| 1806 | # |
| 1807 | # This works by calling _HashHTMLBlocks_InMarkdown, which then calls |
| 1808 | # _HashHTMLBlocks_InHTML when it encounter block tags. When the markdown="1" |
| 1809 | # attribute is found within a tag, _HashHTMLBlocks_InHTML calls back |
| 1810 | # _HashHTMLBlocks_InMarkdown to handle the Markdown syntax within the tag. |
| 1811 | # These two functions are calling each other. It's recursive! |
| 1812 | # |
| 1813 | if ($this->no_markup) return $text; |
| 1814 | |
| 1815 | # |
| 1816 | # Call the HTML-in-Markdown hasher. |
| 1817 | # |
| 1818 | list($text, ) = $this->_hashHTMLBlocks_inMarkdown($text); |
| 1819 | |
| 1820 | return $text; |
| 1821 | } |
| 1822 | function _hashHTMLBlocks_inMarkdown($text, $indent = 0, |
| 1823 | $enclosing_tag_re = '', $span = false) |
| 1824 | { |
| 1825 | # |
| 1826 | # Parse markdown text, calling _HashHTMLBlocks_InHTML for block tags. |
| 1827 | # |
| 1828 | # * $indent is the number of space to be ignored when checking for code |
| 1829 | # blocks. This is important because if we don't take the indent into |
| 1830 | # account, something like this (which looks right) won't work as expected: |
| 1831 | # |
| 1832 | # <div> |
| 1833 | # <div markdown="1"> |
| 1834 | # Hello World. <-- Is this a Markdown code block or text? |
| 1835 | # </div> <-- Is this a Markdown code block or a real tag? |
| 1836 | # <div> |
| 1837 | # |
| 1838 | # If you don't like this, just don't indent the tag on which |
| 1839 | # you apply the markdown="1" attribute. |
| 1840 | # |
| 1841 | # * If $enclosing_tag_re is not empty, stops at the first unmatched closing |
| 1842 | # tag with that name. Nested tags supported. |
| 1843 | # |
| 1844 | # * If $span is true, text inside must treated as span. So any double |
| 1845 | # newline will be replaced by a single newline so that it does not create |
| 1846 | # paragraphs. |
| 1847 | # |
| 1848 | # Returns an array of that form: ( processed text , remaining text ) |
| 1849 | # |
| 1850 | if ($text === '') return array('', ''); |
| 1851 | |
| 1852 | # Regex to check for the presence of newlines around a block tag. |
| 1853 | $newline_before_re = '/(?:^\n?|\n\n)*$/'; |
| 1854 | $newline_after_re = |
| 1855 | '{ |
| 1856 | ^ # Start of text following the tag. |
| 1857 | (?>[ ]*<!--.*?-->)? # Optional comment. |
| 1858 | [ ]*\n # Must be followed by newline. |
| 1859 | }xs'; |
| 1860 | |
| 1861 | # Regex to match any tag. |
| 1862 | $block_tag_re = |
| 1863 | '{ |
| 1864 | ( # $2: Capture whole tag. |
| 1865 | </? # Any opening or closing tag. |
| 1866 | (?> # Tag name. |
| 1867 | '.$this->block_tags_re.' | |
| 1868 | '.$this->context_block_tags_re.' | |
| 1869 | '.$this->clean_tags_re.' | |
| 1870 | (?!\s)'.$enclosing_tag_re.' |
| 1871 | ) |
| 1872 | (?: |
| 1873 | (?=[\s"\'/a-zA-Z0-9]) # Allowed characters after tag name. |
| 1874 | (?> |
| 1875 | ".*?" | # Double quotes (can contain `>`) |
| 1876 | \'.*?\' | # Single quotes (can contain `>`) |
| 1877 | .+? # Anything but quotes and `>`. |
| 1878 | )*? |
| 1879 | )? |
| 1880 | > # End of tag. |
| 1881 | | |
| 1882 | <!-- .*? --> # HTML Comment |
| 1883 | | |
| 1884 | <\?.*?\?> | <%.*?%> # Processing instruction |
| 1885 | | |
| 1886 | <!\[CDATA\[.*?\]\]> # CData Block |
| 1887 | '. ( !$span ? ' # If not in span. |
| 1888 | | |
| 1889 | # Indented code block |
| 1890 | (?: ^[ ]*\n | ^ | \n[ ]*\n ) |
| 1891 | [ ]{'.($indent+4).'}[^\n]* \n |
| 1892 | (?> |
| 1893 | (?: [ ]{'.($indent+4).'}[^\n]* | [ ]* ) \n |
| 1894 | )* |
| 1895 | | |
| 1896 | # Fenced code block marker |
| 1897 | (?<= ^ | \n ) |
| 1898 | [ ]{0,'.($indent+3).'}(?:~{3,}|`{3,}) |
| 1899 | [ ]* |
| 1900 | (?: |
| 1901 | \.?[-_:a-zA-Z0-9]+ # standalone class name |
| 1902 | | |
| 1903 | '.$this->id_class_attr_nocatch_re.' # extra attributes |
| 1904 | )? |
| 1905 | [ ]* |
| 1906 | (?= \n ) |
| 1907 | ' : '' ). ' # End (if not is span). |
| 1908 | | |
| 1909 | # Code span marker |
| 1910 | # Note, this regex needs to go after backtick fenced |
| 1911 | # code blocks but it should also be kept outside of the |
| 1912 | # "if not in span" condition adding backticks to the parser |
| 1913 | `+ |
| 1914 | ) |
| 1915 | }xs'; |
| 1916 | |
| 1917 | |
| 1918 | $depth = 0; # Current depth inside the tag tree. |
| 1919 | $parsed = ""; # Parsed text that will be returned. |
| 1920 | |
| 1921 | # |
| 1922 | # Loop through every tag until we find the closing tag of the parent |
| 1923 | # or loop until reaching the end of text if no parent tag specified. |
| 1924 | # |
| 1925 | do { |
| 1926 | # |
| 1927 | # Split the text using the first $tag_match pattern found. |
| 1928 | # Text before pattern will be first in the array, text after |
| 1929 | # pattern will be at the end, and between will be any catches made |
| 1930 | # by the pattern. |
| 1931 | # |
| 1932 | $parts = preg_split($block_tag_re, $text, 2, |
| 1933 | PREG_SPLIT_DELIM_CAPTURE); |
| 1934 | |
| 1935 | # If in Markdown span mode, add a empty-string span-level hash |
| 1936 | # after each newline to prevent triggering any block element. |
| 1937 | if ($span) { |
| 1938 | $void = $this->hashPart("", ':'); |
| 1939 | $newline = "$void\n"; |
| 1940 | $parts[0] = $void . str_replace("\n", $newline, $parts[0]) . $void; |
| 1941 | } |
| 1942 | |
| 1943 | $parsed .= $parts[0]; # Text before current tag. |
| 1944 | |
| 1945 | # If end of $text has been reached. Stop loop. |
| 1946 | if (count($parts) < 3) { |
| 1947 | $text = ""; |
| 1948 | break; |
| 1949 | } |
| 1950 | |
| 1951 | $tag = $parts[1]; # Tag to handle. |
| 1952 | $text = $parts[2]; # Remaining text after current tag. |
| 1953 | $tag_re = preg_quote($tag); # For use in a regular expression. |
| 1954 | |
| 1955 | # |
| 1956 | # Check for: Fenced code block marker. |
| 1957 | # Note: need to recheck the whole tag to disambiguate backtick |
| 1958 | # fences from code spans |
| 1959 | # |
| 1960 | if (preg_match('{^\n?([ ]{0,'.($indent+3).'})(~{3,}|`{3,})[ ]*(?:\.?[-_:a-zA-Z0-9]+|'.$this->id_class_attr_nocatch_re.')?[ ]*\n?$}', $tag, $capture)) { |
| 1961 | # Fenced code block marker: find matching end marker. |
| 1962 | $fence_indent = strlen($capture[1]); # use captured indent in re |
| 1963 | $fence_re = $capture[2]; # use captured fence in re |
| 1964 | if (preg_match('{^(?>.*\n)*?[ ]{'.($fence_indent).'}'.$fence_re.'[ ]*(?:\n|$)}', $text, |
| 1965 | $matches)) |
| 1966 | { |
| 1967 | # End marker found: pass text unchanged until marker. |
| 1968 | $parsed .= $tag . $matches[0]; |
| 1969 | $text = substr($text, strlen($matches[0])); |
| 1970 | } |
| 1971 | else { |
| 1972 | # No end marker: just skip it. |
| 1973 | $parsed .= $tag; |
| 1974 | } |
| 1975 | } |
| 1976 | # |
| 1977 | # Check for: Indented code block. |
| 1978 | # |
| 1979 | else if ($tag[0] == "\n" || $tag[0] == " ") { |
| 1980 | # Indented code block: pass it unchanged, will be handled |
| 1981 | # later. |
| 1982 | $parsed .= $tag; |
| 1983 | } |
| 1984 | # |
| 1985 | # Check for: Code span marker |
| 1986 | # Note: need to check this after backtick fenced code blocks |
| 1987 | # |
| 1988 | else if ($tag[0] == "`") { |
| 1989 | # Find corresponding end marker. |
| 1990 | $tag_re = preg_quote($tag); |
| 1991 | if (preg_match('{^(?>.+?|\n(?!\n))*?(?<!`)'.$tag_re.'(?!`)}', |
| 1992 | $text, $matches)) |
| 1993 | { |
| 1994 | # End marker found: pass text unchanged until marker. |
| 1995 | $parsed .= $tag . $matches[0]; |
| 1996 | $text = substr($text, strlen($matches[0])); |
| 1997 | } |
| 1998 | else { |
| 1999 | # Unmatched marker: just skip it. |
| 2000 | $parsed .= $tag; |
| 2001 | } |
| 2002 | } |
| 2003 | # |
| 2004 | # Check for: Opening Block level tag or |
| 2005 | # Opening Context Block tag (like ins and del) |
| 2006 | # used as a block tag (tag is alone on it's line). |
| 2007 | # |
| 2008 | else if (preg_match('{^<(?:'.$this->block_tags_re.')\b}', $tag) || |
| 2009 | ( preg_match('{^<(?:'.$this->context_block_tags_re.')\b}', $tag) && |
| 2010 | preg_match($newline_before_re, $parsed) && |
| 2011 | preg_match($newline_after_re, $text) ) |
| 2012 | ) |
| 2013 | { |
| 2014 | # Need to parse tag and following text using the HTML parser. |
| 2015 | list($block_text, $text) = |
| 2016 | $this->_hashHTMLBlocks_inHTML($tag . $text, "hashBlock", true); |
| 2017 | |
| 2018 | # Make sure it stays outside of any paragraph by adding newlines. |
| 2019 | $parsed .= "\n\n$block_text\n\n"; |
| 2020 | } |
| 2021 | # |
| 2022 | # Check for: Clean tag (like script, math) |
| 2023 | # HTML Comments, processing instructions. |
| 2024 | # |
| 2025 | else if (preg_match('{^<(?:'.$this->clean_tags_re.')\b}', $tag) || |
| 2026 | $tag[1] == '!' || $tag[1] == '?') |
| 2027 | { |
| 2028 | # Need to parse tag and following text using the HTML parser. |
| 2029 | # (don't check for markdown attribute) |
| 2030 | list($block_text, $text) = |
| 2031 | $this->_hashHTMLBlocks_inHTML($tag . $text, "hashClean", false); |
| 2032 | |
| 2033 | $parsed .= $block_text; |
| 2034 | } |
| 2035 | # |
| 2036 | # Check for: Tag with same name as enclosing tag. |
| 2037 | # |
| 2038 | else if ($enclosing_tag_re !== '' && |
| 2039 | # Same name as enclosing tag. |
| 2040 | preg_match('{^</?(?:'.$enclosing_tag_re.')\b}', $tag)) |
| 2041 | { |
| 2042 | # |
| 2043 | # Increase/decrease nested tag count. |
| 2044 | # |
| 2045 | if ($tag[1] == '/') $depth--; |
| 2046 | else if ($tag[strlen($tag)-2] != '/') $depth++; |
| 2047 | |
| 2048 | if ($depth < 0) { |
| 2049 | # |
| 2050 | # Going out of parent element. Clean up and break so we |
| 2051 | # return to the calling function. |
| 2052 | # |
| 2053 | $text = $tag . $text; |
| 2054 | break; |
| 2055 | } |
| 2056 | |
| 2057 | $parsed .= $tag; |
| 2058 | } |
| 2059 | else { |
| 2060 | $parsed .= $tag; |
| 2061 | } |
| 2062 | } while ($depth >= 0); |
| 2063 | |
| 2064 | return array($parsed, $text); |
| 2065 | } |
| 2066 | function _hashHTMLBlocks_inHTML($text, $hash_method, $md_attr) { |
| 2067 | # |
| 2068 | # Parse HTML, calling _HashHTMLBlocks_InMarkdown for block tags. |
| 2069 | # |
| 2070 | # * Calls $hash_method to convert any blocks. |
| 2071 | # * Stops when the first opening tag closes. |
| 2072 | # * $md_attr indicate if the use of the `markdown="1"` attribute is allowed. |
| 2073 | # (it is not inside clean tags) |
| 2074 | # |
| 2075 | # Returns an array of that form: ( processed text , remaining text ) |
| 2076 | # |
| 2077 | if ($text === '') return array('', ''); |
| 2078 | |
| 2079 | # Regex to match `markdown` attribute inside of a tag. |
| 2080 | $markdown_attr_re = ' |
| 2081 | { |
| 2082 | \s* # Eat whitespace before the `markdown` attribute |
| 2083 | markdown |
| 2084 | \s*=\s* |
| 2085 | (?> |
| 2086 | (["\']) # $1: quote delimiter |
| 2087 | (.*?) # $2: attribute value |
| 2088 | \1 # matching delimiter |
| 2089 | | |
| 2090 | ([^\s>]*) # $3: unquoted attribute value |
| 2091 | ) |
| 2092 | () # $4: make $3 always defined (avoid warnings) |
| 2093 | }xs'; |
| 2094 | |
| 2095 | # Regex to match any tag. |
| 2096 | $tag_re = '{ |
| 2097 | ( # $2: Capture whole tag. |
| 2098 | </? # Any opening or closing tag. |
| 2099 | [\w:$]+ # Tag name. |
| 2100 | (?: |
| 2101 | (?=[\s"\'/a-zA-Z0-9]) # Allowed characters after tag name. |
| 2102 | (?> |
| 2103 | ".*?" | # Double quotes (can contain `>`) |
| 2104 | \'.*?\' | # Single quotes (can contain `>`) |
| 2105 | .+? # Anything but quotes and `>`. |
| 2106 | )*? |
| 2107 | )? |
| 2108 | > # End of tag. |
| 2109 | | |
| 2110 | <!-- .*? --> # HTML Comment |
| 2111 | | |
| 2112 | <\?.*?\?> | <%.*?%> # Processing instruction |
| 2113 | | |
| 2114 | <!\[CDATA\[.*?\]\]> # CData Block |
| 2115 | ) |
| 2116 | }xs'; |
| 2117 | |
| 2118 | $original_text = $text; # Save original text in case of faliure. |
| 2119 | |
| 2120 | $depth = 0; # Current depth inside the tag tree. |
| 2121 | $block_text = ""; # Temporary text holder for current text. |
| 2122 | $parsed = ""; # Parsed text that will be returned. |
| 2123 | |
| 2124 | # |
| 2125 | # Get the name of the starting tag. |
| 2126 | # (This pattern makes $base_tag_name_re safe without quoting.) |
| 2127 | # |
| 2128 | if (preg_match('/^<([\w:$]*)\b/', $text, $matches)) |
| 2129 | $base_tag_name_re = $matches[1]; |
| 2130 | |
| 2131 | # |
| 2132 | # Loop through every tag until we find the corresponding closing tag. |
| 2133 | # |
| 2134 | do { |
| 2135 | # |
| 2136 | # Split the text using the first $tag_match pattern found. |
| 2137 | # Text before pattern will be first in the array, text after |
| 2138 | # pattern will be at the end, and between will be any catches made |
| 2139 | # by the pattern. |
| 2140 | # |
| 2141 | $parts = preg_split($tag_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE); |
| 2142 | |
| 2143 | if (count($parts) < 3) { |
| 2144 | # |
| 2145 | # End of $text reached with unbalenced tag(s). |
| 2146 | # In that case, we return original text unchanged and pass the |
| 2147 | # first character as filtered to prevent an infinite loop in the |
| 2148 | # parent function. |
| 2149 | # |
| 2150 | return array($original_text[0], substr($original_text, 1)); |
| 2151 | } |
| 2152 | |
| 2153 | $block_text .= $parts[0]; # Text before current tag. |
| 2154 | $tag = $parts[1]; # Tag to handle. |
| 2155 | $text = $parts[2]; # Remaining text after current tag. |
| 2156 | |
| 2157 | # |
| 2158 | # Check for: Auto-close tag (like <hr/>) |
| 2159 | # Comments and Processing Instructions. |
| 2160 | # |
| 2161 | if (preg_match('{^</?(?:'.$this->auto_close_tags_re.')\b}', $tag) || |
| 2162 | $tag[1] == '!' || $tag[1] == '?') |
| 2163 | { |
| 2164 | # Just add the tag to the block as if it was text. |
| 2165 | $block_text .= $tag; |
| 2166 | } |
| 2167 | else { |
| 2168 | # |
| 2169 | # Increase/decrease nested tag count. Only do so if |
| 2170 | # the tag's name match base tag's. |
| 2171 | # |
| 2172 | if (preg_match('{^</?'.$base_tag_name_re.'\b}', $tag)) { |
| 2173 | if ($tag[1] == '/') $depth--; |
| 2174 | else if ($tag[strlen($tag)-2] != '/') $depth++; |
| 2175 | } |
| 2176 | |
| 2177 | # |
| 2178 | # Check for `markdown="1"` attribute and handle it. |
| 2179 | # |
| 2180 | if ($md_attr && |
| 2181 | preg_match($markdown_attr_re, $tag, $attr_m) && |
| 2182 | preg_match('/^1|block|span$/', $attr_m[2] . $attr_m[3])) |
| 2183 | { |
| 2184 | # Remove `markdown` attribute from opening tag. |
| 2185 | $tag = preg_replace($markdown_attr_re, '', $tag); |
| 2186 | |
| 2187 | # Check if text inside this tag must be parsed in span mode. |
| 2188 | $this->mode = $attr_m[2] . $attr_m[3]; |
| 2189 | $span_mode = $this->mode == 'span' || $this->mode != 'block' && |
| 2190 | preg_match('{^<(?:'.$this->contain_span_tags_re.')\b}', $tag); |
| 2191 | |
| 2192 | # Calculate indent before tag. |
| 2193 | if (preg_match('/(?:^|\n)( *?)(?! ).*?$/', $block_text, $matches)) { |
| 2194 | $strlen = $this->utf8_strlen; |
| 2195 | $indent = $strlen($matches[1], 'UTF-8'); |
| 2196 | } else { |
| 2197 | $indent = 0; |
| 2198 | } |
| 2199 | |
| 2200 | # End preceding block with this tag. |
| 2201 | $block_text .= $tag; |
| 2202 | $parsed .= $this->$hash_method($block_text); |
| 2203 | |
| 2204 | # Get enclosing tag name for the ParseMarkdown function. |
| 2205 | # (This pattern makes $tag_name_re safe without quoting.) |
| 2206 | preg_match('/^<([\w:$]*)\b/', $tag, $matches); |
| 2207 | $tag_name_re = $matches[1]; |
| 2208 | |
| 2209 | # Parse the content using the HTML-in-Markdown parser. |
| 2210 | list ($block_text, $text) |
| 2211 | = $this->_hashHTMLBlocks_inMarkdown($text, $indent, |
| 2212 | $tag_name_re, $span_mode); |
| 2213 | |
| 2214 | # Outdent markdown text. |
| 2215 | if ($indent > 0) { |
| 2216 | $block_text = preg_replace("/^[ ]{1,$indent}/m", "", |
| 2217 | $block_text); |
| 2218 | } |
| 2219 | |
| 2220 | # Append tag content to parsed text. |
| 2221 | if (!$span_mode) $parsed .= "\n\n$block_text\n\n"; |
| 2222 | else $parsed .= "$block_text"; |
| 2223 | |
| 2224 | # Start over with a new block. |
| 2225 | $block_text = ""; |
| 2226 | } |
| 2227 | else $block_text .= $tag; |
| 2228 | } |
| 2229 | |
| 2230 | } while ($depth > 0); |
| 2231 | |
| 2232 | # |
| 2233 | # Hash last block text that wasn't processed inside the loop. |
| 2234 | # |
| 2235 | $parsed .= $this->$hash_method($block_text); |
| 2236 | |
| 2237 | return array($parsed, $text); |
| 2238 | } |
| 2239 | |
| 2240 | |
| 2241 | function hashClean($text) { |
| 2242 | # |
| 2243 | # Called whenever a tag must be hashed when a function inserts a "clean" tag |
| 2244 | # in $text, it passes through this function and is automaticaly escaped, |
| 2245 | # blocking invalid nested overlap. |
| 2246 | # |
| 2247 | return $this->hashPart($text, 'C'); |
| 2248 | } |
| 2249 | |
| 2250 | |
| 2251 | function doAnchors($text) { |
| 2252 | # |
| 2253 | # Turn Markdown link shortcuts into XHTML <a> tags. |
| 2254 | # |
| 2255 | if ($this->in_anchor) return $text; |
| 2256 | $this->in_anchor = true; |
| 2257 | |
| 2258 | # |
| 2259 | # First, handle reference-style links: [link text] [id] |
| 2260 | # |
| 2261 | $text = preg_replace_callback('{ |
| 2262 | ( # wrap whole match in $1 |
| 2263 | \[ |
| 2264 | ('.$this->nested_brackets_re.') # link text = $2 |
| 2265 | \] |
| 2266 | |
| 2267 | [ ]? # one optional space |
| 2268 | (?:\n[ ]*)? # one optional newline followed by spaces |
| 2269 | |
| 2270 | \[ |
| 2271 | (.*?) # id = $3 |
| 2272 | \] |
| 2273 | ) |
| 2274 | }xs', |
| 2275 | array(&$this, '_doAnchors_reference_callback'), $text); |
| 2276 | |
| 2277 | # |
| 2278 | # Next, inline-style links: [link text](url "optional title") |
| 2279 | # |
| 2280 | $text = preg_replace_callback('{ |
| 2281 | ( # wrap whole match in $1 |
| 2282 | \[ |
| 2283 | ('.$this->nested_brackets_re.') # link text = $2 |
| 2284 | \] |
| 2285 | \( # literal paren |
| 2286 | [ \n]* |
| 2287 | (?: |
| 2288 | <(.+?)> # href = $3 |
| 2289 | | |
| 2290 | ('.$this->nested_url_parenthesis_re.') # href = $4 |
| 2291 | ) |
| 2292 | [ \n]* |
| 2293 | ( # $5 |
| 2294 | ([\'"]) # quote char = $6 |
| 2295 | (.*?) # Title = $7 |
| 2296 | \6 # matching quote |
| 2297 | [ \n]* # ignore any spaces/tabs between closing quote and ) |
| 2298 | )? # title is optional |
| 2299 | \) |
| 2300 | (?:[ ]? '.$this->id_class_attr_catch_re.' )? # $8 = id/class attributes |
| 2301 | ) |
| 2302 | }xs', |
| 2303 | array(&$this, '_doAnchors_inline_callback'), $text); |
| 2304 | |
| 2305 | # |
| 2306 | # Last, handle reference-style shortcuts: [link text] |
| 2307 | # These must come last in case you've also got [link text][1] |
| 2308 | # or [link text](/foo) |
| 2309 | # |
| 2310 | $text = preg_replace_callback('{ |
| 2311 | ( # wrap whole match in $1 |
| 2312 | \[ |
| 2313 | ([^\[\]]+) # link text = $2; can\'t contain [ or ] |
| 2314 | \] |
| 2315 | ) |
| 2316 | }xs', |
| 2317 | array(&$this, '_doAnchors_reference_callback'), $text); |
| 2318 | |
| 2319 | $this->in_anchor = false; |
| 2320 | return $text; |
| 2321 | } |
| 2322 | function _doAnchors_reference_callback($matches) { |
| 2323 | $whole_match = $matches[1]; |
| 2324 | $link_text = $matches[2]; |
| 2325 | $link_id =& $matches[3]; |
| 2326 | |
| 2327 | if ($link_id == "") { |
| 2328 | # for shortcut links like [this][] or [this]. |
| 2329 | $link_id = $link_text; |
| 2330 | } |
| 2331 | |
| 2332 | # lower-case and turn embedded newlines into spaces |
| 2333 | $link_id = strtolower($link_id); |
| 2334 | $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); |
| 2335 | |
| 2336 | if (isset($this->urls[$link_id])) { |
| 2337 | $url = $this->urls[$link_id]; |
| 2338 | $url = $this->encodeAttribute($url); |
| 2339 | |
| 2340 | $result = "<a href=\"$url\""; |
| 2341 | if ( isset( $this->titles[$link_id] ) ) { |
| 2342 | $title = $this->titles[$link_id]; |
| 2343 | $title = $this->encodeAttribute($title); |
| 2344 | $result .= " title=\"$title\""; |
| 2345 | } |
| 2346 | if (isset($this->ref_attr[$link_id])) |
| 2347 | $result .= $this->ref_attr[$link_id]; |
| 2348 | |
| 2349 | $link_text = $this->runSpanGamut($link_text); |
| 2350 | $result .= ">$link_text</a>"; |
| 2351 | $result = $this->hashPart($result); |
| 2352 | } |
| 2353 | else { |
| 2354 | $result = $whole_match; |
| 2355 | } |
| 2356 | return $result; |
| 2357 | } |
| 2358 | function _doAnchors_inline_callback($matches) { |
| 2359 | $whole_match = $matches[1]; |
| 2360 | $link_text = $this->runSpanGamut($matches[2]); |
| 2361 | $url = $matches[3] == '' ? $matches[4] : $matches[3]; |
| 2362 | $title =& $matches[7]; |
| 2363 | $attr = $this->doExtraAttributes("a", $dummy =& $matches[8]); |
| 2364 | |
| 2365 | |
| 2366 | $url = $this->encodeAttribute($url); |
| 2367 | |
| 2368 | $result = "<a href=\"$url\""; |
| 2369 | if (isset($title)) { |
| 2370 | $title = $this->encodeAttribute($title); |
| 2371 | $result .= " title=\"$title\""; |
| 2372 | } |
| 2373 | $result .= $attr; |
| 2374 | |
| 2375 | $link_text = $this->runSpanGamut($link_text); |
| 2376 | $result .= ">$link_text</a>"; |
| 2377 | |
| 2378 | return $this->hashPart($result); |
| 2379 | } |
| 2380 | |
| 2381 | |
| 2382 | function doImages($text) { |
| 2383 | # |
| 2384 | # Turn Markdown image shortcuts into <img> tags. |
| 2385 | # |
| 2386 | # |
| 2387 | # First, handle reference-style labeled images: ![alt text][id] |
| 2388 | # |
| 2389 | $text = preg_replace_callback('{ |
| 2390 | ( # wrap whole match in $1 |
| 2391 | !\[ |
| 2392 | ('.$this->nested_brackets_re.') # alt text = $2 |
| 2393 | \] |
| 2394 | |
| 2395 | [ ]? # one optional space |
| 2396 | (?:\n[ ]*)? # one optional newline followed by spaces |
| 2397 | |
| 2398 | \[ |
| 2399 | (.*?) # id = $3 |
| 2400 | \] |
| 2401 | |
| 2402 | ) |
| 2403 | }xs', |
| 2404 | array(&$this, '_doImages_reference_callback'), $text); |
| 2405 | |
| 2406 | # |
| 2407 | # Next, handle inline images:  |
| 2408 | # Don't forget: encode * and _ |
| 2409 | # |
| 2410 | $text = preg_replace_callback('{ |
| 2411 | ( # wrap whole match in $1 |
| 2412 | !\[ |
| 2413 | ('.$this->nested_brackets_re.') # alt text = $2 |
| 2414 | \] |
| 2415 | \s? # One optional whitespace character |
| 2416 | \( # literal paren |
| 2417 | [ \n]* |
| 2418 | (?: |
| 2419 | <(\S*)> # src url = $3 |
| 2420 | | |
| 2421 | ('.$this->nested_url_parenthesis_re.') # src url = $4 |
| 2422 | ) |
| 2423 | [ \n]* |
| 2424 | ( # $5 |
| 2425 | ([\'"]) # quote char = $6 |
| 2426 | (.*?) # title = $7 |
| 2427 | \6 # matching quote |
| 2428 | [ \n]* |
| 2429 | )? # title is optional |
| 2430 | \) |
| 2431 | (?:[ ]? '.$this->id_class_attr_catch_re.' )? # $8 = id/class attributes |
| 2432 | ) |
| 2433 | }xs', |
| 2434 | array(&$this, '_doImages_inline_callback'), $text); |
| 2435 | |
| 2436 | return $text; |
| 2437 | } |
| 2438 | function _doImages_reference_callback($matches) { |
| 2439 | $whole_match = $matches[1]; |
| 2440 | $alt_text = $matches[2]; |
| 2441 | $link_id = strtolower($matches[3]); |
| 2442 | |
| 2443 | if ($link_id == "") { |
| 2444 | $link_id = strtolower($alt_text); # for shortcut links like ![this][]. |
| 2445 | } |
| 2446 | |
| 2447 | $alt_text = $this->encodeAttribute($alt_text); |
| 2448 | if (isset($this->urls[$link_id])) { |
| 2449 | $url = $this->encodeAttribute($this->urls[$link_id]); |
| 2450 | $result = "<img src=\"$url\" alt=\"$alt_text\""; |
| 2451 | if (isset($this->titles[$link_id])) { |
| 2452 | $title = $this->titles[$link_id]; |
| 2453 | $title = $this->encodeAttribute($title); |
| 2454 | $result .= " title=\"$title\""; |
| 2455 | } |
| 2456 | if (isset($this->ref_attr[$link_id])) |
| 2457 | $result .= $this->ref_attr[$link_id]; |
| 2458 | $result .= $this->empty_element_suffix; |
| 2459 | $result = $this->hashPart($result); |
| 2460 | } |
| 2461 | else { |
| 2462 | # If there's no such link ID, leave intact: |
| 2463 | $result = $whole_match; |
| 2464 | } |
| 2465 | |
| 2466 | return $result; |
| 2467 | } |
| 2468 | function _doImages_inline_callback($matches) { |
| 2469 | $whole_match = $matches[1]; |
| 2470 | $alt_text = $matches[2]; |
| 2471 | $url = $matches[3] == '' ? $matches[4] : $matches[3]; |
| 2472 | $title =& $matches[7]; |
| 2473 | $attr = $this->doExtraAttributes("img", $dummy =& $matches[8]); |
| 2474 | |
| 2475 | $alt_text = $this->encodeAttribute($alt_text); |
| 2476 | $url = $this->encodeAttribute($url); |
| 2477 | $result = "<img src=\"$url\" alt=\"$alt_text\""; |
| 2478 | if (isset($title)) { |
| 2479 | $title = $this->encodeAttribute($title); |
| 2480 | $result .= " title=\"$title\""; # $title already quoted |
| 2481 | } |
| 2482 | $result .= $attr; |
| 2483 | $result .= $this->empty_element_suffix; |
| 2484 | |
| 2485 | return $this->hashPart($result); |
| 2486 | } |
| 2487 | |
| 2488 | |
| 2489 | function doHeaders($text) { |
| 2490 | # |
| 2491 | # Redefined to add id and class attribute support. |
| 2492 | # |
| 2493 | # Setext-style headers: |
| 2494 | # Header 1 {#header1} |
| 2495 | # ======== |
| 2496 | # |
| 2497 | # Header 2 {#header2 .class1 .class2} |
| 2498 | # -------- |
| 2499 | # |
| 2500 | $text = preg_replace_callback( |
| 2501 | '{ |
| 2502 | (^.+?) # $1: Header text |
| 2503 | (?:[ ]+ '.$this->id_class_attr_catch_re.' )? # $3 = id/class attributes |
| 2504 | [ ]*\n(=+|-+)[ ]*\n+ # $3: Header footer |
| 2505 | }mx', |
| 2506 | array(&$this, '_doHeaders_callback_setext'), $text); |
| 2507 | |
| 2508 | # atx-style headers: |
| 2509 | # # Header 1 {#header1} |
| 2510 | # ## Header 2 {#header2} |
| 2511 | # ## Header 2 with closing hashes ## {#header3.class1.class2} |
| 2512 | # ... |
| 2513 | # ###### Header 6 {.class2} |
| 2514 | # |
| 2515 | $text = preg_replace_callback('{ |
| 2516 | ^(\#{1,6}) # $1 = string of #\'s |
| 2517 | [ ]* |
| 2518 | (.+?) # $2 = Header text |
| 2519 | [ ]* |
| 2520 | \#* # optional closing #\'s (not counted) |
| 2521 | (?:[ ]+ '.$this->id_class_attr_catch_re.' )? # $3 = id/class attributes |
| 2522 | [ ]* |
| 2523 | \n+ |
| 2524 | }xm', |
| 2525 | array(&$this, '_doHeaders_callback_atx'), $text); |
| 2526 | |
| 2527 | return $text; |
| 2528 | } |
| 2529 | function _doHeaders_callback_setext($matches) { |
| 2530 | if ($matches[3] == '-' && preg_match('{^- }', $matches[1])) |
| 2531 | return $matches[0]; |
| 2532 | $level = $matches[3][0] == '=' ? 1 : 2; |
| 2533 | $attr = $this->doExtraAttributes("h$level", $dummy =& $matches[2]); |
| 2534 | $block = "<h$level$attr>".$this->runSpanGamut($matches[1])."</h$level>"; |
| 2535 | return "\n" . $this->hashBlock($block) . "\n\n"; |
| 2536 | } |
| 2537 | function _doHeaders_callback_atx($matches) { |
| 2538 | $level = strlen($matches[1]); |
| 2539 | $attr = $this->doExtraAttributes("h$level", $dummy =& $matches[3]); |
| 2540 | $block = "<h$level$attr>".$this->runSpanGamut($matches[2])."</h$level>"; |
| 2541 | return "\n" . $this->hashBlock($block) . "\n\n"; |
| 2542 | } |
| 2543 | |
| 2544 | |
| 2545 | function doTables($text) { |
| 2546 | # |
| 2547 | # Form HTML tables. |
| 2548 | # |
| 2549 | $less_than_tab = $this->tab_width - 1; |
| 2550 | # |
| 2551 | # Find tables with leading pipe. |
| 2552 | # |
| 2553 | # | Header 1 | Header 2 |
| 2554 | # | -------- | -------- |
| 2555 | # | Cell 1 | Cell 2 |
| 2556 | # | Cell 3 | Cell 4 |
| 2557 | # |
| 2558 | $text = preg_replace_callback(' |
| 2559 | { |
| 2560 | ^ # Start of a line |
| 2561 | [ ]{0,'.$less_than_tab.'} # Allowed whitespace. |
| 2562 | [|] # Optional leading pipe (present) |
| 2563 | (.+) \n # $1: Header row (at least one pipe) |
| 2564 | |
| 2565 | [ ]{0,'.$less_than_tab.'} # Allowed whitespace. |
| 2566 | [|] ([ ]*[-:]+[-| :]*) \n # $2: Header underline |
| 2567 | |
| 2568 | ( # $3: Cells |
| 2569 | (?> |
| 2570 | [ ]* # Allowed whitespace. |
| 2571 | [|] .* \n # Row content. |
| 2572 | )* |
| 2573 | ) |
| 2574 | (?=\n|\Z) # Stop at final double newline. |
| 2575 | }xm', |
| 2576 | array(&$this, '_doTable_leadingPipe_callback'), $text); |
| 2577 | |
| 2578 | # |
| 2579 | # Find tables without leading pipe. |
| 2580 | # |
| 2581 | # Header 1 | Header 2 |
| 2582 | # -------- | -------- |
| 2583 | # Cell 1 | Cell 2 |
| 2584 | # Cell 3 | Cell 4 |
| 2585 | # |
| 2586 | $text = preg_replace_callback(' |
| 2587 | { |
| 2588 | ^ # Start of a line |
| 2589 | [ ]{0,'.$less_than_tab.'} # Allowed whitespace. |
| 2590 | (\S.*[|].*) \n # $1: Header row (at least one pipe) |
| 2591 | |
| 2592 | [ ]{0,'.$less_than_tab.'} # Allowed whitespace. |
| 2593 | ([-:]+[ ]*[|][-| :]*) \n # $2: Header underline |
| 2594 | |
| 2595 | ( # $3: Cells |
| 2596 | (?> |
| 2597 | .* [|] .* \n # Row content |
| 2598 | )* |
| 2599 | ) |
| 2600 | (?=\n|\Z) # Stop at final double newline. |
| 2601 | }xm', |
| 2602 | array(&$this, '_DoTable_callback'), $text); |
| 2603 | |
| 2604 | return $text; |
| 2605 | } |
| 2606 | function _doTable_leadingPipe_callback($matches) { |
| 2607 | $head = $matches[1]; |
| 2608 | $underline = $matches[2]; |
| 2609 | $content = $matches[3]; |
| 2610 | |
| 2611 | # Remove leading pipe for each row. |
| 2612 | $content = preg_replace('/^ *[|]/m', '', $content); |
| 2613 | |
| 2614 | return $this->_doTable_callback(array($matches[0], $head, $underline, $content)); |
| 2615 | } |
| 2616 | function _doTable_callback($matches) { |
| 2617 | $head = $matches[1]; |
| 2618 | $underline = $matches[2]; |
| 2619 | $content = $matches[3]; |
| 2620 | |
| 2621 | # Remove any tailing pipes for each line. |
| 2622 | $head = preg_replace('/[|] *$/m', '', $head); |
| 2623 | $underline = preg_replace('/[|] *$/m', '', $underline); |
| 2624 | $content = preg_replace('/[|] *$/m', '', $content); |
| 2625 | |
| 2626 | # Reading alignement from header underline. |
| 2627 | $separators = preg_split('/ *[|] */', $underline); |
| 2628 | $attr = array(); |
| 2629 | foreach ($separators as $n => $s) { |
| 2630 | if (preg_match('/^ *-+: *$/', $s)) $attr[$n] = ' align="right"'; |
| 2631 | else if (preg_match('/^ *:-+: *$/', $s))$attr[$n] = ' align="center"'; |
| 2632 | else if (preg_match('/^ *:-+ *$/', $s)) $attr[$n] = ' align="left"'; |
| 2633 | else $attr[$n] = ''; |
| 2634 | } |
| 2635 | |
| 2636 | # Parsing span elements, including code spans, character escapes, |
| 2637 | # and inline HTML tags, so that pipes inside those gets ignored. |
| 2638 | $head = $this->parseSpan($head); |
| 2639 | $headers = preg_split('/ *[|] */', $head); |
| 2640 | $col_count = count($headers); |
| 2641 | $attr = array_pad($attr, $col_count, ''); |
| 2642 | |
| 2643 | # Write column headers. |
| 2644 | $text = "<table>\n"; |
| 2645 | $text .= "<thead>\n"; |
| 2646 | $text .= "<tr>\n"; |
| 2647 | foreach ($headers as $n => $header) |
| 2648 | $text .= " <th$attr[$n]>".$this->runSpanGamut(trim($header))."</th>\n"; |
| 2649 | $text .= "</tr>\n"; |
| 2650 | $text .= "</thead>\n"; |
| 2651 | |
| 2652 | # Split content by row. |
| 2653 | $rows = explode("\n", trim($content, "\n")); |
| 2654 | |
| 2655 | $text .= "<tbody>\n"; |
| 2656 | foreach ($rows as $row) { |
| 2657 | # Parsing span elements, including code spans, character escapes, |
| 2658 | # and inline HTML tags, so that pipes inside those gets ignored. |
| 2659 | $row = $this->parseSpan($row); |
| 2660 | |
| 2661 | # Split row by cell. |
| 2662 | $row_cells = preg_split('/ *[|] */', $row, $col_count); |
| 2663 | $row_cells = array_pad($row_cells, $col_count, ''); |
| 2664 | |
| 2665 | $text .= "<tr>\n"; |
| 2666 | foreach ($row_cells as $n => $cell) |
| 2667 | $text .= " <td$attr[$n]>".$this->runSpanGamut(trim($cell))."</td>\n"; |
| 2668 | $text .= "</tr>\n"; |
| 2669 | } |
| 2670 | $text .= "</tbody>\n"; |
| 2671 | $text .= "</table>"; |
| 2672 | |
| 2673 | return $this->hashBlock($text) . "\n"; |
| 2674 | } |
| 2675 | |
| 2676 | |
| 2677 | function doDefLists($text) { |
| 2678 | # |
| 2679 | # Form HTML definition lists. |
| 2680 | # |
| 2681 | $less_than_tab = $this->tab_width - 1; |
| 2682 | |
| 2683 | # Re-usable pattern to match any entire dl list: |
| 2684 | $whole_list_re = '(?> |
| 2685 | ( # $1 = whole list |
| 2686 | ( # $2 |
| 2687 | [ ]{0,'.$less_than_tab.'} |
| 2688 | ((?>.*\S.*\n)+) # $3 = defined term |
| 2689 | \n? |
| 2690 | [ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition |
| 2691 | ) |
| 2692 | (?s:.+?) |
| 2693 | ( # $4 |
| 2694 | \z |
| 2695 | | |
| 2696 | \n{2,} |
| 2697 | (?=\S) |
| 2698 | (?! # Negative lookahead for another term |
| 2699 | [ ]{0,'.$less_than_tab.'} |
| 2700 | (?: \S.*\n )+? # defined term |
| 2701 | \n? |
| 2702 | [ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition |
| 2703 | ) |
| 2704 | (?! # Negative lookahead for another definition |
| 2705 | [ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition |
| 2706 | ) |
| 2707 | ) |
| 2708 | ) |
| 2709 | )'; // mx |
| 2710 | |
| 2711 | $text = preg_replace_callback('{ |
| 2712 | (?>\A\n?|(?<=\n\n)) |
| 2713 | '.$whole_list_re.' |
| 2714 | }mx', |
| 2715 | array(&$this, '_doDefLists_callback'), $text); |
| 2716 | |
| 2717 | return $text; |
| 2718 | } |
| 2719 | function _doDefLists_callback($matches) { |
| 2720 | # Re-usable patterns to match list item bullets and number markers: |
| 2721 | $list = $matches[1]; |
| 2722 | |
| 2723 | # Turn double returns into triple returns, so that we can make a |
| 2724 | # paragraph for the last item in a list, if necessary: |
| 2725 | $result = trim($this->processDefListItems($list)); |
| 2726 | $result = "<dl>\n" . $result . "\n</dl>"; |
| 2727 | return $this->hashBlock($result) . "\n\n"; |
| 2728 | } |
| 2729 | |
| 2730 | |
| 2731 | function processDefListItems($list_str) { |
| 2732 | # |
| 2733 | # Process the contents of a single definition list, splitting it |
| 2734 | # into individual term and definition list items. |
| 2735 | # |
| 2736 | $less_than_tab = $this->tab_width - 1; |
| 2737 | |
| 2738 | # trim trailing blank lines: |
| 2739 | $list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str); |
| 2740 | |
| 2741 | # Process definition terms. |
| 2742 | $list_str = preg_replace_callback('{ |
| 2743 | (?>\A\n?|\n\n+) # leading line |
| 2744 | ( # definition terms = $1 |
| 2745 | [ ]{0,'.$less_than_tab.'} # leading whitespace |
| 2746 | (?!\:[ ]|[ ]) # negative lookahead for a definition |
| 2747 | # mark (colon) or more whitespace. |
| 2748 | (?> \S.* \n)+? # actual term (not whitespace). |
| 2749 | ) |
| 2750 | (?=\n?[ ]{0,3}:[ ]) # lookahead for following line feed |
| 2751 | # with a definition mark. |
| 2752 | }xm', |
| 2753 | array(&$this, '_processDefListItems_callback_dt'), $list_str); |
| 2754 | |
| 2755 | # Process actual definitions. |
| 2756 | $list_str = preg_replace_callback('{ |
| 2757 | \n(\n+)? # leading line = $1 |
| 2758 | ( # marker space = $2 |
| 2759 | [ ]{0,'.$less_than_tab.'} # whitespace before colon |
| 2760 | \:[ ]+ # definition mark (colon) |
| 2761 | ) |
| 2762 | ((?s:.+?)) # definition text = $3 |
| 2763 | (?= \n+ # stop at next definition mark, |
| 2764 | (?: # next term or end of text |
| 2765 | [ ]{0,'.$less_than_tab.'} \:[ ] | |
| 2766 | <dt> | \z |
| 2767 | ) |
| 2768 | ) |
| 2769 | }xm', |
| 2770 | array(&$this, '_processDefListItems_callback_dd'), $list_str); |
| 2771 | |
| 2772 | return $list_str; |
| 2773 | } |
| 2774 | function _processDefListItems_callback_dt($matches) { |
| 2775 | $terms = explode("\n", trim($matches[1])); |
| 2776 | $text = ''; |
| 2777 | foreach ($terms as $term) { |
| 2778 | $term = $this->runSpanGamut(trim($term)); |
| 2779 | $text .= "\n<dt>" . $term . "</dt>"; |
| 2780 | } |
| 2781 | return $text . "\n"; |
| 2782 | } |
| 2783 | function _processDefListItems_callback_dd($matches) { |
| 2784 | $leading_line = $matches[1]; |
| 2785 | $marker_space = $matches[2]; |
| 2786 | $def = $matches[3]; |
| 2787 | |
| 2788 | if ($leading_line || preg_match('/\n{2,}/', $def)) { |
| 2789 | # Replace marker with the appropriate whitespace indentation |
| 2790 | $def = str_repeat(' ', strlen($marker_space)) . $def; |
| 2791 | $def = $this->runBlockGamut($this->outdent($def . "\n\n")); |
| 2792 | $def = "\n". $def ."\n"; |
| 2793 | } |
| 2794 | else { |
| 2795 | $def = rtrim($def); |
| 2796 | $def = $this->runSpanGamut($this->outdent($def)); |
| 2797 | } |
| 2798 | |
| 2799 | return "\n<dd>" . $def . "</dd>\n"; |
| 2800 | } |
| 2801 | |
| 2802 | |
| 2803 | function doFencedCodeBlocks($text) { |
| 2804 | # |
| 2805 | # Adding the fenced code block syntax to regular Markdown: |
| 2806 | # |
| 2807 | # ~~~ |
| 2808 | # Code block |
| 2809 | # ~~~ |
| 2810 | # |
| 2811 | $less_than_tab = $this->tab_width; |
| 2812 | |
| 2813 | $text = preg_replace_callback('{ |
| 2814 | (?:\n|\A) |
| 2815 | # 1: Opening marker |
| 2816 | ( |
| 2817 | (?:~{3,}|`{3,}) # 3 or more tildes/backticks. |
| 2818 | ) |
| 2819 | [ ]* |
| 2820 | (?: |
| 2821 | \.?([-_:a-zA-Z0-9]+) # 2: standalone class name |
| 2822 | | |
| 2823 | '.$this->id_class_attr_catch_re.' # 3: Extra attributes |
| 2824 | )? |
| 2825 | [ ]* \n # Whitespace and newline following marker. |
| 2826 | |
| 2827 | # 4: Content |
| 2828 | ( |
| 2829 | (?> |
| 2830 | (?!\1 [ ]* \n) # Not a closing marker. |
| 2831 | .*\n+ |
| 2832 | )+ |
| 2833 | ) |
| 2834 | |
| 2835 | # Closing marker. |
| 2836 | \1 [ ]* (?= \n ) |
| 2837 | }xm', |
| 2838 | array(&$this, '_doFencedCodeBlocks_callback'), $text); |
| 2839 | |
| 2840 | return $text; |
| 2841 | } |
| 2842 | function _doFencedCodeBlocks_callback($matches) { |
| 2843 | $classname =& $matches[2]; |
| 2844 | $attrs =& $matches[3]; |
| 2845 | $codeblock = $matches[4]; |
| 2846 | $codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES); |
| 2847 | $codeblock = preg_replace_callback('/^\n+/', |
| 2848 | array(&$this, '_doFencedCodeBlocks_newlines'), $codeblock); |
| 2849 | |
| 2850 | if ($classname != "") { |
| 2851 | if ($classname[0] == '.') |
| 2852 | $classname = substr($classname, 1); |
| 2853 | $attr_str = ' class="'.$this->code_class_prefix.$classname.'"'; |
| 2854 | } else { |
| 2855 | $attr_str = $this->doExtraAttributes($this->code_attr_on_pre ? "pre" : "code", $attrs); |
| 2856 | } |
| 2857 | $pre_attr_str = $this->code_attr_on_pre ? $attr_str : ''; |
| 2858 | $code_attr_str = $this->code_attr_on_pre ? '' : $attr_str; |
| 2859 | $codeblock = "<pre$pre_attr_str><code$code_attr_str>$codeblock</code></pre>"; |
| 2860 | |
| 2861 | return "\n\n".$this->hashBlock($codeblock)."\n\n"; |
| 2862 | } |
| 2863 | function _doFencedCodeBlocks_newlines($matches) { |
| 2864 | return str_repeat("<br$this->empty_element_suffix", |
| 2865 | strlen($matches[0])); |
| 2866 | } |
| 2867 | |
| 2868 | |
| 2869 | # |
| 2870 | # Redefining emphasis markers so that emphasis by underscore does not |
| 2871 | # work in the middle of a word. |
| 2872 | # |
| 2873 | public $em_relist = array( |
| 2874 | '' => '(?:(?<!\*)\*(?!\*)|(?<![a-zA-Z0-9_])_(?!_))(?=\S|$)(?![\.,:;]\s)', |
| 2875 | '*' => '(?<=\S|^)(?<!\*)\*(?!\*)', |
| 2876 | '_' => '(?<=\S|^)(?<!_)_(?![a-zA-Z0-9_])', |
| 2877 | ); |
| 2878 | public $strong_relist = array( |
| 2879 | '' => '(?:(?<!\*)\*\*(?!\*)|(?<![a-zA-Z0-9_])__(?!_))(?=\S|$)(?![\.,:;]\s)', |
| 2880 | '**' => '(?<=\S|^)(?<!\*)\*\*(?!\*)', |
| 2881 | '__' => '(?<=\S|^)(?<!_)__(?![a-zA-Z0-9_])', |
| 2882 | ); |
| 2883 | public $em_strong_relist = array( |
| 2884 | '' => '(?:(?<!\*)\*\*\*(?!\*)|(?<![a-zA-Z0-9_])___(?!_))(?=\S|$)(?![\.,:;]\s)', |
| 2885 | '***' => '(?<=\S|^)(?<!\*)\*\*\*(?!\*)', |
| 2886 | '___' => '(?<=\S|^)(?<!_)___(?![a-zA-Z0-9_])', |
| 2887 | ); |
| 2888 | |
| 2889 | |
| 2890 | function formParagraphs($text) { |
| 2891 | # |
| 2892 | # Params: |
| 2893 | # $text - string to process with html <p> tags |
| 2894 | # |
| 2895 | # Strip leading and trailing lines: |
| 2896 | $text = preg_replace('/\A\n+|\n+\z/', '', $text); |
| 2897 | |
| 2898 | $grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY); |
| 2899 | |
| 2900 | # |
| 2901 | # Wrap <p> tags and unhashify HTML blocks |
| 2902 | # |
| 2903 | foreach ($grafs as $key => $value) { |
| 2904 | $value = trim($this->runSpanGamut($value)); |
| 2905 | |
| 2906 | # Check if this should be enclosed in a paragraph. |
| 2907 | # Clean tag hashes & block tag hashes are left alone. |
| 2908 | $is_p = !preg_match('/^B\x1A[0-9]+B|^C\x1A[0-9]+C$/', $value); |
| 2909 | |
| 2910 | if ($is_p) { |
| 2911 | $value = "<p>$value</p>"; |
| 2912 | } |
| 2913 | $grafs[$key] = $value; |
| 2914 | } |
| 2915 | |
| 2916 | # Join grafs in one text, then unhash HTML tags. |
| 2917 | $text = implode("\n\n", $grafs); |
| 2918 | |
| 2919 | # Finish by removing any tag hashes still present in $text. |
| 2920 | $text = $this->unhash($text); |
| 2921 | |
| 2922 | return $text; |
| 2923 | } |
| 2924 | |
| 2925 | |
| 2926 | ### Footnotes |
| 2927 | |
| 2928 | function stripFootnotes($text) { |
| 2929 | # |
| 2930 | # Strips link definitions from text, stores the URLs and titles in |
| 2931 | # hash references. |
| 2932 | # |
| 2933 | $less_than_tab = $this->tab_width - 1; |
| 2934 | |
| 2935 | # Link defs are in the form: [^id]: url "optional title" |
| 2936 | $text = preg_replace_callback('{ |
| 2937 | ^[ ]{0,'.$less_than_tab.'}\[\^(.+?)\][ ]?: # note_id = $1 |
| 2938 | [ ]* |
| 2939 | \n? # maybe *one* newline |
| 2940 | ( # text = $2 (no blank lines allowed) |
| 2941 | (?: |
| 2942 | .+ # actual text |
| 2943 | | |
| 2944 | \n # newlines but |
| 2945 | (?!\[\^.+?\]:\s)# negative lookahead for footnote marker. |
| 2946 | (?!\n+[ ]{0,3}\S)# ensure line is not blank and followed |
| 2947 | # by non-indented content |
| 2948 | )* |
| 2949 | ) |
| 2950 | }xm', |
| 2951 | array(&$this, '_stripFootnotes_callback'), |
| 2952 | $text); |
| 2953 | return $text; |
| 2954 | } |
| 2955 | function _stripFootnotes_callback($matches) { |
| 2956 | $note_id = $this->fn_id_prefix . $matches[1]; |
| 2957 | $this->footnotes[$note_id] = $this->outdent($matches[2]); |
| 2958 | return ''; # String that will replace the block |
| 2959 | } |
| 2960 | |
| 2961 | |
| 2962 | function doFootnotes($text) { |
| 2963 | # |
| 2964 | # Replace footnote references in $text [^id] with a special text-token |
| 2965 | # which will be replaced by the actual footnote marker in appendFootnotes. |
| 2966 | # |
| 2967 | if (!$this->in_anchor) { |
| 2968 | $text = preg_replace('{\[\^(.+?)\]}', "F\x1Afn:\\1\x1A:", $text); |
| 2969 | } |
| 2970 | return $text; |
| 2971 | } |
| 2972 | |
| 2973 | |
| 2974 | function appendFootnotes($text) { |
| 2975 | # |
| 2976 | # Append footnote list to text. |
| 2977 | # |
| 2978 | $text = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', |
| 2979 | array(&$this, '_appendFootnotes_callback'), $text); |
| 2980 | |
| 2981 | if (!empty($this->footnotes_ordered)) { |
| 2982 | $text .= "\n\n"; |
| 2983 | $text .= "<div class=\"footnotes\">\n"; |
| 2984 | $text .= "<hr". $this->empty_element_suffix ."\n"; |
| 2985 | $text .= "<ol>\n\n"; |
| 2986 | |
| 2987 | $attr = ""; |
| 2988 | if ($this->fn_backlink_class != "") { |
| 2989 | $class = $this->fn_backlink_class; |
| 2990 | $class = $this->encodeAttribute($class); |
| 2991 | $attr .= " class=\"$class\""; |
| 2992 | } |
| 2993 | if ($this->fn_backlink_title != "") { |
| 2994 | $title = $this->fn_backlink_title; |
| 2995 | $title = $this->encodeAttribute($title); |
| 2996 | $attr .= " title=\"$title\""; |
| 2997 | } |
| 2998 | $num = 0; |
| 2999 | |
| 3000 | while (!empty($this->footnotes_ordered)) { |
| 3001 | $footnote = reset($this->footnotes_ordered); |
| 3002 | $note_id = key($this->footnotes_ordered); |
| 3003 | unset($this->footnotes_ordered[$note_id]); |
| 3004 | $ref_count = $this->footnotes_ref_count[$note_id]; |
| 3005 | unset($this->footnotes_ref_count[$note_id]); |
| 3006 | unset($this->footnotes[$note_id]); |
| 3007 | |
| 3008 | $footnote .= "\n"; # Need to append newline before parsing. |
| 3009 | $footnote = $this->runBlockGamut("$footnote\n"); |
| 3010 | $footnote = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', |
| 3011 | array(&$this, '_appendFootnotes_callback'), $footnote); |
| 3012 | |
| 3013 | $attr = str_replace("%%", (string) ++$num, $attr); |
| 3014 | $note_id = $this->encodeAttribute($note_id); |
| 3015 | |
| 3016 | # Prepare backlink, multiple backlinks if multiple references |
| 3017 | $backlink = "<a href=\"#fnref:$note_id\"$attr>↩</a>"; |
| 3018 | for ($ref_num = 2; $ref_num <= $ref_count; ++$ref_num) { |
| 3019 | $backlink .= " <a href=\"#fnref$ref_num:$note_id\"$attr>↩</a>"; |
| 3020 | } |
| 3021 | # Add backlink to last paragraph; create new paragraph if needed. |
| 3022 | if (preg_match('{</p>$}', $footnote)) { |
| 3023 | $footnote = substr($footnote, 0, -4) . " $backlink</p>"; |
| 3024 | } else { |
| 3025 | $footnote .= "\n\n<p>$backlink</p>"; |
| 3026 | } |
| 3027 | |
| 3028 | $text .= "<li id=\"fn:$note_id\">\n"; |
| 3029 | $text .= $footnote . "\n"; |
| 3030 | $text .= "</li>\n\n"; |
| 3031 | } |
| 3032 | |
| 3033 | $text .= "</ol>\n"; |
| 3034 | $text .= "</div>"; |
| 3035 | } |
| 3036 | return $text; |
| 3037 | } |
| 3038 | function _appendFootnotes_callback($matches) { |
| 3039 | $node_id = $this->fn_id_prefix . $matches[1]; |
| 3040 | |
| 3041 | # Create footnote marker only if it has a corresponding footnote *and* |
| 3042 | # the footnote hasn't been used by another marker. |
| 3043 | if (isset($this->footnotes[$node_id])) { |
| 3044 | $num =& $this->footnotes_numbers[$node_id]; |
| 3045 | if (!isset($num)) { |
| 3046 | # Transfer footnote content to the ordered list and give it its |
| 3047 | # number |
| 3048 | $this->footnotes_ordered[$node_id] = $this->footnotes[$node_id]; |
| 3049 | $this->footnotes_ref_count[$node_id] = 1; |
| 3050 | $num = $this->footnote_counter++; |
| 3051 | $ref_count_mark = ''; |
| 3052 | } else { |
| 3053 | $ref_count_mark = $this->footnotes_ref_count[$node_id] += 1; |
| 3054 | } |
| 3055 | |
| 3056 | $attr = ""; |
| 3057 | if ($this->fn_link_class != "") { |
| 3058 | $class = $this->fn_link_class; |
| 3059 | $class = $this->encodeAttribute($class); |
| 3060 | $attr .= " class=\"$class\""; |
| 3061 | } |
| 3062 | if ($this->fn_link_title != "") { |
| 3063 | $title = $this->fn_link_title; |
| 3064 | $title = $this->encodeAttribute($title); |
| 3065 | $attr .= " title=\"$title\""; |
| 3066 | } |
| 3067 | |
| 3068 | $attr = str_replace("%%", $num, $attr); |
| 3069 | $node_id = $this->encodeAttribute($node_id); |
| 3070 | |
| 3071 | return |
| 3072 | "<sup id=\"fnref$ref_count_mark:$node_id\">". |
| 3073 | "<a href=\"#fn:$node_id\"$attr>$num</a>". |
| 3074 | "</sup>"; |
| 3075 | } |
| 3076 | |
| 3077 | return "[^".$matches[1]."]"; |
| 3078 | } |
| 3079 | |
| 3080 | |
| 3081 | ### Abbreviations ### |
| 3082 | |
| 3083 | function stripAbbreviations($text) { |
| 3084 | # |
| 3085 | # Strips abbreviations from text, stores titles in hash references. |
| 3086 | # |
| 3087 | $less_than_tab = $this->tab_width - 1; |
| 3088 | |
| 3089 | # Link defs are in the form: [id]*: url "optional title" |
| 3090 | $text = preg_replace_callback('{ |
| 3091 | ^[ ]{0,'.$less_than_tab.'}\*\[(.+?)\][ ]?: # abbr_id = $1 |
| 3092 | (.*) # text = $2 (no blank lines allowed) |
| 3093 | }xm', |
| 3094 | array(&$this, '_stripAbbreviations_callback'), |
| 3095 | $text); |
| 3096 | return $text; |
| 3097 | } |
| 3098 | function _stripAbbreviations_callback($matches) { |
| 3099 | $abbr_word = $matches[1]; |
| 3100 | $abbr_desc = $matches[2]; |
| 3101 | if ($this->abbr_word_re) |
| 3102 | $this->abbr_word_re .= '|'; |
| 3103 | $this->abbr_word_re .= preg_quote($abbr_word); |
| 3104 | $this->abbr_desciptions[$abbr_word] = trim($abbr_desc); |
| 3105 | return ''; # String that will replace the block |
| 3106 | } |
| 3107 | |
| 3108 | |
| 3109 | function doAbbreviations($text) { |
| 3110 | # |
| 3111 | # Find defined abbreviations in text and wrap them in <abbr> elements. |
| 3112 | # |
| 3113 | if ($this->abbr_word_re) { |
| 3114 | // cannot use the /x modifier because abbr_word_re may |
| 3115 | // contain significant spaces: |
| 3116 | $text = preg_replace_callback('{'. |
| 3117 | '(?<![\w\x1A])'. |
| 3118 | '(?:'.$this->abbr_word_re.')'. |
| 3119 | '(?![\w\x1A])'. |
| 3120 | '}', |
| 3121 | array(&$this, '_doAbbreviations_callback'), $text); |
| 3122 | } |
| 3123 | return $text; |
| 3124 | } |
| 3125 | function _doAbbreviations_callback($matches) { |
| 3126 | $abbr = $matches[0]; |
| 3127 | if (isset($this->abbr_desciptions[$abbr])) { |
| 3128 | $desc = $this->abbr_desciptions[$abbr]; |
| 3129 | if (empty($desc)) { |
| 3130 | return $this->hashPart("<abbr>$abbr</abbr>"); |
| 3131 | } else { |
| 3132 | $desc = $this->encodeAttribute($desc); |
| 3133 | return $this->hashPart("<abbr title=\"$desc\">$abbr</abbr>"); |
| 3134 | } |
| 3135 | } else { |
| 3136 | return $matches[0]; |
| 3137 | } |
| 3138 | } |
| 3139 | |
| 3140 | } |
| 3141 | |
| 3142 | |
| 3143 | /* |
| 3144 | |
| 3145 | PHP Markdown Extra |
| 3146 | ================== |
| 3147 | |
| 3148 | Description |
| 3149 | ----------- |
| 3150 | |
| 3151 | This is a PHP port of the original Markdown formatter written in Perl |
| 3152 | by John Gruber. This special "Extra" version of PHP Markdown features |
| 3153 | further enhancements to the syntax for making additional constructs |
| 3154 | such as tables and definition list. |
| 3155 | |
| 3156 | Markdown is a text-to-HTML filter; it translates an easy-to-read / |
| 3157 | easy-to-write structured text format into HTML. Markdown's text format |
| 3158 | is mostly similar to that of plain text email, and supports features such |
| 3159 | as headers, *emphasis*, code blocks, blockquotes, and links. |
| 3160 | |
| 3161 | Markdown's syntax is designed not as a generic markup language, but |
| 3162 | specifically to serve as a front-end to (X)HTML. You can use span-level |
| 3163 | HTML tags anywhere in a Markdown document, and you can use block level |
| 3164 | HTML tags (like <div> and <table> as well). |
| 3165 | |
| 3166 | For more information about Markdown's syntax, see: |
| 3167 | |
| 3168 | <http://daringfireball.net/projects/markdown/> |
| 3169 | |
| 3170 | |
| 3171 | Bugs |
| 3172 | ---- |
| 3173 | |
| 3174 | To file bug reports please send email to: |
| 3175 | |
| 3176 | <michel.fortin@michelf.ca> |
| 3177 | |
| 3178 | Please include with your report: (1) the example input; (2) the output you |
| 3179 | expected; (3) the output Markdown actually produced. |
| 3180 | |
| 3181 | |
| 3182 | Version History |
| 3183 | --------------- |
| 3184 | |
| 3185 | See the readme file for detailed release notes for this version. |
| 3186 | |
| 3187 | |
| 3188 | Copyright and License |
| 3189 | --------------------- |
| 3190 | |
| 3191 | PHP Markdown & Extra |
| 3192 | Copyright (c) 2004-2013 Michel Fortin |
| 3193 | <http://michelf.ca/> |
| 3194 | All rights reserved. |
| 3195 | |
| 3196 | Based on Markdown |
| 3197 | Copyright (c) 2003-2006 John Gruber |
| 3198 | <http://daringfireball.net/> |
| 3199 | All rights reserved. |
| 3200 | |
| 3201 | Redistribution and use in source and binary forms, with or without |
| 3202 | modification, are permitted provided that the following conditions are |
| 3203 | met: |
| 3204 | |
| 3205 | * Redistributions of source code must retain the above copyright notice, |
| 3206 | this list of conditions and the following disclaimer. |
| 3207 | |
| 3208 | * Redistributions in binary form must reproduce the above copyright |
| 3209 | notice, this list of conditions and the following disclaimer in the |
| 3210 | documentation and/or other materials provided with the distribution. |
| 3211 | |
| 3212 | * Neither the name "Markdown" nor the names of its contributors may |
| 3213 | be used to endorse or promote products derived from this software |
| 3214 | without specific prior written permission. |
| 3215 | |
| 3216 | This software is provided by the copyright holders and contributors "as |
| 3217 | is" and any express or implied warranties, including, but not limited |
| 3218 | to, the implied warranties of merchantability and fitness for a |
| 3219 | particular purpose are disclaimed. In no event shall the copyright owner |
| 3220 | or contributors be liable for any direct, indirect, incidental, special, |
| 3221 | exemplary, or consequential damages (including, but not limited to, |
| 3222 | procurement of substitute goods or services; loss of use, data, or |
| 3223 | profits; or business interruption) however caused and on any theory of |
| 3224 | liability, whether in contract, strict liability, or tort (including |
| 3225 | negligence or otherwise) arising in any way out of the use of this |
| 3226 | software, even if advised of the possibility of such damage. |
| 3227 | |
| 3228 | */ |
| 3229 | ?> |
| 3230 |