| [ Index ] |
PHP Cross Reference of moodle-2.8 |
[Summary view] [Print] [Text view]
1 <?php 2 # 3 # Markdown - A text-to-HTML conversion tool for web writers 4 # 5 # PHP Markdown 6 # Copyright (c) 2004-2014 Michel Fortin 7 # <http://michelf.com/projects/php-markdown/> 8 # 9 # Original Markdown 10 # Copyright (c) 2004-2006 John Gruber 11 # <http://daringfireball.net/projects/markdown/> 12 # 13 namespace Michelf; 14 15 16 # 17 # Markdown Parser Class 18 # 19 20 class Markdown implements MarkdownInterface { 21 22 ### Version ### 23 24 const MARKDOWNLIB_VERSION = "1.4.1"; 25 26 ### Simple Function Interface ### 27 28 public static function defaultTransform($text) { 29 # 30 # Initialize the parser and return the result of its transform method. 31 # This will work fine for derived classes too. 32 # 33 # Take parser class on which this function was called. 34 $parser_class = \get_called_class(); 35 36 # try to take parser from the static parser list 37 static $parser_list; 38 $parser =& $parser_list[$parser_class]; 39 40 # create the parser it not already set 41 if (!$parser) 42 $parser = new $parser_class; 43 44 # Transform text using parser. 45 return $parser->transform($text); 46 } 47 48 ### Configuration Variables ### 49 50 # Change to ">" for HTML output. 51 public $empty_element_suffix = " />"; 52 public $tab_width = 4; 53 54 # Change to `true` to disallow markup or entities. 55 public $no_markup = false; 56 public $no_entities = false; 57 58 # Predefined urls and titles for reference links and images. 59 public $predef_urls = array(); 60 public $predef_titles = array(); 61 62 63 ### Parser Implementation ### 64 65 # Regex to match balanced [brackets]. 66 # Needed to insert a maximum bracked depth while converting to PHP. 67 protected $nested_brackets_depth = 6; 68 protected $nested_brackets_re; 69 70 protected $nested_url_parenthesis_depth = 4; 71 protected $nested_url_parenthesis_re; 72 73 # Table of hash values for escaped characters: 74 protected $escape_chars = '\`*_{}[]()>#+-.!'; 75 protected $escape_chars_re; 76 77 78 public function __construct() { 79 # 80 # Constructor function. Initialize appropriate member variables. 81 # 82 $this->_initDetab(); 83 $this->prepareItalicsAndBold(); 84 85 $this->nested_brackets_re = 86 str_repeat('(?>[^\[\]]+|\[', $this->nested_brackets_depth). 87 str_repeat('\])*', $this->nested_brackets_depth); 88 89 $this->nested_url_parenthesis_re = 90 str_repeat('(?>[^()\s]+|\(', $this->nested_url_parenthesis_depth). 91 str_repeat('(?>\)))*', $this->nested_url_parenthesis_depth); 92 93 $this->escape_chars_re = '['.preg_quote($this->escape_chars).']'; 94 95 # Sort document, block, and span gamut in ascendent priority order. 96 asort($this->document_gamut); 97 asort($this->block_gamut); 98 asort($this->span_gamut); 99 } 100 101 102 # Internal hashes used during transformation. 103 protected $urls = array(); 104 protected $titles = array(); 105 protected $html_hashes = array(); 106 107 # Status flag to avoid invalid nesting. 108 protected $in_anchor = false; 109 110 111 protected function setup() { 112 # 113 # Called before the transformation process starts to setup parser 114 # states. 115 # 116 # Clear global hashes. 117 $this->urls = $this->predef_urls; 118 $this->titles = $this->predef_titles; 119 $this->html_hashes = array(); 120 121 $this->in_anchor = false; 122 } 123 124 protected function teardown() { 125 # 126 # Called after the transformation process to clear any variable 127 # which may be taking up memory unnecessarly. 128 # 129 $this->urls = array(); 130 $this->titles = array(); 131 $this->html_hashes = array(); 132 } 133 134 135 public function transform($text) { 136 # 137 # Main function. Performs some preprocessing on the input text 138 # and pass it through the document gamut. 139 # 140 $this->setup(); 141 142 # Remove UTF-8 BOM and marker character in input, if present. 143 $text = preg_replace('{^\xEF\xBB\xBF|\x1A}', '', $text); 144 145 # Standardize line endings: 146 # DOS to Unix and Mac to Unix 147 $text = preg_replace('{\r\n?}', "\n", $text); 148 149 # Make sure $text ends with a couple of newlines: 150 $text .= "\n\n"; 151 152 # Convert all tabs to spaces. 153 $text = $this->detab($text); 154 155 # Turn block-level HTML blocks into hash entries 156 $text = $this->hashHTMLBlocks($text); 157 158 # Strip any lines consisting only of spaces and tabs. 159 # This makes subsequent regexen easier to write, because we can 160 # match consecutive blank lines with /\n+/ instead of something 161 # contorted like /[ ]*\n+/ . 162 $text = preg_replace('/^[ ]+$/m', '', $text); 163 164 # Run document gamut methods. 165 foreach ($this->document_gamut as $method => $priority) { 166 $text = $this->$method($text); 167 } 168 169 $this->teardown(); 170 171 return $text . "\n"; 172 } 173 174 protected $document_gamut = array( 175 # Strip link definitions, store in hashes. 176 "stripLinkDefinitions" => 20, 177 178 "runBasicBlockGamut" => 30, 179 ); 180 181 182 protected function stripLinkDefinitions($text) { 183 # 184 # Strips link definitions from text, stores the URLs and titles in 185 # hash references. 186 # 187 $less_than_tab = $this->tab_width - 1; 188 189 # Link defs are in the form: ^[id]: url "optional title" 190 $text = preg_replace_callback('{ 191 ^[ ]{0,'.$less_than_tab.'}\[(.+)\][ ]?: # id = $1 192 [ ]* 193 \n? # maybe *one* newline 194 [ ]* 195 (?: 196 <(.+?)> # url = $2 197 | 198 (\S+?) # url = $3 199 ) 200 [ ]* 201 \n? # maybe one newline 202 [ ]* 203 (?: 204 (?<=\s) # lookbehind for whitespace 205 ["(] 206 (.*?) # title = $4 207 [")] 208 [ ]* 209 )? # title is optional 210 (?:\n+|\Z) 211 }xm', 212 array($this, '_stripLinkDefinitions_callback'), 213 $text); 214 return $text; 215 } 216 protected function _stripLinkDefinitions_callback($matches) { 217 $link_id = strtolower($matches[1]); 218 $url = $matches[2] == '' ? $matches[3] : $matches[2]; 219 $this->urls[$link_id] = $url; 220 $this->titles[$link_id] =& $matches[4]; 221 return ''; # String that will replace the block 222 } 223 224 225 protected function hashHTMLBlocks($text) { 226 if ($this->no_markup) return $text; 227 228 $less_than_tab = $this->tab_width - 1; 229 230 # Hashify HTML blocks: 231 # We only want to do this for block-level HTML tags, such as headers, 232 # lists, and tables. That's because we still want to wrap <p>s around 233 # "paragraphs" that are wrapped in non-block-level tags, such as anchors, 234 # phrase emphasis, and spans. The list of tags we're looking for is 235 # hard-coded: 236 # 237 # * List "a" is made of tags which can be both inline or block-level. 238 # These will be treated block-level when the start tag is alone on 239 # its line, otherwise they're not matched here and will be taken as 240 # inline later. 241 # * List "b" is made of tags which are always block-level; 242 # 243 $block_tags_a_re = 'ins|del'; 244 $block_tags_b_re = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|address|'. 245 'script|noscript|style|form|fieldset|iframe|math|svg|'. 246 'article|section|nav|aside|hgroup|header|footer|'. 247 'figure'; 248 249 # Regular expression for the content of a block tag. 250 $nested_tags_level = 4; 251 $attr = ' 252 (?> # optional tag attributes 253 \s # starts with whitespace 254 (?> 255 [^>"/]+ # text outside quotes 256 | 257 /+(?!>) # slash not followed by ">" 258 | 259 "[^"]*" # text inside double quotes (tolerate ">") 260 | 261 \'[^\']*\' # text inside single quotes (tolerate ">") 262 )* 263 )? 264 '; 265 $content = 266 str_repeat(' 267 (?> 268 [^<]+ # content without tag 269 | 270 <\2 # nested opening tag 271 '.$attr.' # attributes 272 (?> 273 /> 274 | 275 >', $nested_tags_level). # end of opening tag 276 '.*?'. # last level nested tag content 277 str_repeat(' 278 </\2\s*> # closing nested tag 279 ) 280 | 281 <(?!/\2\s*> # other tags with a different name 282 ) 283 )*', 284 $nested_tags_level); 285 $content2 = str_replace('\2', '\3', $content); 286 287 # First, look for nested blocks, e.g.: 288 # <div> 289 # <div> 290 # tags for inner block must be indented. 291 # </div> 292 # </div> 293 # 294 # The outermost tags must start at the left margin for this to match, and 295 # the inner nested divs must be indented. 296 # We need to do this before the next, more liberal match, because the next 297 # match will start at the first `<div>` and stop at the first `</div>`. 298 $text = preg_replace_callback('{(?> 299 (?> 300 (?<=\n) # Starting on its own line 301 | # or 302 \A\n? # the at beginning of the doc 303 ) 304 ( # save in $1 305 306 # Match from `\n<tag>` to `</tag>\n`, handling nested tags 307 # in between. 308 309 [ ]{0,'.$less_than_tab.'} 310 <('.$block_tags_b_re.')# start tag = $2 311 '.$attr.'> # attributes followed by > and \n 312 '.$content.' # content, support nesting 313 </\2> # the matching end tag 314 [ ]* # trailing spaces/tabs 315 (?=\n+|\Z) # followed by a newline or end of document 316 317 | # Special version for tags of group a. 318 319 [ ]{0,'.$less_than_tab.'} 320 <('.$block_tags_a_re.')# start tag = $3 321 '.$attr.'>[ ]*\n # attributes followed by > 322 '.$content2.' # content, support nesting 323 </\3> # the matching end tag 324 [ ]* # trailing spaces/tabs 325 (?=\n+|\Z) # followed by a newline or end of document 326 327 | # Special case just for <hr />. It was easier to make a special 328 # case than to make the other regex more complicated. 329 330 [ ]{0,'.$less_than_tab.'} 331 <(hr) # start tag = $2 332 '.$attr.' # attributes 333 /?> # the matching end tag 334 [ ]* 335 (?=\n{2,}|\Z) # followed by a blank line or end of document 336 337 | # Special case for standalone HTML comments: 338 339 [ ]{0,'.$less_than_tab.'} 340 (?s: 341 <!-- .*? --> 342 ) 343 [ ]* 344 (?=\n{2,}|\Z) # followed by a blank line or end of document 345 346 | # PHP and ASP-style processor instructions (<? and <%) 347 348 [ ]{0,'.$less_than_tab.'} 349 (?s: 350 <([?%]) # $2 351 .*? 352 \2> 353 ) 354 [ ]* 355 (?=\n{2,}|\Z) # followed by a blank line or end of document 356 357 ) 358 )}Sxmi', 359 array($this, '_hashHTMLBlocks_callback'), 360 $text); 361 362 return $text; 363 } 364 protected function _hashHTMLBlocks_callback($matches) { 365 $text = $matches[1]; 366 $key = $this->hashBlock($text); 367 return "\n\n$key\n\n"; 368 } 369 370 371 protected function hashPart($text, $boundary = 'X') { 372 # 373 # Called whenever a tag must be hashed when a function insert an atomic 374 # element in the text stream. Passing $text to through this function gives 375 # a unique text-token which will be reverted back when calling unhash. 376 # 377 # The $boundary argument specify what character should be used to surround 378 # the token. By convension, "B" is used for block elements that needs not 379 # to be wrapped into paragraph tags at the end, ":" is used for elements 380 # that are word separators and "X" is used in the general case. 381 # 382 # Swap back any tag hash found in $text so we do not have to `unhash` 383 # multiple times at the end. 384 $text = $this->unhash($text); 385 386 # Then hash the block. 387 static $i = 0; 388 $key = "$boundary\x1A" . ++$i . $boundary; 389 $this->html_hashes[$key] = $text; 390 return $key; # String that will replace the tag. 391 } 392 393 394 protected function hashBlock($text) { 395 # 396 # Shortcut function for hashPart with block-level boundaries. 397 # 398 return $this->hashPart($text, 'B'); 399 } 400 401 402 protected $block_gamut = array( 403 # 404 # These are all the transformations that form block-level 405 # tags like paragraphs, headers, and list items. 406 # 407 "doHeaders" => 10, 408 "doHorizontalRules" => 20, 409 410 "doLists" => 40, 411 "doCodeBlocks" => 50, 412 "doBlockQuotes" => 60, 413 ); 414 415 protected function runBlockGamut($text) { 416 # 417 # Run block gamut tranformations. 418 # 419 # We need to escape raw HTML in Markdown source before doing anything 420 # else. This need to be done for each block, and not only at the 421 # begining in the Markdown function since hashed blocks can be part of 422 # list items and could have been indented. Indented blocks would have 423 # been seen as a code block in a previous pass of hashHTMLBlocks. 424 $text = $this->hashHTMLBlocks($text); 425 426 return $this->runBasicBlockGamut($text); 427 } 428 429 protected function runBasicBlockGamut($text) { 430 # 431 # Run block gamut tranformations, without hashing HTML blocks. This is 432 # useful when HTML blocks are known to be already hashed, like in the first 433 # whole-document pass. 434 # 435 foreach ($this->block_gamut as $method => $priority) { 436 $text = $this->$method($text); 437 } 438 439 # Finally form paragraph and restore hashed blocks. 440 $text = $this->formParagraphs($text); 441 442 return $text; 443 } 444 445 446 protected function doHorizontalRules($text) { 447 # Do Horizontal Rules: 448 return preg_replace( 449 '{ 450 ^[ ]{0,3} # Leading space 451 ([-*_]) # $1: First marker 452 (?> # Repeated marker group 453 [ ]{0,2} # Zero, one, or two spaces. 454 \1 # Marker character 455 ){2,} # Group repeated at least twice 456 [ ]* # Tailing spaces 457 $ # End of line. 458 }mx', 459 "\n".$this->hashBlock("<hr$this->empty_element_suffix")."\n", 460 $text); 461 } 462 463 464 protected $span_gamut = array( 465 # 466 # These are all the transformations that occur *within* block-level 467 # tags like paragraphs, headers, and list items. 468 # 469 # Process character escapes, code spans, and inline HTML 470 # in one shot. 471 "parseSpan" => -30, 472 473 # Process anchor and image tags. Images must come first, 474 # because ![foo][f] looks like an anchor. 475 "doImages" => 10, 476 "doAnchors" => 20, 477 478 # Make links out of things like `<http://example.com/>` 479 # Must come after doAnchors, because you can use < and > 480 # delimiters in inline links like [this](<url>). 481 "doAutoLinks" => 30, 482 "encodeAmpsAndAngles" => 40, 483 484 "doItalicsAndBold" => 50, 485 "doHardBreaks" => 60, 486 ); 487 488 protected function runSpanGamut($text) { 489 # 490 # Run span gamut tranformations. 491 # 492 foreach ($this->span_gamut as $method => $priority) { 493 $text = $this->$method($text); 494 } 495 496 return $text; 497 } 498 499 500 protected function doHardBreaks($text) { 501 # Do hard breaks: 502 return preg_replace_callback('/ {2,}\n/', 503 array($this, '_doHardBreaks_callback'), $text); 504 } 505 protected function _doHardBreaks_callback($matches) { 506 return $this->hashPart("<br$this->empty_element_suffix\n"); 507 } 508 509 510 protected function doAnchors($text) { 511 # 512 # Turn Markdown link shortcuts into XHTML <a> tags. 513 # 514 if ($this->in_anchor) return $text; 515 $this->in_anchor = true; 516 517 # 518 # First, handle reference-style links: [link text] [id] 519 # 520 $text = preg_replace_callback('{ 521 ( # wrap whole match in $1 522 \[ 523 ('.$this->nested_brackets_re.') # link text = $2 524 \] 525 526 [ ]? # one optional space 527 (?:\n[ ]*)? # one optional newline followed by spaces 528 529 \[ 530 (.*?) # id = $3 531 \] 532 ) 533 }xs', 534 array($this, '_doAnchors_reference_callback'), $text); 535 536 # 537 # Next, inline-style links: [link text](url "optional title") 538 # 539 $text = preg_replace_callback('{ 540 ( # wrap whole match in $1 541 \[ 542 ('.$this->nested_brackets_re.') # link text = $2 543 \] 544 \( # literal paren 545 [ \n]* 546 (?: 547 <(.+?)> # href = $3 548 | 549 ('.$this->nested_url_parenthesis_re.') # href = $4 550 ) 551 [ \n]* 552 ( # $5 553 ([\'"]) # quote char = $6 554 (.*?) # Title = $7 555 \6 # matching quote 556 [ \n]* # ignore any spaces/tabs between closing quote and ) 557 )? # title is optional 558 \) 559 ) 560 }xs', 561 array($this, '_doAnchors_inline_callback'), $text); 562 563 # 564 # Last, handle reference-style shortcuts: [link text] 565 # These must come last in case you've also got [link text][1] 566 # or [link text](/foo) 567 # 568 $text = preg_replace_callback('{ 569 ( # wrap whole match in $1 570 \[ 571 ([^\[\]]+) # link text = $2; can\'t contain [ or ] 572 \] 573 ) 574 }xs', 575 array($this, '_doAnchors_reference_callback'), $text); 576 577 $this->in_anchor = false; 578 return $text; 579 } 580 protected function _doAnchors_reference_callback($matches) { 581 $whole_match = $matches[1]; 582 $link_text = $matches[2]; 583 $link_id =& $matches[3]; 584 585 if ($link_id == "") { 586 # for shortcut links like [this][] or [this]. 587 $link_id = $link_text; 588 } 589 590 # lower-case and turn embedded newlines into spaces 591 $link_id = strtolower($link_id); 592 $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); 593 594 if (isset($this->urls[$link_id])) { 595 $url = $this->urls[$link_id]; 596 $url = $this->encodeAttribute($url); 597 598 $result = "<a href=\"$url\""; 599 if ( isset( $this->titles[$link_id] ) ) { 600 $title = $this->titles[$link_id]; 601 $title = $this->encodeAttribute($title); 602 $result .= " title=\"$title\""; 603 } 604 605 $link_text = $this->runSpanGamut($link_text); 606 $result .= ">$link_text</a>"; 607 $result = $this->hashPart($result); 608 } 609 else { 610 $result = $whole_match; 611 } 612 return $result; 613 } 614 protected function _doAnchors_inline_callback($matches) { 615 $whole_match = $matches[1]; 616 $link_text = $this->runSpanGamut($matches[2]); 617 $url = $matches[3] == '' ? $matches[4] : $matches[3]; 618 $title =& $matches[7]; 619 620 // if the URL was of the form <s p a c e s> it got caught by the HTML 621 // tag parser and hashed. Need to reverse the process before using the URL. 622 $unhashed = $this->unhash($url); 623 if ($unhashed != $url) 624 $url = preg_replace('/^<(.*)>$/', '\1', $unhashed); 625 626 $url = $this->encodeAttribute($url); 627 628 $result = "<a href=\"$url\""; 629 if (isset($title)) { 630 $title = $this->encodeAttribute($title); 631 $result .= " title=\"$title\""; 632 } 633 634 $link_text = $this->runSpanGamut($link_text); 635 $result .= ">$link_text</a>"; 636 637 return $this->hashPart($result); 638 } 639 640 641 protected function doImages($text) { 642 # 643 # Turn Markdown image shortcuts into <img> tags. 644 # 645 # 646 # First, handle reference-style labeled images: ![alt text][id] 647 # 648 $text = preg_replace_callback('{ 649 ( # wrap whole match in $1 650 !\[ 651 ('.$this->nested_brackets_re.') # alt text = $2 652 \] 653 654 [ ]? # one optional space 655 (?:\n[ ]*)? # one optional newline followed by spaces 656 657 \[ 658 (.*?) # id = $3 659 \] 660 661 ) 662 }xs', 663 array($this, '_doImages_reference_callback'), $text); 664 665 # 666 # Next, handle inline images:  667 # Don't forget: encode * and _ 668 # 669 $text = preg_replace_callback('{ 670 ( # wrap whole match in $1 671 !\[ 672 ('.$this->nested_brackets_re.') # alt text = $2 673 \] 674 \s? # One optional whitespace character 675 \( # literal paren 676 [ \n]* 677 (?: 678 <(\S*)> # src url = $3 679 | 680 ('.$this->nested_url_parenthesis_re.') # src url = $4 681 ) 682 [ \n]* 683 ( # $5 684 ([\'"]) # quote char = $6 685 (.*?) # title = $7 686 \6 # matching quote 687 [ \n]* 688 )? # title is optional 689 \) 690 ) 691 }xs', 692 array($this, '_doImages_inline_callback'), $text); 693 694 return $text; 695 } 696 protected function _doImages_reference_callback($matches) { 697 $whole_match = $matches[1]; 698 $alt_text = $matches[2]; 699 $link_id = strtolower($matches[3]); 700 701 if ($link_id == "") { 702 $link_id = strtolower($alt_text); # for shortcut links like ![this][]. 703 } 704 705 $alt_text = $this->encodeAttribute($alt_text); 706 if (isset($this->urls[$link_id])) { 707 $url = $this->encodeAttribute($this->urls[$link_id]); 708 $result = "<img src=\"$url\" alt=\"$alt_text\""; 709 if (isset($this->titles[$link_id])) { 710 $title = $this->titles[$link_id]; 711 $title = $this->encodeAttribute($title); 712 $result .= " title=\"$title\""; 713 } 714 $result .= $this->empty_element_suffix; 715 $result = $this->hashPart($result); 716 } 717 else { 718 # If there's no such link ID, leave intact: 719 $result = $whole_match; 720 } 721 722 return $result; 723 } 724 protected function _doImages_inline_callback($matches) { 725 $whole_match = $matches[1]; 726 $alt_text = $matches[2]; 727 $url = $matches[3] == '' ? $matches[4] : $matches[3]; 728 $title =& $matches[7]; 729 730 $alt_text = $this->encodeAttribute($alt_text); 731 $url = $this->encodeAttribute($url); 732 $result = "<img src=\"$url\" alt=\"$alt_text\""; 733 if (isset($title)) { 734 $title = $this->encodeAttribute($title); 735 $result .= " title=\"$title\""; # $title already quoted 736 } 737 $result .= $this->empty_element_suffix; 738 739 return $this->hashPart($result); 740 } 741 742 743 protected function doHeaders($text) { 744 # Setext-style headers: 745 # Header 1 746 # ======== 747 # 748 # Header 2 749 # -------- 750 # 751 $text = preg_replace_callback('{ ^(.+?)[ ]*\n(=+|-+)[ ]*\n+ }mx', 752 array($this, '_doHeaders_callback_setext'), $text); 753 754 # atx-style headers: 755 # # Header 1 756 # ## Header 2 757 # ## Header 2 with closing hashes ## 758 # ... 759 # ###### Header 6 760 # 761 $text = preg_replace_callback('{ 762 ^(\#{1,6}) # $1 = string of #\'s 763 [ ]* 764 (.+?) # $2 = Header text 765 [ ]* 766 \#* # optional closing #\'s (not counted) 767 \n+ 768 }xm', 769 array($this, '_doHeaders_callback_atx'), $text); 770 771 return $text; 772 } 773 protected function _doHeaders_callback_setext($matches) { 774 # Terrible hack to check we haven't found an empty list item. 775 if ($matches[2] == '-' && preg_match('{^-(?: |$)}', $matches[1])) 776 return $matches[0]; 777 778 $level = $matches[2]{0} == '=' ? 1 : 2; 779 $block = "<h$level>".$this->runSpanGamut($matches[1])."</h$level>"; 780 return "\n" . $this->hashBlock($block) . "\n\n"; 781 } 782 protected function _doHeaders_callback_atx($matches) { 783 $level = strlen($matches[1]); 784 $block = "<h$level>".$this->runSpanGamut($matches[2])."</h$level>"; 785 return "\n" . $this->hashBlock($block) . "\n\n"; 786 } 787 788 789 protected function doLists($text) { 790 # 791 # Form HTML ordered (numbered) and unordered (bulleted) lists. 792 # 793 $less_than_tab = $this->tab_width - 1; 794 795 # Re-usable patterns to match list item bullets and number markers: 796 $marker_ul_re = '[*+-]'; 797 $marker_ol_re = '\d+[\.]'; 798 799 $markers_relist = array( 800 $marker_ul_re => $marker_ol_re, 801 $marker_ol_re => $marker_ul_re, 802 ); 803 804 foreach ($markers_relist as $marker_re => $other_marker_re) { 805 # Re-usable pattern to match any entirel ul or ol list: 806 $whole_list_re = ' 807 ( # $1 = whole list 808 ( # $2 809 ([ ]{0,'.$less_than_tab.'}) # $3 = number of spaces 810 ('.$marker_re.') # $4 = first list item marker 811 [ ]+ 812 ) 813 (?s:.+?) 814 ( # $5 815 \z 816 | 817 \n{2,} 818 (?=\S) 819 (?! # Negative lookahead for another list item marker 820 [ ]* 821 '.$marker_re.'[ ]+ 822 ) 823 | 824 (?= # Lookahead for another kind of list 825 \n 826 \3 # Must have the same indentation 827 '.$other_marker_re.'[ ]+ 828 ) 829 ) 830 ) 831 '; // mx 832 833 # We use a different prefix before nested lists than top-level lists. 834 # See extended comment in _ProcessListItems(). 835 836 if ($this->list_level) { 837 $text = preg_replace_callback('{ 838 ^ 839 '.$whole_list_re.' 840 }mx', 841 array($this, '_doLists_callback'), $text); 842 } 843 else { 844 $text = preg_replace_callback('{ 845 (?:(?<=\n)\n|\A\n?) # Must eat the newline 846 '.$whole_list_re.' 847 }mx', 848 array($this, '_doLists_callback'), $text); 849 } 850 } 851 852 return $text; 853 } 854 protected function _doLists_callback($matches) { 855 # Re-usable patterns to match list item bullets and number markers: 856 $marker_ul_re = '[*+-]'; 857 $marker_ol_re = '\d+[\.]'; 858 $marker_any_re = "(?:$marker_ul_re|$marker_ol_re)"; 859 860 $list = $matches[1]; 861 $list_type = preg_match("/$marker_ul_re/", $matches[4]) ? "ul" : "ol"; 862 863 $marker_any_re = ( $list_type == "ul" ? $marker_ul_re : $marker_ol_re ); 864 865 $list .= "\n"; 866 $result = $this->processListItems($list, $marker_any_re); 867 868 $result = $this->hashBlock("<$list_type>\n" . $result . "</$list_type>"); 869 return "\n". $result ."\n\n"; 870 } 871 872 protected $list_level = 0; 873 874 protected function processListItems($list_str, $marker_any_re) { 875 # 876 # Process the contents of a single ordered or unordered list, splitting it 877 # into individual list items. 878 # 879 # The $this->list_level global keeps track of when we're inside a list. 880 # Each time we enter a list, we increment it; when we leave a list, 881 # we decrement. If it's zero, we're not in a list anymore. 882 # 883 # We do this because when we're not inside a list, we want to treat 884 # something like this: 885 # 886 # I recommend upgrading to version 887 # 8. Oops, now this line is treated 888 # as a sub-list. 889 # 890 # As a single paragraph, despite the fact that the second line starts 891 # with a digit-period-space sequence. 892 # 893 # Whereas when we're inside a list (or sub-list), that line will be 894 # treated as the start of a sub-list. What a kludge, huh? This is 895 # an aspect of Markdown's syntax that's hard to parse perfectly 896 # without resorting to mind-reading. Perhaps the solution is to 897 # change the syntax rules such that sub-lists must start with a 898 # starting cardinal number; e.g. "1." or "a.". 899 900 $this->list_level++; 901 902 # trim trailing blank lines: 903 $list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str); 904 905 $list_str = preg_replace_callback('{ 906 (\n)? # leading line = $1 907 (^[ ]*) # leading whitespace = $2 908 ('.$marker_any_re.' # list marker and space = $3 909 (?:[ ]+|(?=\n)) # space only required if item is not empty 910 ) 911 ((?s:.*?)) # list item text = $4 912 (?:(\n+(?=\n))|\n) # tailing blank line = $5 913 (?= \n* (\z | \2 ('.$marker_any_re.') (?:[ ]+|(?=\n)))) 914 }xm', 915 array($this, '_processListItems_callback'), $list_str); 916 917 $this->list_level--; 918 return $list_str; 919 } 920 protected function _processListItems_callback($matches) { 921 $item = $matches[4]; 922 $leading_line =& $matches[1]; 923 $leading_space =& $matches[2]; 924 $marker_space = $matches[3]; 925 $tailing_blank_line =& $matches[5]; 926 927 if ($leading_line || $tailing_blank_line || 928 preg_match('/\n{2,}/', $item)) 929 { 930 # Replace marker with the appropriate whitespace indentation 931 $item = $leading_space . str_repeat(' ', strlen($marker_space)) . $item; 932 $item = $this->runBlockGamut($this->outdent($item)."\n"); 933 } 934 else { 935 # Recursion for sub-lists: 936 $item = $this->doLists($this->outdent($item)); 937 $item = preg_replace('/\n+$/', '', $item); 938 $item = $this->runSpanGamut($item); 939 } 940 941 return "<li>" . $item . "</li>\n"; 942 } 943 944 945 protected function doCodeBlocks($text) { 946 # 947 # Process Markdown `<pre><code>` blocks. 948 # 949 $text = preg_replace_callback('{ 950 (?:\n\n|\A\n?) 951 ( # $1 = the code block -- one or more lines, starting with a space/tab 952 (?> 953 [ ]{'.$this->tab_width.'} # Lines must start with a tab or a tab-width of spaces 954 .*\n+ 955 )+ 956 ) 957 ((?=^[ ]{0,'.$this->tab_width.'}\S)|\Z) # Lookahead for non-space at line-start, or end of doc 958 }xm', 959 array($this, '_doCodeBlocks_callback'), $text); 960 961 return $text; 962 } 963 protected function _doCodeBlocks_callback($matches) { 964 $codeblock = $matches[1]; 965 966 $codeblock = $this->outdent($codeblock); 967 $codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES); 968 969 # trim leading newlines and trailing newlines 970 $codeblock = preg_replace('/\A\n+|\n+\z/', '', $codeblock); 971 972 $codeblock = "<pre><code>$codeblock\n</code></pre>"; 973 return "\n\n".$this->hashBlock($codeblock)."\n\n"; 974 } 975 976 977 protected function makeCodeSpan($code) { 978 # 979 # Create a code span markup for $code. Called from handleSpanToken. 980 # 981 $code = htmlspecialchars(trim($code), ENT_NOQUOTES); 982 return $this->hashPart("<code>$code</code>"); 983 } 984 985 986 protected $em_relist = array( 987 '' => '(?:(?<!\*)\*(?!\*)|(?<!_)_(?!_))(?![\.,:;]?\s)', 988 '*' => '(?<![\s*])\*(?!\*)', 989 '_' => '(?<![\s_])_(?!_)', 990 ); 991 protected $strong_relist = array( 992 '' => '(?:(?<!\*)\*\*(?!\*)|(?<!_)__(?!_))(?![\.,:;]?\s)', 993 '**' => '(?<![\s*])\*\*(?!\*)', 994 '__' => '(?<![\s_])__(?!_)', 995 ); 996 protected $em_strong_relist = array( 997 '' => '(?:(?<!\*)\*\*\*(?!\*)|(?<!_)___(?!_))(?![\.,:;]?\s)', 998 '***' => '(?<![\s*])\*\*\*(?!\*)', 999 '___' => '(?<![\s_])___(?!_)', 1000 ); 1001 protected $em_strong_prepared_relist; 1002 1003 protected function prepareItalicsAndBold() { 1004 # 1005 # Prepare regular expressions for searching emphasis tokens in any 1006 # context. 1007 # 1008 foreach ($this->em_relist as $em => $em_re) { 1009 foreach ($this->strong_relist as $strong => $strong_re) { 1010 # Construct list of allowed token expressions. 1011 $token_relist = array(); 1012 if (isset($this->em_strong_relist["$em$strong"])) { 1013 $token_relist[] = $this->em_strong_relist["$em$strong"]; 1014 } 1015 $token_relist[] = $em_re; 1016 $token_relist[] = $strong_re; 1017 1018 # Construct master expression from list. 1019 $token_re = '{('. implode('|', $token_relist) .')}'; 1020 $this->em_strong_prepared_relist["$em$strong"] = $token_re; 1021 } 1022 } 1023 } 1024 1025 protected function doItalicsAndBold($text) { 1026 $token_stack = array(''); 1027 $text_stack = array(''); 1028 $em = ''; 1029 $strong = ''; 1030 $tree_char_em = false; 1031 1032 while (1) { 1033 # 1034 # Get prepared regular expression for seraching emphasis tokens 1035 # in current context. 1036 # 1037 $token_re = $this->em_strong_prepared_relist["$em$strong"]; 1038 1039 # 1040 # Each loop iteration search for the next emphasis token. 1041 # Each token is then passed to handleSpanToken. 1042 # 1043 $parts = preg_split($token_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE); 1044 $text_stack[0] .= $parts[0]; 1045 $token =& $parts[1]; 1046 $text =& $parts[2]; 1047 1048 if (empty($token)) { 1049 # Reached end of text span: empty stack without emitting. 1050 # any more emphasis. 1051 while ($token_stack[0]) { 1052 $text_stack[1] .= array_shift($token_stack); 1053 $text_stack[0] .= array_shift($text_stack); 1054 } 1055 break; 1056 } 1057 1058 $token_len = strlen($token); 1059 if ($tree_char_em) { 1060 # Reached closing marker while inside a three-char emphasis. 1061 if ($token_len == 3) { 1062 # Three-char closing marker, close em and strong. 1063 array_shift($token_stack); 1064 $span = array_shift($text_stack); 1065 $span = $this->runSpanGamut($span); 1066 $span = "<strong><em>$span</em></strong>"; 1067 $text_stack[0] .= $this->hashPart($span); 1068 $em = ''; 1069 $strong = ''; 1070 } else { 1071 # Other closing marker: close one em or strong and 1072 # change current token state to match the other 1073 $token_stack[0] = str_repeat($token{0}, 3-$token_len); 1074 $tag = $token_len == 2 ? "strong" : "em"; 1075 $span = $text_stack[0]; 1076 $span = $this->runSpanGamut($span); 1077 $span = "<$tag>$span</$tag>"; 1078 $text_stack[0] = $this->hashPart($span); 1079 $$tag = ''; # $$tag stands for $em or $strong 1080 } 1081 $tree_char_em = false; 1082 } else if ($token_len == 3) { 1083 if ($em) { 1084 # Reached closing marker for both em and strong. 1085 # Closing strong marker: 1086 for ($i = 0; $i < 2; ++$i) { 1087 $shifted_token = array_shift($token_stack); 1088 $tag = strlen($shifted_token) == 2 ? "strong" : "em"; 1089 $span = array_shift($text_stack); 1090 $span = $this->runSpanGamut($span); 1091 $span = "<$tag>$span</$tag>"; 1092 $text_stack[0] .= $this->hashPart($span); 1093 $$tag = ''; # $$tag stands for $em or $strong 1094 } 1095 } else { 1096 # Reached opening three-char emphasis marker. Push on token 1097 # stack; will be handled by the special condition above. 1098 $em = $token{0}; 1099 $strong = "$em$em"; 1100 array_unshift($token_stack, $token); 1101 array_unshift($text_stack, ''); 1102 $tree_char_em = true; 1103 } 1104 } else if ($token_len == 2) { 1105 if ($strong) { 1106 # Unwind any dangling emphasis marker: 1107 if (strlen($token_stack[0]) == 1) { 1108 $text_stack[1] .= array_shift($token_stack); 1109 $text_stack[0] .= array_shift($text_stack); 1110 } 1111 # Closing strong marker: 1112 array_shift($token_stack); 1113 $span = array_shift($text_stack); 1114 $span = $this->runSpanGamut($span); 1115 $span = "<strong>$span</strong>"; 1116 $text_stack[0] .= $this->hashPart($span); 1117 $strong = ''; 1118 } else { 1119 array_unshift($token_stack, $token); 1120 array_unshift($text_stack, ''); 1121 $strong = $token; 1122 } 1123 } else { 1124 # Here $token_len == 1 1125 if ($em) { 1126 if (strlen($token_stack[0]) == 1) { 1127 # Closing emphasis marker: 1128 array_shift($token_stack); 1129 $span = array_shift($text_stack); 1130 $span = $this->runSpanGamut($span); 1131 $span = "<em>$span</em>"; 1132 $text_stack[0] .= $this->hashPart($span); 1133 $em = ''; 1134 } else { 1135 $text_stack[0] .= $token; 1136 } 1137 } else { 1138 array_unshift($token_stack, $token); 1139 array_unshift($text_stack, ''); 1140 $em = $token; 1141 } 1142 } 1143 } 1144 return $text_stack[0]; 1145 } 1146 1147 1148 protected function doBlockQuotes($text) { 1149 $text = preg_replace_callback('/ 1150 ( # Wrap whole match in $1 1151 (?> 1152 ^[ ]*>[ ]? # ">" at the start of a line 1153 .+\n # rest of the first line 1154 (.+\n)* # subsequent consecutive lines 1155 \n* # blanks 1156 )+ 1157 ) 1158 /xm', 1159 array($this, '_doBlockQuotes_callback'), $text); 1160 1161 return $text; 1162 } 1163 protected function _doBlockQuotes_callback($matches) { 1164 $bq = $matches[1]; 1165 # trim one level of quoting - trim whitespace-only lines 1166 $bq = preg_replace('/^[ ]*>[ ]?|^[ ]+$/m', '', $bq); 1167 $bq = $this->runBlockGamut($bq); # recurse 1168 1169 $bq = preg_replace('/^/m', " ", $bq); 1170 # These leading spaces cause problem with <pre> content, 1171 # so we need to fix that: 1172 $bq = preg_replace_callback('{(\s*<pre>.+?</pre>)}sx', 1173 array($this, '_doBlockQuotes_callback2'), $bq); 1174 1175 return "\n". $this->hashBlock("<blockquote>\n$bq\n</blockquote>")."\n\n"; 1176 } 1177 protected function _doBlockQuotes_callback2($matches) { 1178 $pre = $matches[1]; 1179 $pre = preg_replace('/^ /m', '', $pre); 1180 return $pre; 1181 } 1182 1183 1184 protected function formParagraphs($text) { 1185 # 1186 # Params: 1187 # $text - string to process with html <p> tags 1188 # 1189 # Strip leading and trailing lines: 1190 $text = preg_replace('/\A\n+|\n+\z/', '', $text); 1191 1192 $grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY); 1193 1194 # 1195 # Wrap <p> tags and unhashify HTML blocks 1196 # 1197 foreach ($grafs as $key => $value) { 1198 if (!preg_match('/^B\x1A[0-9]+B$/', $value)) { 1199 # Is a paragraph. 1200 $value = $this->runSpanGamut($value); 1201 $value = preg_replace('/^([ ]*)/', "<p>", $value); 1202 $value .= "</p>"; 1203 $grafs[$key] = $this->unhash($value); 1204 } 1205 else { 1206 # Is a block. 1207 # Modify elements of @grafs in-place... 1208 $graf = $value; 1209 $block = $this->html_hashes[$graf]; 1210 $graf = $block; 1211 // if (preg_match('{ 1212 // \A 1213 // ( # $1 = <div> tag 1214 // <div \s+ 1215 // [^>]* 1216 // \b 1217 // markdown\s*=\s* ([\'"]) # $2 = attr quote char 1218 // 1 1219 // \2 1220 // [^>]* 1221 // > 1222 // ) 1223 // ( # $3 = contents 1224 // .* 1225 // ) 1226 // (</div>) # $4 = closing tag 1227 // \z 1228 // }xs', $block, $matches)) 1229 // { 1230 // list(, $div_open, , $div_content, $div_close) = $matches; 1231 // 1232 // # We can't call Markdown(), because that resets the hash; 1233 // # that initialization code should be pulled into its own sub, though. 1234 // $div_content = $this->hashHTMLBlocks($div_content); 1235 // 1236 // # Run document gamut methods on the content. 1237 // foreach ($this->document_gamut as $method => $priority) { 1238 // $div_content = $this->$method($div_content); 1239 // } 1240 // 1241 // $div_open = preg_replace( 1242 // '{\smarkdown\s*=\s*([\'"]).+?\1}', '', $div_open); 1243 // 1244 // $graf = $div_open . "\n" . $div_content . "\n" . $div_close; 1245 // } 1246 $grafs[$key] = $graf; 1247 } 1248 } 1249 1250 return implode("\n\n", $grafs); 1251 } 1252 1253 1254 protected function encodeAttribute($text) { 1255 # 1256 # Encode text for a double-quoted HTML attribute. This function 1257 # is *not* suitable for attributes enclosed in single quotes. 1258 # 1259 $text = $this->encodeAmpsAndAngles($text); 1260 $text = str_replace('"', '"', $text); 1261 return $text; 1262 } 1263 1264 1265 protected function encodeAmpsAndAngles($text) { 1266 # 1267 # Smart processing for ampersands and angle brackets that need to 1268 # be encoded. Valid character entities are left alone unless the 1269 # no-entities mode is set. 1270 # 1271 if ($this->no_entities) { 1272 $text = str_replace('&', '&', $text); 1273 } else { 1274 # Ampersand-encoding based entirely on Nat Irons's Amputator 1275 # MT plugin: <http://bumppo.net/projects/amputator/> 1276 $text = preg_replace('/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/', 1277 '&', $text); 1278 } 1279 # Encode remaining <'s 1280 $text = str_replace('<', '<', $text); 1281 1282 return $text; 1283 } 1284 1285 1286 protected function doAutoLinks($text) { 1287 $text = preg_replace_callback('{<((https?|ftp|dict):[^\'">\s]+)>}i', 1288 array($this, '_doAutoLinks_url_callback'), $text); 1289 1290 # Email addresses: <[email protected]> 1291 $text = preg_replace_callback('{ 1292 < 1293 (?:mailto:)? 1294 ( 1295 (?: 1296 [-!#$%&\'*+/=?^_`.{|}~\w\x80-\xFF]+ 1297 | 1298 ".*?" 1299 ) 1300 \@ 1301 (?: 1302 [-a-z0-9\x80-\xFF]+(\.[-a-z0-9\x80-\xFF]+)*\.[a-z]+ 1303 | 1304 \[[\d.a-fA-F:]+\] # IPv4 & IPv6 1305 ) 1306 ) 1307 > 1308 }xi', 1309 array($this, '_doAutoLinks_email_callback'), $text); 1310 $text = preg_replace_callback('{<(tel:([^\'">\s]+))>}i',array($this, '_doAutoLinks_tel_callback'), $text); 1311 1312 return $text; 1313 } 1314 protected function _doAutoLinks_tel_callback($matches) { 1315 $url = $this->encodeAttribute($matches[1]); 1316 $tel = $this->encodeAttribute($matches[2]); 1317 $link = "<a href=\"$url\">$tel</a>"; 1318 return $this->hashPart($link); 1319 } 1320 protected function _doAutoLinks_url_callback($matches) { 1321 $url = $this->encodeAttribute($matches[1]); 1322 $link = "<a href=\"$url\">$url</a>"; 1323 return $this->hashPart($link); 1324 } 1325 protected function _doAutoLinks_email_callback($matches) { 1326 $address = $matches[1]; 1327 $link = $this->encodeEmailAddress($address); 1328 return $this->hashPart($link); 1329 } 1330 1331 1332 protected function encodeEmailAddress($addr) { 1333 # 1334 # Input: an email address, e.g. "[email protected]" 1335 # 1336 # Output: the email address as a mailto link, with each character 1337 # of the address encoded as either a decimal or hex entity, in 1338 # the hopes of foiling most address harvesting spam bots. E.g.: 1339 # 1340 # <p><a href="mailto:foo 1341 # @example.co 1342 # m">foo@exampl 1343 # e.com</a></p> 1344 # 1345 # Based by a filter by Matthew Wickline, posted to BBEdit-Talk. 1346 # With some optimizations by Milian Wolff. 1347 # 1348 $addr = "mailto:" . $addr; 1349 $chars = preg_split('/(?<!^)(?!$)/', $addr); 1350 $seed = (int)abs(crc32($addr) / strlen($addr)); # Deterministic seed. 1351 1352 foreach ($chars as $key => $char) { 1353 $ord = ord($char); 1354 # Ignore non-ascii chars. 1355 if ($ord < 128) { 1356 $r = ($seed * (1 + $key)) % 100; # Pseudo-random function. 1357 # roughly 10% raw, 45% hex, 45% dec 1358 # '@' *must* be encoded. I insist. 1359 # '"' has to be encoded inside the attribute 1360 if ($r > 90 && $char != '@' && $char != '"') /* do nothing */; 1361 else if ($r < 45) $chars[$key] = '&#x'.dechex($ord).';'; 1362 else $chars[$key] = '&#'.$ord.';'; 1363 } 1364 } 1365 1366 $addr = implode('', $chars); 1367 $text = implode('', array_slice($chars, 7)); # text without `mailto:` 1368 $addr = "<a href=\"$addr\">$text</a>"; 1369 1370 return $addr; 1371 } 1372 1373 1374 protected function parseSpan($str) { 1375 # 1376 # Take the string $str and parse it into tokens, hashing embeded HTML, 1377 # escaped characters and handling code spans. 1378 # 1379 $output = ''; 1380 1381 $span_re = '{ 1382 ( 1383 \\\\'.$this->escape_chars_re.' 1384 | 1385 (?<![`\\\\]) 1386 `+ # code span marker 1387 '.( $this->no_markup ? '' : ' 1388 | 1389 <!-- .*? --> # comment 1390 | 1391 <\?.*?\?> | <%.*?%> # processing instruction 1392 | 1393 <[!$]?[-a-zA-Z0-9:_]+ # regular tags 1394 (?> 1395 \s 1396 (?>[^"\'>]+|"[^"]*"|\'[^\']*\')* 1397 )? 1398 > 1399 | 1400 <[-a-zA-Z0-9:_]+\s*/> # xml-style empty tag 1401 | 1402 </[-a-zA-Z0-9:_]+\s*> # closing tag 1403 ').' 1404 ) 1405 }xs'; 1406 1407 while (1) { 1408 # 1409 # Each loop iteration seach for either the next tag, the next 1410 # openning code span marker, or the next escaped character. 1411 # Each token is then passed to handleSpanToken. 1412 # 1413 $parts = preg_split($span_re, $str, 2, PREG_SPLIT_DELIM_CAPTURE); 1414 1415 # Create token from text preceding tag. 1416 if ($parts[0] != "") { 1417 $output .= $parts[0]; 1418 } 1419 1420 # Check if we reach the end. 1421 if (isset($parts[1])) { 1422 $output .= $this->handleSpanToken($parts[1], $parts[2]); 1423 $str = $parts[2]; 1424 } 1425 else { 1426 break; 1427 } 1428 } 1429 1430 return $output; 1431 } 1432 1433 1434 protected function handleSpanToken($token, &$str) { 1435 # 1436 # Handle $token provided by parseSpan by determining its nature and 1437 # returning the corresponding value that should replace it. 1438 # 1439 switch ($token{0}) { 1440 case "\\": 1441 return $this->hashPart("&#". ord($token{1}). ";"); 1442 case "`": 1443 # Search for end marker in remaining text. 1444 if (preg_match('/^(.*?[^`])'.preg_quote($token).'(?!`)(.*)$/sm', 1445 $str, $matches)) 1446 { 1447 $str = $matches[2]; 1448 $codespan = $this->makeCodeSpan($matches[1]); 1449 return $this->hashPart($codespan); 1450 } 1451 return $token; // return as text since no ending marker found. 1452 default: 1453 return $this->hashPart($token); 1454 } 1455 } 1456 1457 1458 protected function outdent($text) { 1459 # 1460 # Remove one level of line-leading tabs or spaces 1461 # 1462 return preg_replace('/^(\t|[ ]{1,'.$this->tab_width.'})/m', '', $text); 1463 } 1464 1465 1466 # String length function for detab. `_initDetab` will create a function to 1467 # hanlde UTF-8 if the default function does not exist. 1468 protected $utf8_strlen = 'mb_strlen'; 1469 1470 protected function detab($text) { 1471 # 1472 # Replace tabs with the appropriate amount of space. 1473 # 1474 # For each line we separate the line in blocks delemited by 1475 # tab characters. Then we reconstruct every line by adding the 1476 # appropriate number of space between each blocks. 1477 1478 $text = preg_replace_callback('/^.*\t.*$/m', 1479 array($this, '_detab_callback'), $text); 1480 1481 return $text; 1482 } 1483 protected function _detab_callback($matches) { 1484 $line = $matches[0]; 1485 $strlen = $this->utf8_strlen; # strlen function for UTF-8. 1486 1487 # Split in blocks. 1488 $blocks = explode("\t", $line); 1489 # Add each blocks to the line. 1490 $line = $blocks[0]; 1491 unset($blocks[0]); # Do not add first block twice. 1492 foreach ($blocks as $block) { 1493 # Calculate amount of space, insert spaces, insert block. 1494 $amount = $this->tab_width - 1495 $strlen($line, 'UTF-8') % $this->tab_width; 1496 $line .= str_repeat(" ", $amount) . $block; 1497 } 1498 return $line; 1499 } 1500 protected function _initDetab() { 1501 # 1502 # Check for the availability of the function in the `utf8_strlen` property 1503 # (initially `mb_strlen`). If the function is not available, create a 1504 # function that will loosely count the number of UTF-8 characters with a 1505 # regular expression. 1506 # 1507 if (function_exists($this->utf8_strlen)) return; 1508 $this->utf8_strlen = create_function('$text', 'return preg_match_all( 1509 "/[\\\\x00-\\\\xBF]|[\\\\xC0-\\\\xFF][\\\\x80-\\\\xBF]*/", 1510 $text, $m);'); 1511 } 1512 1513 1514 protected function unhash($text) { 1515 # 1516 # Swap back in all the tags hashed by _HashHTMLBlocks. 1517 # 1518 return preg_replace_callback('/(.)\x1A[0-9]+\1/', 1519 array($this, '_unhash_callback'), $text); 1520 } 1521 protected function _unhash_callback($matches) { 1522 return $this->html_hashes[$matches[0]]; 1523 } 1524 1525 } 1526 1527 1528 # 1529 # Temporary Markdown Extra Parser Implementation Class 1530 # 1531 # NOTE: DON'T USE THIS CLASS 1532 # Currently the implementation of of Extra resides here in this temporary class. 1533 # This makes it easier to propagate the changes between the three different 1534 # packaging styles of PHP Markdown. When this issue is resolved, this 1535 # MarkdownExtra_TmpImpl class here will disappear and \Michelf\MarkdownExtra 1536 # will contain the code. So please use \Michelf\MarkdownExtra and ignore this 1537 # one. 1538 # 1539 1540 abstract class _MarkdownExtra_TmpImpl extends \Michelf\Markdown { 1541 1542 ### Configuration Variables ### 1543 1544 # Prefix for footnote ids. 1545 public $fn_id_prefix = ""; 1546 1547 # Optional title attribute for footnote links and backlinks. 1548 public $fn_link_title = ""; 1549 public $fn_backlink_title = ""; 1550 1551 # Optional class attribute for footnote links and backlinks. 1552 public $fn_link_class = "footnote-ref"; 1553 public $fn_backlink_class = "footnote-backref"; 1554 1555 # Class name for table cell alignment (%% replaced left/center/right) 1556 # For instance: 'go-%%' becomes 'go-left' or 'go-right' or 'go-center' 1557 # If empty, the align attribute is used instead of a class name. 1558 public $table_align_class_tmpl = ''; 1559 1560 # Optional class prefix for fenced code block. 1561 public $code_class_prefix = ""; 1562 # Class attribute for code blocks goes on the `code` tag; 1563 # setting this to true will put attributes on the `pre` tag instead. 1564 public $code_attr_on_pre = false; 1565 1566 # Predefined abbreviations. 1567 public $predef_abbr = array(); 1568 1569 1570 ### Parser Implementation ### 1571 1572 public function __construct() { 1573 # 1574 # Constructor function. Initialize the parser object. 1575 # 1576 # Add extra escapable characters before parent constructor 1577 # initialize the table. 1578 $this->escape_chars .= ':|'; 1579 1580 # Insert extra document, block, and span transformations. 1581 # Parent constructor will do the sorting. 1582 $this->document_gamut += array( 1583 "doFencedCodeBlocks" => 5, 1584 "stripFootnotes" => 15, 1585 "stripAbbreviations" => 25, 1586 "appendFootnotes" => 50, 1587 ); 1588 $this->block_gamut += array( 1589 "doFencedCodeBlocks" => 5, 1590 "doTables" => 15, 1591 "doDefLists" => 45, 1592 ); 1593 $this->span_gamut += array( 1594 "doFootnotes" => 5, 1595 "doAbbreviations" => 70, 1596 ); 1597 1598 parent::__construct(); 1599 } 1600 1601 1602 # Extra variables used during extra transformations. 1603 protected $footnotes = array(); 1604 protected $footnotes_ordered = array(); 1605 protected $footnotes_ref_count = array(); 1606 protected $footnotes_numbers = array(); 1607 protected $abbr_desciptions = array(); 1608 protected $abbr_word_re = ''; 1609 1610 # Give the current footnote number. 1611 protected $footnote_counter = 1; 1612 1613 1614 protected function setup() { 1615 # 1616 # Setting up Extra-specific variables. 1617 # 1618 parent::setup(); 1619 1620 $this->footnotes = array(); 1621 $this->footnotes_ordered = array(); 1622 $this->footnotes_ref_count = array(); 1623 $this->footnotes_numbers = array(); 1624 $this->abbr_desciptions = array(); 1625 $this->abbr_word_re = ''; 1626 $this->footnote_counter = 1; 1627 1628 foreach ($this->predef_abbr as $abbr_word => $abbr_desc) { 1629 if ($this->abbr_word_re) 1630 $this->abbr_word_re .= '|'; 1631 $this->abbr_word_re .= preg_quote($abbr_word); 1632 $this->abbr_desciptions[$abbr_word] = trim($abbr_desc); 1633 } 1634 } 1635 1636 protected function teardown() { 1637 # 1638 # Clearing Extra-specific variables. 1639 # 1640 $this->footnotes = array(); 1641 $this->footnotes_ordered = array(); 1642 $this->footnotes_ref_count = array(); 1643 $this->footnotes_numbers = array(); 1644 $this->abbr_desciptions = array(); 1645 $this->abbr_word_re = ''; 1646 1647 parent::teardown(); 1648 } 1649 1650 1651 ### Extra Attribute Parser ### 1652 1653 # Expression to use to catch attributes (includes the braces) 1654 protected $id_class_attr_catch_re = '\{((?:[ ]*[#.][-_:a-zA-Z0-9]+){1,})[ ]*\}'; 1655 # Expression to use when parsing in a context when no capture is desired 1656 protected $id_class_attr_nocatch_re = '\{(?:[ ]*[#.][-_:a-zA-Z0-9]+){1,}[ ]*\}'; 1657 1658 protected function doExtraAttributes($tag_name, $attr) { 1659 # 1660 # Parse attributes caught by the $this->id_class_attr_catch_re expression 1661 # and return the HTML-formatted list of attributes. 1662 # 1663 # Currently supported attributes are .class and #id. 1664 # 1665 if (empty($attr)) return ""; 1666 1667 # Split on components 1668 preg_match_all('/[#.][-_:a-zA-Z0-9]+/', $attr, $matches); 1669 $elements = $matches[0]; 1670 1671 # handle classes and ids (only first id taken into account) 1672 $classes = array(); 1673 $id = false; 1674 foreach ($elements as $element) { 1675 if ($element{0} == '.') { 1676 $classes[] = substr($element, 1); 1677 } else if ($element{0} == '#') { 1678 if ($id === false) $id = substr($element, 1); 1679 } 1680 } 1681 1682 # compose attributes as string 1683 $attr_str = ""; 1684 if (!empty($id)) { 1685 $attr_str .= ' id="'.$id.'"'; 1686 } 1687 if (!empty($classes)) { 1688 $attr_str .= ' class="'.implode(" ", $classes).'"'; 1689 } 1690 return $attr_str; 1691 } 1692 1693 1694 protected function stripLinkDefinitions($text) { 1695 # 1696 # Strips link definitions from text, stores the URLs and titles in 1697 # hash references. 1698 # 1699 $less_than_tab = $this->tab_width - 1; 1700 1701 # Link defs are in the form: ^[id]: url "optional title" 1702 $text = preg_replace_callback('{ 1703 ^[ ]{0,'.$less_than_tab.'}\[(.+)\][ ]?: # id = $1 1704 [ ]* 1705 \n? # maybe *one* newline 1706 [ ]* 1707 (?: 1708 <(.+?)> # url = $2 1709 | 1710 (\S+?) # url = $3 1711 ) 1712 [ ]* 1713 \n? # maybe one newline 1714 [ ]* 1715 (?: 1716 (?<=\s) # lookbehind for whitespace 1717 ["(] 1718 (.*?) # title = $4 1719 [")] 1720 [ ]* 1721 )? # title is optional 1722 (?:[ ]* '.$this->id_class_attr_catch_re.' )? # $5 = extra id & class attr 1723 (?:\n+|\Z) 1724 }xm', 1725 array($this, '_stripLinkDefinitions_callback'), 1726 $text); 1727 return $text; 1728 } 1729 protected function _stripLinkDefinitions_callback($matches) { 1730 $link_id = strtolower($matches[1]); 1731 $url = $matches[2] == '' ? $matches[3] : $matches[2]; 1732 $this->urls[$link_id] = $url; 1733 $this->titles[$link_id] =& $matches[4]; 1734 $this->ref_attr[$link_id] = $this->doExtraAttributes("", $dummy =& $matches[5]); 1735 return ''; # String that will replace the block 1736 } 1737 1738 1739 ### HTML Block Parser ### 1740 1741 # Tags that are always treated as block tags: 1742 protected $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|figure'; 1743 1744 # Tags treated as block tags only if the opening tag is alone on its line: 1745 protected $context_block_tags_re = 'script|noscript|style|ins|del|iframe|object|source|track|param|math|svg|canvas|audio|video'; 1746 1747 # Tags where markdown="1" default to span mode: 1748 protected $contain_span_tags_re = 'p|h[1-6]|li|dd|dt|td|th|legend|address'; 1749 1750 # Tags which must not have their contents modified, no matter where 1751 # they appear: 1752 protected $clean_tags_re = 'script|style|math|svg'; 1753 1754 # Tags that do not need to be closed. 1755 protected $auto_close_tags_re = 'hr|img|param|source|track'; 1756 1757 1758 protected function hashHTMLBlocks($text) { 1759 # 1760 # Hashify HTML Blocks and "clean tags". 1761 # 1762 # We only want to do this for block-level HTML tags, such as headers, 1763 # lists, and tables. That's because we still want to wrap <p>s around 1764 # "paragraphs" that are wrapped in non-block-level tags, such as anchors, 1765 # phrase emphasis, and spans. The list of tags we're looking for is 1766 # hard-coded. 1767 # 1768 # This works by calling _HashHTMLBlocks_InMarkdown, which then calls 1769 # _HashHTMLBlocks_InHTML when it encounter block tags. When the markdown="1" 1770 # attribute is found within a tag, _HashHTMLBlocks_InHTML calls back 1771 # _HashHTMLBlocks_InMarkdown to handle the Markdown syntax within the tag. 1772 # These two functions are calling each other. It's recursive! 1773 # 1774 if ($this->no_markup) return $text; 1775 1776 # 1777 # Call the HTML-in-Markdown hasher. 1778 # 1779 list($text, ) = $this->_hashHTMLBlocks_inMarkdown($text); 1780 1781 return $text; 1782 } 1783 protected function _hashHTMLBlocks_inMarkdown($text, $indent = 0, 1784 $enclosing_tag_re = '', $span = false) 1785 { 1786 # 1787 # Parse markdown text, calling _HashHTMLBlocks_InHTML for block tags. 1788 # 1789 # * $indent is the number of space to be ignored when checking for code 1790 # blocks. This is important because if we don't take the indent into 1791 # account, something like this (which looks right) won't work as expected: 1792 # 1793 # <div> 1794 # <div markdown="1"> 1795 # Hello World. <-- Is this a Markdown code block or text? 1796 # </div> <-- Is this a Markdown code block or a real tag? 1797 # <div> 1798 # 1799 # If you don't like this, just don't indent the tag on which 1800 # you apply the markdown="1" attribute. 1801 # 1802 # * If $enclosing_tag_re is not empty, stops at the first unmatched closing 1803 # tag with that name. Nested tags supported. 1804 # 1805 # * If $span is true, text inside must treated as span. So any double 1806 # newline will be replaced by a single newline so that it does not create 1807 # paragraphs. 1808 # 1809 # Returns an array of that form: ( processed text , remaining text ) 1810 # 1811 if ($text === '') return array('', ''); 1812 1813 # Regex to check for the presense of newlines around a block tag. 1814 $newline_before_re = '/(?:^\n?|\n\n)*$/'; 1815 $newline_after_re = 1816 '{ 1817 ^ # Start of text following the tag. 1818 (?>[ ]*<!--.*?-->)? # Optional comment. 1819 [ ]*\n # Must be followed by newline. 1820 }xs'; 1821 1822 # Regex to match any tag. 1823 $block_tag_re = 1824 '{ 1825 ( # $2: Capture whole tag. 1826 </? # Any opening or closing tag. 1827 (?> # Tag name. 1828 '.$this->block_tags_re.' | 1829 '.$this->context_block_tags_re.' | 1830 '.$this->clean_tags_re.' | 1831 (?!\s)'.$enclosing_tag_re.' 1832 ) 1833 (?: 1834 (?=[\s"\'/a-zA-Z0-9]) # Allowed characters after tag name. 1835 (?> 1836 ".*?" | # Double quotes (can contain `>`) 1837 \'.*?\' | # Single quotes (can contain `>`) 1838 .+? # Anything but quotes and `>`. 1839 )*? 1840 )? 1841 > # End of tag. 1842 | 1843 <!-- .*? --> # HTML Comment 1844 | 1845 <\?.*?\?> | <%.*?%> # Processing instruction 1846 | 1847 <!\[CDATA\[.*?\]\]> # CData Block 1848 '. ( !$span ? ' # If not in span. 1849 | 1850 # Indented code block 1851 (?: ^[ ]*\n | ^ | \n[ ]*\n ) 1852 [ ]{'.($indent+4).'}[^\n]* \n 1853 (?> 1854 (?: [ ]{'.($indent+4).'}[^\n]* | [ ]* ) \n 1855 )* 1856 | 1857 # Fenced code block marker 1858 (?<= ^ | \n ) 1859 [ ]{0,'.($indent+3).'}(?:~{3,}|`{3,}) 1860 [ ]* 1861 (?: 1862 \.?[-_:a-zA-Z0-9]+ # standalone class name 1863 | 1864 '.$this->id_class_attr_nocatch_re.' # extra attributes 1865 )? 1866 [ ]* 1867 (?= \n ) 1868 ' : '' ). ' # End (if not is span). 1869 | 1870 # Code span marker 1871 # Note, this regex needs to go after backtick fenced 1872 # code blocks but it should also be kept outside of the 1873 # "if not in span" condition adding backticks to the parser 1874 `+ 1875 ) 1876 }xs'; 1877 1878 1879 $depth = 0; # Current depth inside the tag tree. 1880 $parsed = ""; # Parsed text that will be returned. 1881 1882 # 1883 # Loop through every tag until we find the closing tag of the parent 1884 # or loop until reaching the end of text if no parent tag specified. 1885 # 1886 do { 1887 # 1888 # Split the text using the first $tag_match pattern found. 1889 # Text before pattern will be first in the array, text after 1890 # pattern will be at the end, and between will be any catches made 1891 # by the pattern. 1892 # 1893 $parts = preg_split($block_tag_re, $text, 2, 1894 PREG_SPLIT_DELIM_CAPTURE); 1895 1896 # If in Markdown span mode, add a empty-string span-level hash 1897 # after each newline to prevent triggering any block element. 1898 if ($span) { 1899 $void = $this->hashPart("", ':'); 1900 $newline = "$void\n"; 1901 $parts[0] = $void . str_replace("\n", $newline, $parts[0]) . $void; 1902 } 1903 1904 $parsed .= $parts[0]; # Text before current tag. 1905 1906 # If end of $text has been reached. Stop loop. 1907 if (count($parts) < 3) { 1908 $text = ""; 1909 break; 1910 } 1911 1912 $tag = $parts[1]; # Tag to handle. 1913 $text = $parts[2]; # Remaining text after current tag. 1914 $tag_re = preg_quote($tag); # For use in a regular expression. 1915 1916 # 1917 # Check for: Fenced code block marker. 1918 # Note: need to recheck the whole tag to disambiguate backtick 1919 # fences from code spans 1920 # 1921 if (preg_match('{^\n?([ ]{0,'.($indent+3).'})(~{3,}|`{3,})[ ]*(?:\.?[-_:a-zA-Z0-9]+|'.$this->id_class_attr_nocatch_re.')?[ ]*\n?$}', $tag, $capture)) { 1922 # Fenced code block marker: find matching end marker. 1923 $fence_indent = strlen($capture[1]); # use captured indent in re 1924 $fence_re = $capture[2]; # use captured fence in re 1925 if (preg_match('{^(?>.*\n)*?[ ]{'.($fence_indent).'}'.$fence_re.'[ ]*(?:\n|$)}', $text, 1926 $matches)) 1927 { 1928 # End marker found: pass text unchanged until marker. 1929 $parsed .= $tag . $matches[0]; 1930 $text = substr($text, strlen($matches[0])); 1931 } 1932 else { 1933 # No end marker: just skip it. 1934 $parsed .= $tag; 1935 } 1936 } 1937 # 1938 # Check for: Indented code block. 1939 # 1940 else if ($tag{0} == "\n" || $tag{0} == " ") { 1941 # Indented code block: pass it unchanged, will be handled 1942 # later. 1943 $parsed .= $tag; 1944 } 1945 # 1946 # Check for: Code span marker 1947 # Note: need to check this after backtick fenced code blocks 1948 # 1949 else if ($tag{0} == "`") { 1950 # Find corresponding end marker. 1951 $tag_re = preg_quote($tag); 1952 if (preg_match('{^(?>.+?|\n(?!\n))*?(?<!`)'.$tag_re.'(?!`)}', 1953 $text, $matches)) 1954 { 1955 # End marker found: pass text unchanged until marker. 1956 $parsed .= $tag . $matches[0]; 1957 $text = substr($text, strlen($matches[0])); 1958 } 1959 else { 1960 # Unmatched marker: just skip it. 1961 $parsed .= $tag; 1962 } 1963 } 1964 # 1965 # Check for: Opening Block level tag or 1966 # Opening Context Block tag (like ins and del) 1967 # used as a block tag (tag is alone on it's line). 1968 # 1969 else if (preg_match('{^<(?:'.$this->block_tags_re.')\b}', $tag) || 1970 ( preg_match('{^<(?:'.$this->context_block_tags_re.')\b}', $tag) && 1971 preg_match($newline_before_re, $parsed) && 1972 preg_match($newline_after_re, $text) ) 1973 ) 1974 { 1975 # Need to parse tag and following text using the HTML parser. 1976 list($block_text, $text) = 1977 $this->_hashHTMLBlocks_inHTML($tag . $text, "hashBlock", true); 1978 1979 # Make sure it stays outside of any paragraph by adding newlines. 1980 $parsed .= "\n\n$block_text\n\n"; 1981 } 1982 # 1983 # Check for: Clean tag (like script, math) 1984 # HTML Comments, processing instructions. 1985 # 1986 else if (preg_match('{^<(?:'.$this->clean_tags_re.')\b}', $tag) || 1987 $tag{1} == '!' || $tag{1} == '?') 1988 { 1989 # Need to parse tag and following text using the HTML parser. 1990 # (don't check for markdown attribute) 1991 list($block_text, $text) = 1992 $this->_hashHTMLBlocks_inHTML($tag . $text, "hashClean", false); 1993 1994 $parsed .= $block_text; 1995 } 1996 # 1997 # Check for: Tag with same name as enclosing tag. 1998 # 1999 else if ($enclosing_tag_re !== '' && 2000 # Same name as enclosing tag. 2001 preg_match('{^</?(?:'.$enclosing_tag_re.')\b}', $tag)) 2002 { 2003 # 2004 # Increase/decrease nested tag count. 2005 # 2006 if ($tag{1} == '/') $depth--; 2007 else if ($tag{strlen($tag)-2} != '/') $depth++; 2008 2009 if ($depth < 0) { 2010 # 2011 # Going out of parent element. Clean up and break so we 2012 # return to the calling function. 2013 # 2014 $text = $tag . $text; 2015 break; 2016 } 2017 2018 $parsed .= $tag; 2019 } 2020 else { 2021 $parsed .= $tag; 2022 } 2023 } while ($depth >= 0); 2024 2025 return array($parsed, $text); 2026 } 2027 protected function _hashHTMLBlocks_inHTML($text, $hash_method, $md_attr) { 2028 # 2029 # Parse HTML, calling _HashHTMLBlocks_InMarkdown for block tags. 2030 # 2031 # * Calls $hash_method to convert any blocks. 2032 # * Stops when the first opening tag closes. 2033 # * $md_attr indicate if the use of the `markdown="1"` attribute is allowed. 2034 # (it is not inside clean tags) 2035 # 2036 # Returns an array of that form: ( processed text , remaining text ) 2037 # 2038 if ($text === '') return array('', ''); 2039 2040 # Regex to match `markdown` attribute inside of a tag. 2041 $markdown_attr_re = ' 2042 { 2043 \s* # Eat whitespace before the `markdown` attribute 2044 markdown 2045 \s*=\s* 2046 (?> 2047 (["\']) # $1: quote delimiter 2048 (.*?) # $2: attribute value 2049 \1 # matching delimiter 2050 | 2051 ([^\s>]*) # $3: unquoted attribute value 2052 ) 2053 () # $4: make $3 always defined (avoid warnings) 2054 }xs'; 2055 2056 # Regex to match any tag. 2057 $tag_re = '{ 2058 ( # $2: Capture whole tag. 2059 </? # Any opening or closing tag. 2060 [\w:$]+ # Tag name. 2061 (?: 2062 (?=[\s"\'/a-zA-Z0-9]) # Allowed characters after tag name. 2063 (?> 2064 ".*?" | # Double quotes (can contain `>`) 2065 \'.*?\' | # Single quotes (can contain `>`) 2066 .+? # Anything but quotes and `>`. 2067 )*? 2068 )? 2069 > # End of tag. 2070 | 2071 <!-- .*? --> # HTML Comment 2072 | 2073 <\?.*?\?> | <%.*?%> # Processing instruction 2074 | 2075 <!\[CDATA\[.*?\]\]> # CData Block 2076 ) 2077 }xs'; 2078 2079 $original_text = $text; # Save original text in case of faliure. 2080 2081 $depth = 0; # Current depth inside the tag tree. 2082 $block_text = ""; # Temporary text holder for current text. 2083 $parsed = ""; # Parsed text that will be returned. 2084 2085 # 2086 # Get the name of the starting tag. 2087 # (This pattern makes $base_tag_name_re safe without quoting.) 2088 # 2089 if (preg_match('/^<([\w:$]*)\b/', $text, $matches)) 2090 $base_tag_name_re = $matches[1]; 2091 2092 # 2093 # Loop through every tag until we find the corresponding closing tag. 2094 # 2095 do { 2096 # 2097 # Split the text using the first $tag_match pattern found. 2098 # Text before pattern will be first in the array, text after 2099 # pattern will be at the end, and between will be any catches made 2100 # by the pattern. 2101 # 2102 $parts = preg_split($tag_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE); 2103 2104 if (count($parts) < 3) { 2105 # 2106 # End of $text reached with unbalenced tag(s). 2107 # In that case, we return original text unchanged and pass the 2108 # first character as filtered to prevent an infinite loop in the 2109 # parent function. 2110 # 2111 return array($original_text{0}, substr($original_text, 1)); 2112 } 2113 2114 $block_text .= $parts[0]; # Text before current tag. 2115 $tag = $parts[1]; # Tag to handle. 2116 $text = $parts[2]; # Remaining text after current tag. 2117 2118 # 2119 # Check for: Auto-close tag (like <hr/>) 2120 # Comments and Processing Instructions. 2121 # 2122 if (preg_match('{^</?(?:'.$this->auto_close_tags_re.')\b}', $tag) || 2123 $tag{1} == '!' || $tag{1} == '?') 2124 { 2125 # Just add the tag to the block as if it was text. 2126 $block_text .= $tag; 2127 } 2128 else { 2129 # 2130 # Increase/decrease nested tag count. Only do so if 2131 # the tag's name match base tag's. 2132 # 2133 if (preg_match('{^</?'.$base_tag_name_re.'\b}', $tag)) { 2134 if ($tag{1} == '/') $depth--; 2135 else if ($tag{strlen($tag)-2} != '/') $depth++; 2136 } 2137 2138 # 2139 # Check for `markdown="1"` attribute and handle it. 2140 # 2141 if ($md_attr && 2142 preg_match($markdown_attr_re, $tag, $attr_m) && 2143 preg_match('/^1|block|span$/', $attr_m[2] . $attr_m[3])) 2144 { 2145 # Remove `markdown` attribute from opening tag. 2146 $tag = preg_replace($markdown_attr_re, '', $tag); 2147 2148 # Check if text inside this tag must be parsed in span mode. 2149 $this->mode = $attr_m[2] . $attr_m[3]; 2150 $span_mode = $this->mode == 'span' || $this->mode != 'block' && 2151 preg_match('{^<(?:'.$this->contain_span_tags_re.')\b}', $tag); 2152 2153 # Calculate indent before tag. 2154 if (preg_match('/(?:^|\n)( *?)(?! ).*?$/', $block_text, $matches)) { 2155 $strlen = $this->utf8_strlen; 2156 $indent = $strlen($matches[1], 'UTF-8'); 2157 } else { 2158 $indent = 0; 2159 } 2160 2161 # End preceding block with this tag. 2162 $block_text .= $tag; 2163 $parsed .= $this->$hash_method($block_text); 2164 2165 # Get enclosing tag name for the ParseMarkdown function. 2166 # (This pattern makes $tag_name_re safe without quoting.) 2167 preg_match('/^<([\w:$]*)\b/', $tag, $matches); 2168 $tag_name_re = $matches[1]; 2169 2170 # Parse the content using the HTML-in-Markdown parser. 2171 list ($block_text, $text) 2172 = $this->_hashHTMLBlocks_inMarkdown($text, $indent, 2173 $tag_name_re, $span_mode); 2174 2175 # Outdent markdown text. 2176 if ($indent > 0) { 2177 $block_text = preg_replace("/^[ ]{1,$indent}/m", "", 2178 $block_text); 2179 } 2180 2181 # Append tag content to parsed text. 2182 if (!$span_mode) $parsed .= "\n\n$block_text\n\n"; 2183 else $parsed .= "$block_text"; 2184 2185 # Start over with a new block. 2186 $block_text = ""; 2187 } 2188 else $block_text .= $tag; 2189 } 2190 2191 } while ($depth > 0); 2192 2193 # 2194 # Hash last block text that wasn't processed inside the loop. 2195 # 2196 $parsed .= $this->$hash_method($block_text); 2197 2198 return array($parsed, $text); 2199 } 2200 2201 2202 protected function hashClean($text) { 2203 # 2204 # Called whenever a tag must be hashed when a function inserts a "clean" tag 2205 # in $text, it passes through this function and is automaticaly escaped, 2206 # blocking invalid nested overlap. 2207 # 2208 return $this->hashPart($text, 'C'); 2209 } 2210 2211 2212 protected function doAnchors($text) { 2213 # 2214 # Turn Markdown link shortcuts into XHTML <a> tags. 2215 # 2216 if ($this->in_anchor) return $text; 2217 $this->in_anchor = true; 2218 2219 # 2220 # First, handle reference-style links: [link text] [id] 2221 # 2222 $text = preg_replace_callback('{ 2223 ( # wrap whole match in $1 2224 \[ 2225 ('.$this->nested_brackets_re.') # link text = $2 2226 \] 2227 2228 [ ]? # one optional space 2229 (?:\n[ ]*)? # one optional newline followed by spaces 2230 2231 \[ 2232 (.*?) # id = $3 2233 \] 2234 ) 2235 }xs', 2236 array($this, '_doAnchors_reference_callback'), $text); 2237 2238 # 2239 # Next, inline-style links: [link text](url "optional title") 2240 # 2241 $text = preg_replace_callback('{ 2242 ( # wrap whole match in $1 2243 \[ 2244 ('.$this->nested_brackets_re.') # link text = $2 2245 \] 2246 \( # literal paren 2247 [ \n]* 2248 (?: 2249 <(.+?)> # href = $3 2250 | 2251 ('.$this->nested_url_parenthesis_re.') # href = $4 2252 ) 2253 [ \n]* 2254 ( # $5 2255 ([\'"]) # quote char = $6 2256 (.*?) # Title = $7 2257 \6 # matching quote 2258 [ \n]* # ignore any spaces/tabs between closing quote and ) 2259 )? # title is optional 2260 \) 2261 (?:[ ]? '.$this->id_class_attr_catch_re.' )? # $8 = id/class attributes 2262 ) 2263 }xs', 2264 array($this, '_doAnchors_inline_callback'), $text); 2265 2266 # 2267 # Last, handle reference-style shortcuts: [link text] 2268 # These must come last in case you've also got [link text][1] 2269 # or [link text](/foo) 2270 # 2271 $text = preg_replace_callback('{ 2272 ( # wrap whole match in $1 2273 \[ 2274 ([^\[\]]+) # link text = $2; can\'t contain [ or ] 2275 \] 2276 ) 2277 }xs', 2278 array($this, '_doAnchors_reference_callback'), $text); 2279 2280 $this->in_anchor = false; 2281 return $text; 2282 } 2283 protected function _doAnchors_reference_callback($matches) { 2284 $whole_match = $matches[1]; 2285 $link_text = $matches[2]; 2286 $link_id =& $matches[3]; 2287 2288 if ($link_id == "") { 2289 # for shortcut links like [this][] or [this]. 2290 $link_id = $link_text; 2291 } 2292 2293 # lower-case and turn embedded newlines into spaces 2294 $link_id = strtolower($link_id); 2295 $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); 2296 2297 if (isset($this->urls[$link_id])) { 2298 $url = $this->urls[$link_id]; 2299 $url = $this->encodeAttribute($url); 2300 2301 $result = "<a href=\"$url\""; 2302 if ( isset( $this->titles[$link_id] ) ) { 2303 $title = $this->titles[$link_id]; 2304 $title = $this->encodeAttribute($title); 2305 $result .= " title=\"$title\""; 2306 } 2307 if (isset($this->ref_attr[$link_id])) 2308 $result .= $this->ref_attr[$link_id]; 2309 2310 $link_text = $this->runSpanGamut($link_text); 2311 $result .= ">$link_text</a>"; 2312 $result = $this->hashPart($result); 2313 } 2314 else { 2315 $result = $whole_match; 2316 } 2317 return $result; 2318 } 2319 protected function _doAnchors_inline_callback($matches) { 2320 $whole_match = $matches[1]; 2321 $link_text = $this->runSpanGamut($matches[2]); 2322 $url = $matches[3] == '' ? $matches[4] : $matches[3]; 2323 $title =& $matches[7]; 2324 $attr = $this->doExtraAttributes("a", $dummy =& $matches[8]); 2325 2326 // if the URL was of the form <s p a c e s> it got caught by the HTML 2327 // tag parser and hashed. Need to reverse the process before using the URL. 2328 $unhashed = $this->unhash($url); 2329 if ($unhashed != $url) 2330 $url = preg_replace('/^<(.*)>$/', '\1', $unhashed); 2331 2332 $url = $this->encodeAttribute($url); 2333 2334 $result = "<a href=\"$url\""; 2335 if (isset($title)) { 2336 $title = $this->encodeAttribute($title); 2337 $result .= " title=\"$title\""; 2338 } 2339 $result .= $attr; 2340 2341 $link_text = $this->runSpanGamut($link_text); 2342 $result .= ">$link_text</a>"; 2343 2344 return $this->hashPart($result); 2345 } 2346 2347 2348 protected function doImages($text) { 2349 # 2350 # Turn Markdown image shortcuts into <img> tags. 2351 # 2352 # 2353 # First, handle reference-style labeled images: ![alt text][id] 2354 # 2355 $text = preg_replace_callback('{ 2356 ( # wrap whole match in $1 2357 !\[ 2358 ('.$this->nested_brackets_re.') # alt text = $2 2359 \] 2360 2361 [ ]? # one optional space 2362 (?:\n[ ]*)? # one optional newline followed by spaces 2363 2364 \[ 2365 (.*?) # id = $3 2366 \] 2367 2368 ) 2369 }xs', 2370 array($this, '_doImages_reference_callback'), $text); 2371 2372 # 2373 # Next, handle inline images:  2374 # Don't forget: encode * and _ 2375 # 2376 $text = preg_replace_callback('{ 2377 ( # wrap whole match in $1 2378 !\[ 2379 ('.$this->nested_brackets_re.') # alt text = $2 2380 \] 2381 \s? # One optional whitespace character 2382 \( # literal paren 2383 [ \n]* 2384 (?: 2385 <(\S*)> # src url = $3 2386 | 2387 ('.$this->nested_url_parenthesis_re.') # src url = $4 2388 ) 2389 [ \n]* 2390 ( # $5 2391 ([\'"]) # quote char = $6 2392 (.*?) # title = $7 2393 \6 # matching quote 2394 [ \n]* 2395 )? # title is optional 2396 \) 2397 (?:[ ]? '.$this->id_class_attr_catch_re.' )? # $8 = id/class attributes 2398 ) 2399 }xs', 2400 array($this, '_doImages_inline_callback'), $text); 2401 2402 return $text; 2403 } 2404 protected function _doImages_reference_callback($matches) { 2405 $whole_match = $matches[1]; 2406 $alt_text = $matches[2]; 2407 $link_id = strtolower($matches[3]); 2408 2409 if ($link_id == "") { 2410 $link_id = strtolower($alt_text); # for shortcut links like ![this][]. 2411 } 2412 2413 $alt_text = $this->encodeAttribute($alt_text); 2414 if (isset($this->urls[$link_id])) { 2415 $url = $this->encodeAttribute($this->urls[$link_id]); 2416 $result = "<img src=\"$url\" alt=\"$alt_text\""; 2417 if (isset($this->titles[$link_id])) { 2418 $title = $this->titles[$link_id]; 2419 $title = $this->encodeAttribute($title); 2420 $result .= " title=\"$title\""; 2421 } 2422 if (isset($this->ref_attr[$link_id])) 2423 $result .= $this->ref_attr[$link_id]; 2424 $result .= $this->empty_element_suffix; 2425 $result = $this->hashPart($result); 2426 } 2427 else { 2428 # If there's no such link ID, leave intact: 2429 $result = $whole_match; 2430 } 2431 2432 return $result; 2433 } 2434 protected function _doImages_inline_callback($matches) { 2435 $whole_match = $matches[1]; 2436 $alt_text = $matches[2]; 2437 $url = $matches[3] == '' ? $matches[4] : $matches[3]; 2438 $title =& $matches[7]; 2439 $attr = $this->doExtraAttributes("img", $dummy =& $matches[8]); 2440 2441 $alt_text = $this->encodeAttribute($alt_text); 2442 $url = $this->encodeAttribute($url); 2443 $result = "<img src=\"$url\" alt=\"$alt_text\""; 2444 if (isset($title)) { 2445 $title = $this->encodeAttribute($title); 2446 $result .= " title=\"$title\""; # $title already quoted 2447 } 2448 $result .= $attr; 2449 $result .= $this->empty_element_suffix; 2450 2451 return $this->hashPart($result); 2452 } 2453 2454 2455 protected function doHeaders($text) { 2456 # 2457 # Redefined to add id and class attribute support. 2458 # 2459 # Setext-style headers: 2460 # Header 1 {#header1} 2461 # ======== 2462 # 2463 # Header 2 {#header2 .class1 .class2} 2464 # -------- 2465 # 2466 $text = preg_replace_callback( 2467 '{ 2468 (^.+?) # $1: Header text 2469 (?:[ ]+ '.$this->id_class_attr_catch_re.' )? # $3 = id/class attributes 2470 [ ]*\n(=+|-+)[ ]*\n+ # $3: Header footer 2471 }mx', 2472 array($this, '_doHeaders_callback_setext'), $text); 2473 2474 # atx-style headers: 2475 # # Header 1 {#header1} 2476 # ## Header 2 {#header2} 2477 # ## Header 2 with closing hashes ## {#header3.class1.class2} 2478 # ... 2479 # ###### Header 6 {.class2} 2480 # 2481 $text = preg_replace_callback('{ 2482 ^(\#{1,6}) # $1 = string of #\'s 2483 [ ]* 2484 (.+?) # $2 = Header text 2485 [ ]* 2486 \#* # optional closing #\'s (not counted) 2487 (?:[ ]+ '.$this->id_class_attr_catch_re.' )? # $3 = id/class attributes 2488 [ ]* 2489 \n+ 2490 }xm', 2491 array($this, '_doHeaders_callback_atx'), $text); 2492 2493 return $text; 2494 } 2495 protected function _doHeaders_callback_setext($matches) { 2496 if ($matches[3] == '-' && preg_match('{^- }', $matches[1])) 2497 return $matches[0]; 2498 $level = $matches[3]{0} == '=' ? 1 : 2; 2499 $attr = $this->doExtraAttributes("h$level", $dummy =& $matches[2]); 2500 $block = "<h$level$attr>".$this->runSpanGamut($matches[1])."</h$level>"; 2501 return "\n" . $this->hashBlock($block) . "\n\n"; 2502 } 2503 protected function _doHeaders_callback_atx($matches) { 2504 $level = strlen($matches[1]); 2505 $attr = $this->doExtraAttributes("h$level", $dummy =& $matches[3]); 2506 $block = "<h$level$attr>".$this->runSpanGamut($matches[2])."</h$level>"; 2507 return "\n" . $this->hashBlock($block) . "\n\n"; 2508 } 2509 2510 2511 protected function doTables($text) { 2512 # 2513 # Form HTML tables. 2514 # 2515 $less_than_tab = $this->tab_width - 1; 2516 # 2517 # Find tables with leading pipe. 2518 # 2519 # | Header 1 | Header 2 2520 # | -------- | -------- 2521 # | Cell 1 | Cell 2 2522 # | Cell 3 | Cell 4 2523 # 2524 $text = preg_replace_callback(' 2525 { 2526 ^ # Start of a line 2527 [ ]{0,'.$less_than_tab.'} # Allowed whitespace. 2528 [|] # Optional leading pipe (present) 2529 (.+) \n # $1: Header row (at least one pipe) 2530 2531 [ ]{0,'.$less_than_tab.'} # Allowed whitespace. 2532 [|] ([ ]*[-:]+[-| :]*) \n # $2: Header underline 2533 2534 ( # $3: Cells 2535 (?> 2536 [ ]* # Allowed whitespace. 2537 [|] .* \n # Row content. 2538 )* 2539 ) 2540 (?=\n|\Z) # Stop at final double newline. 2541 }xm', 2542 array($this, '_doTable_leadingPipe_callback'), $text); 2543 2544 # 2545 # Find tables without leading pipe. 2546 # 2547 # Header 1 | Header 2 2548 # -------- | -------- 2549 # Cell 1 | Cell 2 2550 # Cell 3 | Cell 4 2551 # 2552 $text = preg_replace_callback(' 2553 { 2554 ^ # Start of a line 2555 [ ]{0,'.$less_than_tab.'} # Allowed whitespace. 2556 (\S.*[|].*) \n # $1: Header row (at least one pipe) 2557 2558 [ ]{0,'.$less_than_tab.'} # Allowed whitespace. 2559 ([-:]+[ ]*[|][-| :]*) \n # $2: Header underline 2560 2561 ( # $3: Cells 2562 (?> 2563 .* [|] .* \n # Row content 2564 )* 2565 ) 2566 (?=\n|\Z) # Stop at final double newline. 2567 }xm', 2568 array($this, '_DoTable_callback'), $text); 2569 2570 return $text; 2571 } 2572 protected function _doTable_leadingPipe_callback($matches) { 2573 $head = $matches[1]; 2574 $underline = $matches[2]; 2575 $content = $matches[3]; 2576 2577 # Remove leading pipe for each row. 2578 $content = preg_replace('/^ *[|]/m', '', $content); 2579 2580 return $this->_doTable_callback(array($matches[0], $head, $underline, $content)); 2581 } 2582 protected function _doTable_makeAlignAttr($alignname) 2583 { 2584 if (empty($this->table_align_class_tmpl)) 2585 return " align=\"$alignname\""; 2586 2587 $classname = str_replace('%%', $alignname, $this->table_align_class_tmpl); 2588 return " class=\"$classname\""; 2589 } 2590 protected function _doTable_callback($matches) { 2591 $head = $matches[1]; 2592 $underline = $matches[2]; 2593 $content = $matches[3]; 2594 2595 # Remove any tailing pipes for each line. 2596 $head = preg_replace('/[|] *$/m', '', $head); 2597 $underline = preg_replace('/[|] *$/m', '', $underline); 2598 $content = preg_replace('/[|] *$/m', '', $content); 2599 2600 # Reading alignement from header underline. 2601 $separators = preg_split('/ *[|] */', $underline); 2602 foreach ($separators as $n => $s) { 2603 if (preg_match('/^ *-+: *$/', $s)) 2604 $attr[$n] = $this->_doTable_makeAlignAttr('right'); 2605 else if (preg_match('/^ *:-+: *$/', $s)) 2606 $attr[$n] = $this->_doTable_makeAlignAttr('center'); 2607 else if (preg_match('/^ *:-+ *$/', $s)) 2608 $attr[$n] = $this->_doTable_makeAlignAttr('left'); 2609 else 2610 $attr[$n] = ''; 2611 } 2612 2613 # Parsing span elements, including code spans, character escapes, 2614 # and inline HTML tags, so that pipes inside those gets ignored. 2615 $head = $this->parseSpan($head); 2616 $headers = preg_split('/ *[|] */', $head); 2617 $col_count = count($headers); 2618 $attr = array_pad($attr, $col_count, ''); 2619 2620 # Write column headers. 2621 $text = "<table>\n"; 2622 $text .= "<thead>\n"; 2623 $text .= "<tr>\n"; 2624 foreach ($headers as $n => $header) 2625 $text .= " <th$attr[$n]>".$this->runSpanGamut(trim($header))."</th>\n"; 2626 $text .= "</tr>\n"; 2627 $text .= "</thead>\n"; 2628 2629 # Split content by row. 2630 $rows = explode("\n", trim($content, "\n")); 2631 2632 $text .= "<tbody>\n"; 2633 foreach ($rows as $row) { 2634 # Parsing span elements, including code spans, character escapes, 2635 # and inline HTML tags, so that pipes inside those gets ignored. 2636 $row = $this->parseSpan($row); 2637 2638 # Split row by cell. 2639 $row_cells = preg_split('/ *[|] */', $row, $col_count); 2640 $row_cells = array_pad($row_cells, $col_count, ''); 2641 2642 $text .= "<tr>\n"; 2643 foreach ($row_cells as $n => $cell) 2644 $text .= " <td$attr[$n]>".$this->runSpanGamut(trim($cell))."</td>\n"; 2645 $text .= "</tr>\n"; 2646 } 2647 $text .= "</tbody>\n"; 2648 $text .= "</table>"; 2649 2650 return $this->hashBlock($text) . "\n"; 2651 } 2652 2653 2654 protected function doDefLists($text) { 2655 # 2656 # Form HTML definition lists. 2657 # 2658 $less_than_tab = $this->tab_width - 1; 2659 2660 # Re-usable pattern to match any entire dl list: 2661 $whole_list_re = '(?> 2662 ( # $1 = whole list 2663 ( # $2 2664 [ ]{0,'.$less_than_tab.'} 2665 ((?>.*\S.*\n)+) # $3 = defined term 2666 \n? 2667 [ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition 2668 ) 2669 (?s:.+?) 2670 ( # $4 2671 \z 2672 | 2673 \n{2,} 2674 (?=\S) 2675 (?! # Negative lookahead for another term 2676 [ ]{0,'.$less_than_tab.'} 2677 (?: \S.*\n )+? # defined term 2678 \n? 2679 [ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition 2680 ) 2681 (?! # Negative lookahead for another definition 2682 [ ]{0,'.$less_than_tab.'}:[ ]+ # colon starting definition 2683 ) 2684 ) 2685 ) 2686 )'; // mx 2687 2688 $text = preg_replace_callback('{ 2689 (?>\A\n?|(?<=\n\n)) 2690 '.$whole_list_re.' 2691 }mx', 2692 array($this, '_doDefLists_callback'), $text); 2693 2694 return $text; 2695 } 2696 protected function _doDefLists_callback($matches) { 2697 # Re-usable patterns to match list item bullets and number markers: 2698 $list = $matches[1]; 2699 2700 # Turn double returns into triple returns, so that we can make a 2701 # paragraph for the last item in a list, if necessary: 2702 $result = trim($this->processDefListItems($list)); 2703 $result = "<dl>\n" . $result . "\n</dl>"; 2704 return $this->hashBlock($result) . "\n\n"; 2705 } 2706 2707 2708 protected function processDefListItems($list_str) { 2709 # 2710 # Process the contents of a single definition list, splitting it 2711 # into individual term and definition list items. 2712 # 2713 $less_than_tab = $this->tab_width - 1; 2714 2715 # trim trailing blank lines: 2716 $list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str); 2717 2718 # Process definition terms. 2719 $list_str = preg_replace_callback('{ 2720 (?>\A\n?|\n\n+) # leading line 2721 ( # definition terms = $1 2722 [ ]{0,'.$less_than_tab.'} # leading whitespace 2723 (?!\:[ ]|[ ]) # negative lookahead for a definition 2724 # mark (colon) or more whitespace. 2725 (?> \S.* \n)+? # actual term (not whitespace). 2726 ) 2727 (?=\n?[ ]{0,3}:[ ]) # lookahead for following line feed 2728 # with a definition mark. 2729 }xm', 2730 array($this, '_processDefListItems_callback_dt'), $list_str); 2731 2732 # Process actual definitions. 2733 $list_str = preg_replace_callback('{ 2734 \n(\n+)? # leading line = $1 2735 ( # marker space = $2 2736 [ ]{0,'.$less_than_tab.'} # whitespace before colon 2737 \:[ ]+ # definition mark (colon) 2738 ) 2739 ((?s:.+?)) # definition text = $3 2740 (?= \n+ # stop at next definition mark, 2741 (?: # next term or end of text 2742 [ ]{0,'.$less_than_tab.'} \:[ ] | 2743 <dt> | \z 2744 ) 2745 ) 2746 }xm', 2747 array($this, '_processDefListItems_callback_dd'), $list_str); 2748 2749 return $list_str; 2750 } 2751 protected function _processDefListItems_callback_dt($matches) { 2752 $terms = explode("\n", trim($matches[1])); 2753 $text = ''; 2754 foreach ($terms as $term) { 2755 $term = $this->runSpanGamut(trim($term)); 2756 $text .= "\n<dt>" . $term . "</dt>"; 2757 } 2758 return $text . "\n"; 2759 } 2760 protected function _processDefListItems_callback_dd($matches) { 2761 $leading_line = $matches[1]; 2762 $marker_space = $matches[2]; 2763 $def = $matches[3]; 2764 2765 if ($leading_line || preg_match('/\n{2,}/', $def)) { 2766 # Replace marker with the appropriate whitespace indentation 2767 $def = str_repeat(' ', strlen($marker_space)) . $def; 2768 $def = $this->runBlockGamut($this->outdent($def . "\n\n")); 2769 $def = "\n". $def ."\n"; 2770 } 2771 else { 2772 $def = rtrim($def); 2773 $def = $this->runSpanGamut($this->outdent($def)); 2774 } 2775 2776 return "\n<dd>" . $def . "</dd>\n"; 2777 } 2778 2779 2780 protected function doFencedCodeBlocks($text) { 2781 # 2782 # Adding the fenced code block syntax to regular Markdown: 2783 # 2784 # ~~~ 2785 # Code block 2786 # ~~~ 2787 # 2788 $less_than_tab = $this->tab_width; 2789 2790 $text = preg_replace_callback('{ 2791 (?:\n|\A) 2792 # 1: Opening marker 2793 ( 2794 (?:~{3,}|`{3,}) # 3 or more tildes/backticks. 2795 ) 2796 [ ]* 2797 (?: 2798 \.?([-_:a-zA-Z0-9]+) # 2: standalone class name 2799 | 2800 '.$this->id_class_attr_catch_re.' # 3: Extra attributes 2801 )? 2802 [ ]* \n # Whitespace and newline following marker. 2803 2804 # 4: Content 2805 ( 2806 (?> 2807 (?!\1 [ ]* \n) # Not a closing marker. 2808 .*\n+ 2809 )+ 2810 ) 2811 2812 # Closing marker. 2813 \1 [ ]* (?= \n ) 2814 }xm', 2815 array($this, '_doFencedCodeBlocks_callback'), $text); 2816 2817 return $text; 2818 } 2819 protected function _doFencedCodeBlocks_callback($matches) { 2820 $classname =& $matches[2]; 2821 $attrs =& $matches[3]; 2822 $codeblock = $matches[4]; 2823 $codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES); 2824 $codeblock = preg_replace_callback('/^\n+/', 2825 array($this, '_doFencedCodeBlocks_newlines'), $codeblock); 2826 2827 if ($classname != "") { 2828 if ($classname{0} == '.') 2829 $classname = substr($classname, 1); 2830 $attr_str = ' class="'.$this->code_class_prefix.$classname.'"'; 2831 } else { 2832 $attr_str = $this->doExtraAttributes($this->code_attr_on_pre ? "pre" : "code", $attrs); 2833 } 2834 $pre_attr_str = $this->code_attr_on_pre ? $attr_str : ''; 2835 $code_attr_str = $this->code_attr_on_pre ? '' : $attr_str; 2836 $codeblock = "<pre$pre_attr_str><code$code_attr_str>$codeblock</code></pre>"; 2837 2838 return "\n\n".$this->hashBlock($codeblock)."\n\n"; 2839 } 2840 protected function _doFencedCodeBlocks_newlines($matches) { 2841 return str_repeat("<br$this->empty_element_suffix", 2842 strlen($matches[0])); 2843 } 2844 2845 2846 # 2847 # Redefining emphasis markers so that emphasis by underscore does not 2848 # work in the middle of a word. 2849 # 2850 protected $em_relist = array( 2851 '' => '(?:(?<!\*)\*(?!\*)|(?<![a-zA-Z0-9_])_(?!_))(?![\.,:;]?\s)', 2852 '*' => '(?<![\s*])\*(?!\*)', 2853 '_' => '(?<![\s_])_(?![a-zA-Z0-9_])', 2854 ); 2855 protected $strong_relist = array( 2856 '' => '(?:(?<!\*)\*\*(?!\*)|(?<![a-zA-Z0-9_])__(?!_))(?![\.,:;]?\s)', 2857 '**' => '(?<![\s*])\*\*(?!\*)', 2858 '__' => '(?<![\s_])__(?![a-zA-Z0-9_])', 2859 ); 2860 protected $em_strong_relist = array( 2861 '' => '(?:(?<!\*)\*\*\*(?!\*)|(?<![a-zA-Z0-9_])___(?!_))(?![\.,:;]?\s)', 2862 '***' => '(?<![\s*])\*\*\*(?!\*)', 2863 '___' => '(?<![\s_])___(?![a-zA-Z0-9_])', 2864 ); 2865 2866 2867 protected function formParagraphs($text) { 2868 # 2869 # Params: 2870 # $text - string to process with html <p> tags 2871 # 2872 # Strip leading and trailing lines: 2873 $text = preg_replace('/\A\n+|\n+\z/', '', $text); 2874 2875 $grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY); 2876 2877 # 2878 # Wrap <p> tags and unhashify HTML blocks 2879 # 2880 foreach ($grafs as $key => $value) { 2881 $value = trim($this->runSpanGamut($value)); 2882 2883 # Check if this should be enclosed in a paragraph. 2884 # Clean tag hashes & block tag hashes are left alone. 2885 $is_p = !preg_match('/^B\x1A[0-9]+B|^C\x1A[0-9]+C$/', $value); 2886 2887 if ($is_p) { 2888 $value = "<p>$value</p>"; 2889 } 2890 $grafs[$key] = $value; 2891 } 2892 2893 # Join grafs in one text, then unhash HTML tags. 2894 $text = implode("\n\n", $grafs); 2895 2896 # Finish by removing any tag hashes still present in $text. 2897 $text = $this->unhash($text); 2898 2899 return $text; 2900 } 2901 2902 2903 ### Footnotes 2904 2905 protected function stripFootnotes($text) { 2906 # 2907 # Strips link definitions from text, stores the URLs and titles in 2908 # hash references. 2909 # 2910 $less_than_tab = $this->tab_width - 1; 2911 2912 # Link defs are in the form: [^id]: url "optional title" 2913 $text = preg_replace_callback('{ 2914 ^[ ]{0,'.$less_than_tab.'}\[\^(.+?)\][ ]?: # note_id = $1 2915 [ ]* 2916 \n? # maybe *one* newline 2917 ( # text = $2 (no blank lines allowed) 2918 (?: 2919 .+ # actual text 2920 | 2921 \n # newlines but 2922 (?!\[.+?\][ ]?:\s)# negative lookahead for footnote or link definition marker. 2923 (?!\n+[ ]{0,3}\S)# ensure line is not blank and followed 2924 # by non-indented content 2925 )* 2926 ) 2927 }xm', 2928 array($this, '_stripFootnotes_callback'), 2929 $text); 2930 return $text; 2931 } 2932 protected function _stripFootnotes_callback($matches) { 2933 $note_id = $this->fn_id_prefix . $matches[1]; 2934 $this->footnotes[$note_id] = $this->outdent($matches[2]); 2935 return ''; # String that will replace the block 2936 } 2937 2938 2939 protected function doFootnotes($text) { 2940 # 2941 # Replace footnote references in $text [^id] with a special text-token 2942 # which will be replaced by the actual footnote marker in appendFootnotes. 2943 # 2944 if (!$this->in_anchor) { 2945 $text = preg_replace('{\[\^(.+?)\]}', "F\x1Afn:\\1\x1A:", $text); 2946 } 2947 return $text; 2948 } 2949 2950 2951 protected function appendFootnotes($text) { 2952 # 2953 # Append footnote list to text. 2954 # 2955 $text = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', 2956 array($this, '_appendFootnotes_callback'), $text); 2957 2958 if (!empty($this->footnotes_ordered)) { 2959 $text .= "\n\n"; 2960 $text .= "<div class=\"footnotes\">\n"; 2961 $text .= "<hr". $this->empty_element_suffix ."\n"; 2962 $text .= "<ol>\n\n"; 2963 2964 $attr = ""; 2965 if ($this->fn_backlink_class != "") { 2966 $class = $this->fn_backlink_class; 2967 $class = $this->encodeAttribute($class); 2968 $attr .= " class=\"$class\""; 2969 } 2970 if ($this->fn_backlink_title != "") { 2971 $title = $this->fn_backlink_title; 2972 $title = $this->encodeAttribute($title); 2973 $attr .= " title=\"$title\""; 2974 } 2975 $num = 0; 2976 2977 while (!empty($this->footnotes_ordered)) { 2978 $footnote = reset($this->footnotes_ordered); 2979 $note_id = key($this->footnotes_ordered); 2980 unset($this->footnotes_ordered[$note_id]); 2981 $ref_count = $this->footnotes_ref_count[$note_id]; 2982 unset($this->footnotes_ref_count[$note_id]); 2983 unset($this->footnotes[$note_id]); 2984 2985 $footnote .= "\n"; # Need to append newline before parsing. 2986 $footnote = $this->runBlockGamut("$footnote\n"); 2987 $footnote = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', 2988 array($this, '_appendFootnotes_callback'), $footnote); 2989 2990 $attr = str_replace("%%", ++$num, $attr); 2991 $note_id = $this->encodeAttribute($note_id); 2992 2993 # Prepare backlink, multiple backlinks if multiple references 2994 $backlink = "<a href=\"#fnref:$note_id\"$attr>↩</a>"; 2995 for ($ref_num = 2; $ref_num <= $ref_count; ++$ref_num) { 2996 $backlink .= " <a href=\"#fnref$ref_num:$note_id\"$attr>↩</a>"; 2997 } 2998 # Add backlink to last paragraph; create new paragraph if needed. 2999 if (preg_match('{</p>$}', $footnote)) { 3000 $footnote = substr($footnote, 0, -4) . " $backlink</p>"; 3001 } else { 3002 $footnote .= "\n\n<p>$backlink</p>"; 3003 } 3004 3005 $text .= "<li id=\"fn:$note_id\">\n"; 3006 $text .= $footnote . "\n"; 3007 $text .= "</li>\n\n"; 3008 } 3009 3010 $text .= "</ol>\n"; 3011 $text .= "</div>"; 3012 } 3013 return $text; 3014 } 3015 protected function _appendFootnotes_callback($matches) { 3016 $node_id = $this->fn_id_prefix . $matches[1]; 3017 3018 # Create footnote marker only if it has a corresponding footnote *and* 3019 # the footnote hasn't been used by another marker. 3020 if (isset($this->footnotes[$node_id])) { 3021 $num =& $this->footnotes_numbers[$node_id]; 3022 if (!isset($num)) { 3023 # Transfer footnote content to the ordered list and give it its 3024 # number 3025 $this->footnotes_ordered[$node_id] = $this->footnotes[$node_id]; 3026 $this->footnotes_ref_count[$node_id] = 1; 3027 $num = $this->footnote_counter++; 3028 $ref_count_mark = ''; 3029 } else { 3030 $ref_count_mark = $this->footnotes_ref_count[$node_id] += 1; 3031 } 3032 3033 $attr = ""; 3034 if ($this->fn_link_class != "") { 3035 $class = $this->fn_link_class; 3036 $class = $this->encodeAttribute($class); 3037 $attr .= " class=\"$class\""; 3038 } 3039 if ($this->fn_link_title != "") { 3040 $title = $this->fn_link_title; 3041 $title = $this->encodeAttribute($title); 3042 $attr .= " title=\"$title\""; 3043 } 3044 3045 $attr = str_replace("%%", $num, $attr); 3046 $node_id = $this->encodeAttribute($node_id); 3047 3048 return 3049 "<sup id=\"fnref$ref_count_mark:$node_id\">". 3050 "<a href=\"#fn:$node_id\"$attr>$num</a>". 3051 "</sup>"; 3052 } 3053 3054 return "[^".$matches[1]."]"; 3055 } 3056 3057 3058 ### Abbreviations ### 3059 3060 protected function stripAbbreviations($text) { 3061 # 3062 # Strips abbreviations from text, stores titles in hash references. 3063 # 3064 $less_than_tab = $this->tab_width - 1; 3065 3066 # Link defs are in the form: [id]*: url "optional title" 3067 $text = preg_replace_callback('{ 3068 ^[ ]{0,'.$less_than_tab.'}\*\[(.+?)\][ ]?: # abbr_id = $1 3069 (.*) # text = $2 (no blank lines allowed) 3070 }xm', 3071 array($this, '_stripAbbreviations_callback'), 3072 $text); 3073 return $text; 3074 } 3075 protected function _stripAbbreviations_callback($matches) { 3076 $abbr_word = $matches[1]; 3077 $abbr_desc = $matches[2]; 3078 if ($this->abbr_word_re) 3079 $this->abbr_word_re .= '|'; 3080 $this->abbr_word_re .= preg_quote($abbr_word); 3081 $this->abbr_desciptions[$abbr_word] = trim($abbr_desc); 3082 return ''; # String that will replace the block 3083 } 3084 3085 3086 protected function doAbbreviations($text) { 3087 # 3088 # Find defined abbreviations in text and wrap them in <abbr> elements. 3089 # 3090 if ($this->abbr_word_re) { 3091 // cannot use the /x modifier because abbr_word_re may 3092 // contain significant spaces: 3093 $text = preg_replace_callback('{'. 3094 '(?<![\w\x1A])'. 3095 '(?:'.$this->abbr_word_re.')'. 3096 '(?![\w\x1A])'. 3097 '}', 3098 array($this, '_doAbbreviations_callback'), $text); 3099 } 3100 return $text; 3101 } 3102 protected function _doAbbreviations_callback($matches) { 3103 $abbr = $matches[0]; 3104 if (isset($this->abbr_desciptions[$abbr])) { 3105 $desc = $this->abbr_desciptions[$abbr]; 3106 if (empty($desc)) { 3107 return $this->hashPart("<abbr>$abbr</abbr>"); 3108 } else { 3109 $desc = $this->encodeAttribute($desc); 3110 return $this->hashPart("<abbr title=\"$desc\">$abbr</abbr>"); 3111 } 3112 } else { 3113 return $matches[0]; 3114 } 3115 } 3116 3117 }
title
Description
Body
title
Description
Body
title
Description
Body
title
Body
| Generated: Fri Nov 28 20:29:05 2014 | Cross-referenced by PHPXref 0.7.1 |