| 1 |
<?php |
| 2 |
/** |
| 3 |
* Markdown - A text-to-HTML conversion tool for web writers |
| 4 |
* |
| 5 |
* @package php-markdown |
| 6 |
* @author Michel Fortin <michel.fortin@michelf.com> |
| 7 |
* @copyright 2004-2022 Michel Fortin <https://michelf.com/projects/php-markdown/> |
| 8 |
* @copyright (Original Markdown) 2004-2006 John Gruber <https://daringfireball.net/projects/markdown/> |
| 9 |
*/ |
| 10 |
|
| 11 |
namespace Michelf; |
| 12 |
|
| 13 |
/** |
| 14 |
* Markdown Parser Class |
| 15 |
*/ |
| 16 |
class Markdown implements MarkdownInterface { |
| 17 |
/** |
| 18 |
* Define the package version |
| 19 |
* @var string |
| 20 |
*/ |
| 21 |
const MARKDOWNLIB_VERSION = "2.0.0"; |
| 22 |
|
| 23 |
/** |
| 24 |
* Simple function interface - Initialize the parser and return the result |
| 25 |
* of its transform method. This will work fine for derived classes too. |
| 26 |
* |
| 27 |
* @api |
| 28 |
* |
| 29 |
* @param string $text |
| 30 |
* @return string |
| 31 |
*/ |
| 32 |
public static function defaultTransform(string $text): string { |
| 33 |
// Take parser class on which this function was called. |
| 34 |
$parser_class = static::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 |
|
| 45 |
// Transform text using parser. |
| 46 |
return $parser->transform($text); |
| 47 |
} |
| 48 |
|
| 49 |
/** |
| 50 |
* Configuration variables |
| 51 |
*/ |
| 52 |
/** |
| 53 |
* Change to ">" for HTML output. |
| 54 |
*/ |
| 55 |
public string $empty_element_suffix = " />"; |
| 56 |
|
| 57 |
/** |
| 58 |
* The width of indentation of the output markup |
| 59 |
*/ |
| 60 |
public int $tab_width = 4; |
| 61 |
|
| 62 |
/** |
| 63 |
* Change to `true` to disallow markup or entities. |
| 64 |
*/ |
| 65 |
public bool $no_markup = false; |
| 66 |
public bool $no_entities = false; |
| 67 |
|
| 68 |
|
| 69 |
/** |
| 70 |
* Change to `true` to enable line breaks on \n without two trailling spaces |
| 71 |
* @var boolean |
| 72 |
*/ |
| 73 |
public bool $hard_wrap = false; |
| 74 |
|
| 75 |
/** |
| 76 |
* Predefined URLs and titles for reference links and images. |
| 77 |
*/ |
| 78 |
public array $predef_urls = array(); |
| 79 |
public array $predef_titles = array(); |
| 80 |
|
| 81 |
/** |
| 82 |
* Optional filter function for URLs |
| 83 |
* @var callable|null |
| 84 |
*/ |
| 85 |
public $url_filter_func = null; |
| 86 |
|
| 87 |
/** |
| 88 |
* Optional header id="" generation callback function. |
| 89 |
* @var callable|null |
| 90 |
*/ |
| 91 |
public $header_id_func = null; |
| 92 |
|
| 93 |
/** |
| 94 |
* Optional function for converting code block content to HTML |
| 95 |
* @var callable|null |
| 96 |
*/ |
| 97 |
public $code_block_content_func = null; |
| 98 |
|
| 99 |
/** |
| 100 |
* Optional function for converting code span content to HTML. |
| 101 |
* @var callable|null |
| 102 |
*/ |
| 103 |
public $code_span_content_func = null; |
| 104 |
|
| 105 |
/** |
| 106 |
* Class attribute to toggle "enhanced ordered list" behaviour |
| 107 |
* setting this to true will allow ordered lists to start from the index |
| 108 |
* number that is defined first. |
| 109 |
* |
| 110 |
* For example: |
| 111 |
* 2. List item two |
| 112 |
* 3. List item three |
| 113 |
* |
| 114 |
* Becomes: |
| 115 |
* <ol start="2"> |
| 116 |
* <li>List item two</li> |
| 117 |
* <li>List item three</li> |
| 118 |
* </ol> |
| 119 |
*/ |
| 120 |
public bool $enhanced_ordered_list = false; |
| 121 |
|
| 122 |
/** |
| 123 |
* Parser implementation |
| 124 |
*/ |
| 125 |
/** |
| 126 |
* Regex to match balanced [brackets]. |
| 127 |
* Needed to insert a maximum bracked depth while converting to PHP. |
| 128 |
*/ |
| 129 |
protected int $nested_brackets_depth = 6; |
| 130 |
protected string $nested_brackets_re; |
| 131 |
|
| 132 |
protected int $nested_url_parenthesis_depth = 4; |
| 133 |
protected string $nested_url_parenthesis_re; |
| 134 |
|
| 135 |
/** |
| 136 |
* Table of hash values for escaped characters: |
| 137 |
*/ |
| 138 |
protected string $escape_chars = '\`*_{}[]()>#+-.!'; |
| 139 |
protected string $escape_chars_re; |
| 140 |
|
| 141 |
/** |
| 142 |
* Constructor function. Initialize appropriate member variables. |
| 143 |
* @return void |
| 144 |
*/ |
| 145 |
public function __construct() { |
| 146 |
$this->_initDetab(); |
| 147 |
$this->prepareItalicsAndBold(); |
| 148 |
|
| 149 |
$this->nested_brackets_re = |
| 150 |
str_repeat('(?>[^\[\]]+|\[', $this->nested_brackets_depth). |
| 151 |
str_repeat('\])*', $this->nested_brackets_depth); |
| 152 |
|
| 153 |
$this->nested_url_parenthesis_re = |
| 154 |
str_repeat('(?>[^()\s]+|\(', $this->nested_url_parenthesis_depth). |
| 155 |
str_repeat('(?>\)))*', $this->nested_url_parenthesis_depth); |
| 156 |
|
| 157 |
$this->escape_chars_re = '['.preg_quote($this->escape_chars).']'; |
| 158 |
|
| 159 |
// Sort document, block, and span gamut in ascendent priority order. |
| 160 |
asort($this->document_gamut); |
| 161 |
asort($this->block_gamut); |
| 162 |
asort($this->span_gamut); |
| 163 |
} |
| 164 |
|
| 165 |
|
| 166 |
/** |
| 167 |
* Internal hashes used during transformation. |
| 168 |
*/ |
| 169 |
protected array $urls = array(); |
| 170 |
protected array $titles = array(); |
| 171 |
protected array $html_hashes = array(); |
| 172 |
|
| 173 |
/** |
| 174 |
* Status flag to avoid invalid nesting. |
| 175 |
*/ |
| 176 |
protected bool $in_anchor = false; |
| 177 |
|
| 178 |
/** |
| 179 |
* Status flag to avoid invalid nesting. |
| 180 |
*/ |
| 181 |
protected bool $in_emphasis_processing = false; |
| 182 |
|
| 183 |
/** |
| 184 |
* Called before the transformation process starts to setup parser states. |
| 185 |
* @return void |
| 186 |
*/ |
| 187 |
protected function setup() { |
| 188 |
// Clear global hashes. |
| 189 |
$this->urls = $this->predef_urls; |
| 190 |
$this->titles = $this->predef_titles; |
| 191 |
$this->html_hashes = array(); |
| 192 |
$this->in_anchor = false; |
| 193 |
$this->in_emphasis_processing = false; |
| 194 |
} |
| 195 |
|
| 196 |
/** |
| 197 |
* Called after the transformation process to clear any variable which may |
| 198 |
* be taking up memory unnecessarly. |
| 199 |
* @return void |
| 200 |
*/ |
| 201 |
protected function teardown() { |
| 202 |
$this->urls = array(); |
| 203 |
$this->titles = array(); |
| 204 |
$this->html_hashes = array(); |
| 205 |
} |
| 206 |
|
| 207 |
/** |
| 208 |
* Main function. Performs some preprocessing on the input text and pass |
| 209 |
* it through the document gamut. |
| 210 |
* |
| 211 |
* @api |
| 212 |
* |
| 213 |
* @param string $text |
| 214 |
* @return string |
| 215 |
*/ |
| 216 |
public function transform(string $text): string { |
| 217 |
$this->setup(); |
| 218 |
|
| 219 |
# Remove UTF-8 BOM and marker character in input, if present. |
| 220 |
$text = preg_replace('{^\xEF\xBB\xBF|\x1A}', '', $text); |
| 221 |
|
| 222 |
# Standardize line endings: |
| 223 |
# DOS to Unix and Mac to Unix |
| 224 |
$text = preg_replace('{\r\n?}', "\n", $text); |
| 225 |
|
| 226 |
# Make sure $text ends with a couple of newlines: |
| 227 |
$text .= "\n\n"; |
| 228 |
|
| 229 |
# Convert all tabs to spaces. |
| 230 |
$text = $this->detab($text); |
| 231 |
|
| 232 |
# Turn block-level HTML blocks into hash entries |
| 233 |
$text = $this->hashHTMLBlocks($text); |
| 234 |
|
| 235 |
# Strip any lines consisting only of spaces and tabs. |
| 236 |
# This makes subsequent regexen easier to write, because we can |
| 237 |
# match consecutive blank lines with /\n+/ instead of something |
| 238 |
# contorted like /[ ]*\n+/ . |
| 239 |
$text = preg_replace('/^[ ]+$/m', '', $text); |
| 240 |
|
| 241 |
# Run document gamut methods. |
| 242 |
foreach ($this->document_gamut as $method => $priority) { |
| 243 |
$text = $this->$method($text); |
| 244 |
} |
| 245 |
|
| 246 |
$this->teardown(); |
| 247 |
|
| 248 |
return $text . "\n"; |
| 249 |
} |
| 250 |
|
| 251 |
/** |
| 252 |
* Define the document gamut |
| 253 |
*/ |
| 254 |
protected array $document_gamut = array( |
| 255 |
// Strip link definitions, store in hashes. |
| 256 |
"stripLinkDefinitions" => 20, |
| 257 |
"runBasicBlockGamut" => 30, |
| 258 |
); |
| 259 |
|
| 260 |
/** |
| 261 |
* Strips link definitions from text, stores the URLs and titles in |
| 262 |
* hash references |
| 263 |
* @param string $text |
| 264 |
* @return string |
| 265 |
*/ |
| 266 |
protected function stripLinkDefinitions($text) { |
| 267 |
|
| 268 |
$less_than_tab = $this->tab_width - 1; |
| 269 |
|
| 270 |
// Link defs are in the form: ^[id]: url "optional title" |
| 271 |
$text = preg_replace_callback('{ |
| 272 |
^[ ]{0,'.$less_than_tab.'}\[(.+)\][ ]?: # id = $1 |
| 273 |
[ ]* |
| 274 |
\n? # maybe *one* newline |
| 275 |
[ ]* |
| 276 |
(?: |
| 277 |
<(.+?)> # url = $2 |
| 278 |
| |
| 279 |
(\S+?) # url = $3 |
| 280 |
) |
| 281 |
[ ]* |
| 282 |
\n? # maybe one newline |
| 283 |
[ ]* |
| 284 |
(?: |
| 285 |
(?<=\s) # lookbehind for whitespace |
| 286 |
["(] |
| 287 |
(.*?) # title = $4 |
| 288 |
[")] |
| 289 |
[ ]* |
| 290 |
)? # title is optional |
| 291 |
(?:\n+|\Z) |
| 292 |
}xm', |
| 293 |
array($this, '_stripLinkDefinitions_callback'), |
| 294 |
$text |
| 295 |
); |
| 296 |
return $text; |
| 297 |
} |
| 298 |
|
| 299 |
/** |
| 300 |
* The callback to strip link definitions |
| 301 |
* @param array $matches |
| 302 |
* @return string |
| 303 |
*/ |
| 304 |
protected function _stripLinkDefinitions_callback($matches) { |
| 305 |
$link_id = strtolower($matches[1]); |
| 306 |
$url = $matches[2] == '' ? $matches[3] : $matches[2]; |
| 307 |
$this->urls[$link_id] = $url; |
| 308 |
$this->titles[$link_id] =& $matches[4]; |
| 309 |
return ''; // String that will replace the block |
| 310 |
} |
| 311 |
|
| 312 |
/** |
| 313 |
* Hashify HTML blocks |
| 314 |
* @param string $text |
| 315 |
* @return string |
| 316 |
*/ |
| 317 |
protected function hashHTMLBlocks($text) { |
| 318 |
if ($this->no_markup) { |
| 319 |
return $text; |
| 320 |
} |
| 321 |
|
| 322 |
$less_than_tab = $this->tab_width - 1; |
| 323 |
|
| 324 |
/** |
| 325 |
* Hashify HTML blocks: |
| 326 |
* |
| 327 |
* We only want to do this for block-level HTML tags, such as headers, |
| 328 |
* lists, and tables. That's because we still want to wrap <p>s around |
| 329 |
* "paragraphs" that are wrapped in non-block-level tags, such as |
| 330 |
* anchors, phrase emphasis, and spans. The list of tags we're looking |
| 331 |
* for is hard-coded: |
| 332 |
* |
| 333 |
* * List "a" is made of tags which can be both inline or block-level. |
| 334 |
* These will be treated block-level when the start tag is alone on |
| 335 |
* its line, otherwise they're not matched here and will be taken as |
| 336 |
* inline later. |
| 337 |
* * List "b" is made of tags which are always block-level; |
| 338 |
*/ |
| 339 |
$block_tags_a_re = 'ins|del'; |
| 340 |
$block_tags_b_re = 'p|div|h[1-6]|blockquote|pre|table|dl|ol|ul|address|'. |
| 341 |
'script|noscript|style|form|fieldset|iframe|math|svg|'. |
| 342 |
'article|section|nav|aside|hgroup|header|footer|'. |
| 343 |
'figure|details|summary'; |
| 344 |
|
| 345 |
// Regular expression for the content of a block tag. |
| 346 |
$nested_tags_level = 4; |
| 347 |
$attr = ' |
| 348 |
(?> # optional tag attributes |
| 349 |
\s # starts with whitespace |
| 350 |
(?> |
| 351 |
[^>"/]+ # text outside quotes |
| 352 |
| |
| 353 |
/+(?!>) # slash not followed by ">" |
| 354 |
| |
| 355 |
"[^"]*" # text inside double quotes (tolerate ">") |
| 356 |
| |
| 357 |
\'[^\']*\' # text inside single quotes (tolerate ">") |
| 358 |
)* |
| 359 |
)? |
| 360 |
'; |
| 361 |
$content = |
| 362 |
str_repeat(' |
| 363 |
(?> |
| 364 |
[^<]+ # content without tag |
| 365 |
| |
| 366 |
<\2 # nested opening tag |
| 367 |
'.$attr.' # attributes |
| 368 |
(?> |
| 369 |
/> |
| 370 |
| |
| 371 |
>', $nested_tags_level). // end of opening tag |
| 372 |
'.*?'. // last level nested tag content |
| 373 |
str_repeat(' |
| 374 |
</\2\s*> # closing nested tag |
| 375 |
) |
| 376 |
| |
| 377 |
<(?!/\2\s*> # other tags with a different name |
| 378 |
) |
| 379 |
)*', |
| 380 |
$nested_tags_level); |
| 381 |
$content2 = str_replace('\2', '\3', $content); |
| 382 |
|
| 383 |
/** |
| 384 |
* First, look for nested blocks, e.g.: |
| 385 |
* <div> |
| 386 |
* <div> |
| 387 |
* tags for inner block must be indented. |
| 388 |
* </div> |
| 389 |
* </div> |
| 390 |
* |
| 391 |
* The outermost tags must start at the left margin for this to match, |
| 392 |
* and the inner nested divs must be indented. |
| 393 |
* We need to do this before the next, more liberal match, because the |
| 394 |
* next match will start at the first `<div>` and stop at the |
| 395 |
* first `</div>`. |
| 396 |
*/ |
| 397 |
$text = preg_replace_callback('{(?> |
| 398 |
(?> |
| 399 |
(?<=\n) # Starting on its own line |
| 400 |
| # or |
| 401 |
\A\n? # the at beginning of the doc |
| 402 |
) |
| 403 |
( # save in $1 |
| 404 |
|
| 405 |
# Match from `\n<tag>` to `</tag>\n`, handling nested tags |
| 406 |
# in between. |
| 407 |
|
| 408 |
[ ]{0,'.$less_than_tab.'} |
| 409 |
<('.$block_tags_b_re.')# start tag = $2 |
| 410 |
'.$attr.'> # attributes followed by > and \n |
| 411 |
'.$content.' # content, support nesting |
| 412 |
</\2> # the matching end tag |
| 413 |
[ ]* # trailing spaces/tabs |
| 414 |
(?=\n+|\Z) # followed by a newline or end of document |
| 415 |
|
| 416 |
| # Special version for tags of group a. |
| 417 |
|
| 418 |
[ ]{0,'.$less_than_tab.'} |
| 419 |
<('.$block_tags_a_re.')# start tag = $3 |
| 420 |
'.$attr.'>[ ]*\n # attributes followed by > |
| 421 |
'.$content2.' # content, support nesting |
| 422 |
</\3> # the matching end tag |
| 423 |
[ ]* # trailing spaces/tabs |
| 424 |
(?=\n+|\Z) # followed by a newline or end of document |
| 425 |
|
| 426 |
| # Special case just for <hr />. It was easier to make a special |
| 427 |
# case than to make the other regex more complicated. |
| 428 |
|
| 429 |
[ ]{0,'.$less_than_tab.'} |
| 430 |
<(hr) # start tag = $2 |
| 431 |
'.$attr.' # attributes |
| 432 |
/?> # the matching end tag |
| 433 |
[ ]* |
| 434 |
(?=\n{2,}|\Z) # followed by a blank line or end of document |
| 435 |
|
| 436 |
| # Special case for standalone HTML comments: |
| 437 |
|
| 438 |
[ ]{0,'.$less_than_tab.'} |
| 439 |
(?s: |
| 440 |
<!-- .*? --> |
| 441 |
) |
| 442 |
[ ]* |
| 443 |
(?=\n{2,}|\Z) # followed by a blank line or end of document |
| 444 |
|
| 445 |
| # PHP and ASP-style processor instructions (<? and <%) |
| 446 |
|
| 447 |
[ ]{0,'.$less_than_tab.'} |
| 448 |
(?s: |
| 449 |
<([?%]) # $2 |
| 450 |
.*? |
| 451 |
\2> |
| 452 |
) |
| 453 |
[ ]* |
| 454 |
(?=\n{2,}|\Z) # followed by a blank line or end of document |
| 455 |
|
| 456 |
) |
| 457 |
)}Sxmi', |
| 458 |
array($this, '_hashHTMLBlocks_callback'), |
| 459 |
$text |
| 460 |
); |
| 461 |
|
| 462 |
return $text; |
| 463 |
} |
| 464 |
|
| 465 |
/** |
| 466 |
* The callback for hashing HTML blocks |
| 467 |
* @param string $matches |
| 468 |
* @return string |
| 469 |
*/ |
| 470 |
protected function _hashHTMLBlocks_callback($matches) { |
| 471 |
$text = $matches[1]; |
| 472 |
$key = $this->hashBlock($text); |
| 473 |
return "\n\n$key\n\n"; |
| 474 |
} |
| 475 |
|
| 476 |
/** |
| 477 |
* Called whenever a tag must be hashed when a function insert an atomic |
| 478 |
* element in the text stream. Passing $text to through this function gives |
| 479 |
* a unique text-token which will be reverted back when calling unhash. |
| 480 |
* |
| 481 |
* The $boundary argument specify what character should be used to surround |
| 482 |
* the token. By convension, "B" is used for block elements that needs not |
| 483 |
* to be wrapped into paragraph tags at the end, ":" is used for elements |
| 484 |
* that are word separators and "X" is used in the general case. |
| 485 |
* |
| 486 |
* @param string $text |
| 487 |
* @param string $boundary |
| 488 |
* @return string |
| 489 |
*/ |
| 490 |
protected function hashPart($text, $boundary = 'X') { |
| 491 |
// Swap back any tag hash found in $text so we do not have to `unhash` |
| 492 |
// multiple times at the end. |
| 493 |
$text = $this->unhash($text); |
| 494 |
|
| 495 |
// Then hash the block. |
| 496 |
static $i = 0; |
| 497 |
$key = "$boundary\x1A" . ++$i . $boundary; |
| 498 |
$this->html_hashes[$key] = $text; |
| 499 |
return $key; // String that will replace the tag. |
| 500 |
} |
| 501 |
|
| 502 |
/** |
| 503 |
* Shortcut function for hashPart with block-level boundaries. |
| 504 |
* @param string $text |
| 505 |
* @return string |
| 506 |
*/ |
| 507 |
protected function hashBlock($text) { |
| 508 |
return $this->hashPart($text, 'B'); |
| 509 |
} |
| 510 |
|
| 511 |
/** |
| 512 |
* Define the block gamut - these are all the transformations that form |
| 513 |
* block-level tags like paragraphs, headers, and list items. |
| 514 |
*/ |
| 515 |
protected array $block_gamut = array( |
| 516 |
"doHeaders" => 10, |
| 517 |
"doHorizontalRules" => 20, |
| 518 |
"doLists" => 40, |
| 519 |
"doCodeBlocks" => 50, |
| 520 |
"doBlockQuotes" => 60, |
| 521 |
); |
| 522 |
|
| 523 |
/** |
| 524 |
* Run block gamut tranformations. |
| 525 |
* |
| 526 |
* We need to escape raw HTML in Markdown source before doing anything |
| 527 |
* else. This need to be done for each block, and not only at the |
| 528 |
* begining in the Markdown function since hashed blocks can be part of |
| 529 |
* list items and could have been indented. Indented blocks would have |
| 530 |
* been seen as a code block in a previous pass of hashHTMLBlocks. |
| 531 |
* |
| 532 |
* @param string $text |
| 533 |
* @return string |
| 534 |
*/ |
| 535 |
protected function runBlockGamut($text) { |
| 536 |
$text = $this->hashHTMLBlocks($text); |
| 537 |
return $this->runBasicBlockGamut($text); |
| 538 |
} |
| 539 |
|
| 540 |
/** |
| 541 |
* Run block gamut tranformations, without hashing HTML blocks. This is |
| 542 |
* useful when HTML blocks are known to be already hashed, like in the first |
| 543 |
* whole-document pass. |
| 544 |
* |
| 545 |
* @param string $text |
| 546 |
* @return string |
| 547 |
*/ |
| 548 |
protected function runBasicBlockGamut($text) { |
| 549 |
|
| 550 |
foreach ($this->block_gamut as $method => $priority) { |
| 551 |
$text = $this->$method($text); |
| 552 |
} |
| 553 |
|
| 554 |
// Finally form paragraph and restore hashed blocks. |
| 555 |
$text = $this->formParagraphs($text); |
| 556 |
|
| 557 |
return $text; |
| 558 |
} |
| 559 |
|
| 560 |
/** |
| 561 |
* Convert horizontal rules |
| 562 |
* @param string $text |
| 563 |
* @return string |
| 564 |
*/ |
| 565 |
protected function doHorizontalRules($text) { |
| 566 |
return preg_replace( |
| 567 |
'{ |
| 568 |
^[ ]{0,3} # Leading space |
| 569 |
([-*_]) # $1: First marker |
| 570 |
(?> # Repeated marker group |
| 571 |
[ ]{0,2} # Zero, one, or two spaces. |
| 572 |
\1 # Marker character |
| 573 |
){2,} # Group repeated at least twice |
| 574 |
[ ]* # Tailing spaces |
| 575 |
$ # End of line. |
| 576 |
}mx', |
| 577 |
"\n".$this->hashBlock("<hr$this->empty_element_suffix")."\n", |
| 578 |
$text |
| 579 |
); |
| 580 |
} |
| 581 |
|
| 582 |
/** |
| 583 |
* These are all the transformations that occur *within* block-level |
| 584 |
* tags like paragraphs, headers, and list items. |
| 585 |
*/ |
| 586 |
protected array $span_gamut = array( |
| 587 |
// Process character escapes, code spans, and inline HTML |
| 588 |
// in one shot. |
| 589 |
"parseSpan" => -30, |
| 590 |
// Process anchor and image tags. Images must come first, |
| 591 |
// because ![foo][f] looks like an anchor. |
| 592 |
"doImages" => 10, |
| 593 |
"doAnchors" => 20, |
| 594 |
// Make links out of things like `<https://example.com/>` |
| 595 |
// Must come after doAnchors, because you can use < and > |
| 596 |
// delimiters in inline links like [this](<url>). |
| 597 |
"doAutoLinks" => 30, |
| 598 |
"encodeAmpsAndAngles" => 40, |
| 599 |
"doItalicsAndBold" => 50, |
| 600 |
"doHardBreaks" => 60, |
| 601 |
); |
| 602 |
|
| 603 |
/** |
| 604 |
* Run span gamut transformations |
| 605 |
* @param string $text |
| 606 |
* @return string |
| 607 |
*/ |
| 608 |
protected function runSpanGamut($text) { |
| 609 |
foreach ($this->span_gamut as $method => $priority) { |
| 610 |
$text = $this->$method($text); |
| 611 |
} |
| 612 |
|
| 613 |
return $text; |
| 614 |
} |
| 615 |
|
| 616 |
/** |
| 617 |
* Do hard breaks |
| 618 |
* @param string $text |
| 619 |
* @return string |
| 620 |
*/ |
| 621 |
protected function doHardBreaks($text) { |
| 622 |
if ($this->hard_wrap) { |
| 623 |
return preg_replace_callback('/ *\n/', |
| 624 |
array($this, '_doHardBreaks_callback'), $text); |
| 625 |
} else { |
| 626 |
return preg_replace_callback('/ {2,}\n/', |
| 627 |
array($this, '_doHardBreaks_callback'), $text); |
| 628 |
} |
| 629 |
} |
| 630 |
|
| 631 |
/** |
| 632 |
* Trigger part hashing for the hard break (callback method) |
| 633 |
* @param array $matches |
| 634 |
* @return string |
| 635 |
*/ |
| 636 |
protected function _doHardBreaks_callback($matches) { |
| 637 |
return $this->hashPart("<br$this->empty_element_suffix\n"); |
| 638 |
} |
| 639 |
|
| 640 |
/** |
| 641 |
* Turn Markdown link shortcuts into XHTML <a> tags. |
| 642 |
* @param string $text |
| 643 |
* @return string |
| 644 |
*/ |
| 645 |
protected function doAnchors($text) { |
| 646 |
if ($this->in_anchor) { |
| 647 |
return $text; |
| 648 |
} |
| 649 |
$this->in_anchor = true; |
| 650 |
|
| 651 |
// First, handle reference-style links: [link text] [id] |
| 652 |
$text = preg_replace_callback('{ |
| 653 |
( # wrap whole match in $1 |
| 654 |
\[ |
| 655 |
('.$this->nested_brackets_re.') # link text = $2 |
| 656 |
\] |
| 657 |
|
| 658 |
[ ]? # one optional space |
| 659 |
(?:\n[ ]*)? # one optional newline followed by spaces |
| 660 |
|
| 661 |
\[ |
| 662 |
(.*?) # id = $3 |
| 663 |
\] |
| 664 |
) |
| 665 |
}xs', |
| 666 |
array($this, '_doAnchors_reference_callback'), $text); |
| 667 |
|
| 668 |
// Next, inline-style links: [link text](url "optional title") |
| 669 |
$text = preg_replace_callback('{ |
| 670 |
( # wrap whole match in $1 |
| 671 |
\[ |
| 672 |
('.$this->nested_brackets_re.') # link text = $2 |
| 673 |
\] |
| 674 |
\( # literal paren |
| 675 |
[ \n]* |
| 676 |
(?: |
| 677 |
<(.+?)> # href = $3 |
| 678 |
| |
| 679 |
('.$this->nested_url_parenthesis_re.') # href = $4 |
| 680 |
) |
| 681 |
[ \n]* |
| 682 |
( # $5 |
| 683 |
([\'"]) # quote char = $6 |
| 684 |
(.*?) # Title = $7 |
| 685 |
\6 # matching quote |
| 686 |
[ \n]* # ignore any spaces/tabs between closing quote and ) |
| 687 |
)? # title is optional |
| 688 |
\) |
| 689 |
) |
| 690 |
}xs', |
| 691 |
array($this, '_doAnchors_inline_callback'), $text); |
| 692 |
|
| 693 |
// Last, handle reference-style shortcuts: [link text] |
| 694 |
// These must come last in case you've also got [link text][1] |
| 695 |
// or [link text](/foo) |
| 696 |
$text = preg_replace_callback('{ |
| 697 |
( # wrap whole match in $1 |
| 698 |
\[ |
| 699 |
([^\[\]]+) # link text = $2; can\'t contain [ or ] |
| 700 |
\] |
| 701 |
) |
| 702 |
}xs', |
| 703 |
array($this, '_doAnchors_reference_callback'), $text); |
| 704 |
|
| 705 |
$this->in_anchor = false; |
| 706 |
return $text; |
| 707 |
} |
| 708 |
|
| 709 |
/** |
| 710 |
* Callback method to parse referenced anchors |
| 711 |
* @param array $matches |
| 712 |
* @return string |
| 713 |
*/ |
| 714 |
protected function _doAnchors_reference_callback($matches) { |
| 715 |
$whole_match = $matches[1]; |
| 716 |
$link_text = $matches[2]; |
| 717 |
$link_id =& $matches[3]; |
| 718 |
|
| 719 |
if ($link_id == "") { |
| 720 |
// for shortcut links like [this][] or [this]. |
| 721 |
$link_id = $link_text; |
| 722 |
} |
| 723 |
|
| 724 |
// lower-case and turn embedded newlines into spaces |
| 725 |
$link_id = strtolower($link_id); |
| 726 |
$link_id = preg_replace('{[ ]?\n}', ' ', $link_id); |
| 727 |
|
| 728 |
if (isset($this->urls[$link_id])) { |
| 729 |
$url = $this->urls[$link_id]; |
| 730 |
$url = $this->encodeURLAttribute($url); |
| 731 |
|
| 732 |
$result = "<a href=\"$url\""; |
| 733 |
if ( isset( $this->titles[$link_id] ) ) { |
| 734 |
$title = $this->titles[$link_id]; |
| 735 |
$title = $this->encodeAttribute($title); |
| 736 |
$result .= " title=\"$title\""; |
| 737 |
} |
| 738 |
|
| 739 |
$link_text = $this->runSpanGamut($link_text); |
| 740 |
$result .= ">$link_text</a>"; |
| 741 |
$result = $this->hashPart($result); |
| 742 |
} else { |
| 743 |
$result = $whole_match; |
| 744 |
} |
| 745 |
return $result; |
| 746 |
} |
| 747 |
|
| 748 |
/** |
| 749 |
* Callback method to parse inline anchors |
| 750 |
* @param array $matches |
| 751 |
* @return string |
| 752 |
*/ |
| 753 |
protected function _doAnchors_inline_callback($matches) { |
| 754 |
$link_text = $this->runSpanGamut($matches[2]); |
| 755 |
$url = $matches[3] === '' ? $matches[4] : $matches[3]; |
| 756 |
$title =& $matches[7]; |
| 757 |
|
| 758 |
// If the URL was of the form <s p a c e s> it got caught by the HTML |
| 759 |
// tag parser and hashed. Need to reverse the process before using |
| 760 |
// the URL. |
| 761 |
$unhashed = $this->unhash($url); |
| 762 |
if ($unhashed !== $url) |
| 763 |
$url = preg_replace('/^<(.*)>$/', '\1', $unhashed); |
| 764 |
|
| 765 |
$url = $this->encodeURLAttribute($url); |
| 766 |
|
| 767 |
$result = "<a href=\"$url\""; |
| 768 |
if ($title) { |
| 769 |
$title = $this->encodeAttribute($title); |
| 770 |
$result .= " title=\"$title\""; |
| 771 |
} |
| 772 |
|
| 773 |
$link_text = $this->runSpanGamut($link_text); |
| 774 |
$result .= ">$link_text</a>"; |
| 775 |
|
| 776 |
return $this->hashPart($result); |
| 777 |
} |
| 778 |
|
| 779 |
/** |
| 780 |
* Turn Markdown image shortcuts into <img> tags. |
| 781 |
* @param string $text |
| 782 |
* @return string |
| 783 |
*/ |
| 784 |
protected function doImages($text) { |
| 785 |
// First, handle reference-style labeled images: ![alt text][id] |
| 786 |
$text = preg_replace_callback('{ |
| 787 |
( # wrap whole match in $1 |
| 788 |
!\[ |
| 789 |
('.$this->nested_brackets_re.') # alt text = $2 |
| 790 |
\] |
| 791 |
|
| 792 |
[ ]? # one optional space |
| 793 |
(?:\n[ ]*)? # one optional newline followed by spaces |
| 794 |
|
| 795 |
\[ |
| 796 |
(.*?) # id = $3 |
| 797 |
\] |
| 798 |
|
| 799 |
) |
| 800 |
}xs', |
| 801 |
array($this, '_doImages_reference_callback'), $text); |
| 802 |
|
| 803 |
// Next, handle inline images:  |
| 804 |
// Don't forget: encode * and _ |
| 805 |
$text = preg_replace_callback('{ |
| 806 |
( # wrap whole match in $1 |
| 807 |
!\[ |
| 808 |
('.$this->nested_brackets_re.') # alt text = $2 |
| 809 |
\] |
| 810 |
\s? # One optional whitespace character |
| 811 |
\( # literal paren |
| 812 |
[ \n]* |
| 813 |
(?: |
| 814 |
<(\S*)> # src url = $3 |
| 815 |
| |
| 816 |
('.$this->nested_url_parenthesis_re.') # src url = $4 |
| 817 |
) |
| 818 |
[ \n]* |
| 819 |
( # $5 |
| 820 |
([\'"]) # quote char = $6 |
| 821 |
(.*?) # title = $7 |
| 822 |
\6 # matching quote |
| 823 |
[ \n]* |
| 824 |
)? # title is optional |
| 825 |
\) |
| 826 |
) |
| 827 |
}xs', |
| 828 |
array($this, '_doImages_inline_callback'), $text); |
| 829 |
|
| 830 |
return $text; |
| 831 |
} |
| 832 |
|
| 833 |
/** |
| 834 |
* Callback to parse references image tags |
| 835 |
* @param array $matches |
| 836 |
* @return string |
| 837 |
*/ |
| 838 |
protected function _doImages_reference_callback($matches) { |
| 839 |
$whole_match = $matches[1]; |
| 840 |
$alt_text = $matches[2]; |
| 841 |
$link_id = strtolower($matches[3]); |
| 842 |
|
| 843 |
if ($link_id == "") { |
| 844 |
$link_id = strtolower($alt_text); // for shortcut links like ![this][]. |
| 845 |
} |
| 846 |
|
| 847 |
$alt_text = $this->encodeAttribute($alt_text); |
| 848 |
if (isset($this->urls[$link_id])) { |
| 849 |
$url = $this->encodeURLAttribute($this->urls[$link_id]); |
| 850 |
$result = "<img src=\"$url\" alt=\"$alt_text\""; |
| 851 |
if (isset($this->titles[$link_id])) { |
| 852 |
$title = $this->titles[$link_id]; |
| 853 |
$title = $this->encodeAttribute($title); |
| 854 |
$result .= " title=\"$title\""; |
| 855 |
} |
| 856 |
$result .= $this->empty_element_suffix; |
| 857 |
$result = $this->hashPart($result); |
| 858 |
} else { |
| 859 |
// If there's no such link ID, leave intact: |
| 860 |
$result = $whole_match; |
| 861 |
} |
| 862 |
|
| 863 |
return $result; |
| 864 |
} |
| 865 |
|
| 866 |
/** |
| 867 |
* Callback to parse inline image tags |
| 868 |
* @param array $matches |
| 869 |
* @return string |
| 870 |
*/ |
| 871 |
protected function _doImages_inline_callback($matches) { |
| 872 |
$whole_match = $matches[1]; |
| 873 |
$alt_text = $matches[2]; |
| 874 |
$url = $matches[3] == '' ? $matches[4] : $matches[3]; |
| 875 |
$title =& $matches[7]; |
| 876 |
|
| 877 |
$alt_text = $this->encodeAttribute($alt_text); |
| 878 |
$url = $this->encodeURLAttribute($url); |
| 879 |
$result = "<img src=\"$url\" alt=\"$alt_text\""; |
| 880 |
if (isset($title)) { |
| 881 |
$title = $this->encodeAttribute($title); |
| 882 |
$result .= " title=\"$title\""; // $title already quoted |
| 883 |
} |
| 884 |
$result .= $this->empty_element_suffix; |
| 885 |
|
| 886 |
return $this->hashPart($result); |
| 887 |
} |
| 888 |
|
| 889 |
/** |
| 890 |
* Parse Markdown heading elements to HTML |
| 891 |
* @param string $text |
| 892 |
* @return string |
| 893 |
*/ |
| 894 |
protected function doHeaders($text) { |
| 895 |
/** |
| 896 |
* Setext-style headers: |
| 897 |
* Header 1 |
| 898 |
* ======== |
| 899 |
* |
| 900 |
* Header 2 |
| 901 |
* -------- |
| 902 |
*/ |
| 903 |
$text = preg_replace_callback('{ ^(.+?)[ ]*\n(=+|-+)[ ]*\n+ }mx', |
| 904 |
array($this, '_doHeaders_callback_setext'), $text); |
| 905 |
|
| 906 |
/** |
| 907 |
* atx-style headers: |
| 908 |
* # Header 1 |
| 909 |
* ## Header 2 |
| 910 |
* ## Header 2 with closing hashes ## |
| 911 |
* ... |
| 912 |
* ###### Header 6 |
| 913 |
*/ |
| 914 |
$text = preg_replace_callback('{ |
| 915 |
^(\#{1,6}) # $1 = string of #\'s |
| 916 |
[ ]* |
| 917 |
(.+?) # $2 = Header text |
| 918 |
[ ]* |
| 919 |
\#* # optional closing #\'s (not counted) |
| 920 |
\n+ |
| 921 |
}xm', |
| 922 |
array($this, '_doHeaders_callback_atx'), $text); |
| 923 |
|
| 924 |
return $text; |
| 925 |
} |
| 926 |
|
| 927 |
/** |
| 928 |
* Setext header parsing callback |
| 929 |
* @param array $matches |
| 930 |
* @return string |
| 931 |
*/ |
| 932 |
protected function _doHeaders_callback_setext($matches) { |
| 933 |
// Terrible hack to check we haven't found an empty list item. |
| 934 |
if ($matches[2] == '-' && preg_match('{^-(?: |$)}', $matches[1])) { |
| 935 |
return $matches[0]; |
| 936 |
} |
| 937 |
|
| 938 |
$level = $matches[2][0] == '=' ? 1 : 2; |
| 939 |
|
| 940 |
// ID attribute generation |
| 941 |
$idAtt = $this->_generateIdFromHeaderValue($matches[1]); |
| 942 |
|
| 943 |
$block = "<h$level$idAtt>".$this->runSpanGamut($matches[1])."</h$level>"; |
| 944 |
return "\n" . $this->hashBlock($block) . "\n\n"; |
| 945 |
} |
| 946 |
|
| 947 |
/** |
| 948 |
* ATX header parsing callback |
| 949 |
* @param array $matches |
| 950 |
* @return string |
| 951 |
*/ |
| 952 |
protected function _doHeaders_callback_atx($matches) { |
| 953 |
// ID attribute generation |
| 954 |
$idAtt = $this->_generateIdFromHeaderValue($matches[2]); |
| 955 |
|
| 956 |
$level = strlen($matches[1]); |
| 957 |
$block = "<h$level$idAtt>".$this->runSpanGamut($matches[2])."</h$level>"; |
| 958 |
return "\n" . $this->hashBlock($block) . "\n\n"; |
| 959 |
} |
| 960 |
|
| 961 |
/** |
| 962 |
* If a header_id_func property is set, we can use it to automatically |
| 963 |
* generate an id attribute. |
| 964 |
* |
| 965 |
* This method returns a string in the form id="foo", or an empty string |
| 966 |
* otherwise. |
| 967 |
* @param string $headerValue |
| 968 |
* @return string |
| 969 |
*/ |
| 970 |
protected function _generateIdFromHeaderValue($headerValue) { |
| 971 |
if (!is_callable($this->header_id_func)) { |
| 972 |
return ""; |
| 973 |
} |
| 974 |
|
| 975 |
$idValue = call_user_func($this->header_id_func, $headerValue); |
| 976 |
if (!$idValue) { |
| 977 |
return ""; |
| 978 |
} |
| 979 |
|
| 980 |
return ' id="' . $this->encodeAttribute($idValue) . '"'; |
| 981 |
} |
| 982 |
|
| 983 |
/** |
| 984 |
* Form HTML ordered (numbered) and unordered (bulleted) lists. |
| 985 |
* @param string $text |
| 986 |
* @return string |
| 987 |
*/ |
| 988 |
protected function doLists($text) { |
| 989 |
$less_than_tab = $this->tab_width - 1; |
| 990 |
|
| 991 |
// Re-usable patterns to match list item bullets and number markers: |
| 992 |
$marker_ul_re = '[*+-]'; |
| 993 |
$marker_ol_re = '\d+[\.]'; |
| 994 |
|
| 995 |
$markers_relist = array( |
| 996 |
$marker_ul_re => $marker_ol_re, |
| 997 |
$marker_ol_re => $marker_ul_re, |
| 998 |
); |
| 999 |
|
| 1000 |
foreach ($markers_relist as $marker_re => $other_marker_re) { |
| 1001 |
// Re-usable pattern to match any entirel ul or ol list: |
| 1002 |
$whole_list_re = ' |
| 1003 |
( # $1 = whole list |
| 1004 |
( # $2 |
| 1005 |
([ ]{0,'.$less_than_tab.'}) # $3 = number of spaces |
| 1006 |
('.$marker_re.') # $4 = first list item marker |
| 1007 |
[ ]+ |
| 1008 |
) |
| 1009 |
(?s:.+?) |
| 1010 |
( # $5 |
| 1011 |
\z |
| 1012 |
| |
| 1013 |
\n{2,} |
| 1014 |
(?=\S) |
| 1015 |
(?! # Negative lookahead for another list item marker |
| 1016 |
[ ]* |
| 1017 |
'.$marker_re.'[ ]+ |
| 1018 |
) |
| 1019 |
| |
| 1020 |
(?= # Lookahead for another kind of list |
| 1021 |
\n |
| 1022 |
\3 # Must have the same indentation |
| 1023 |
'.$other_marker_re.'[ ]+ |
| 1024 |
) |
| 1025 |
) |
| 1026 |
) |
| 1027 |
'; // mx |
| 1028 |
|
| 1029 |
// We use a different prefix before nested lists than top-level lists. |
| 1030 |
//See extended comment in _ProcessListItems(). |
| 1031 |
|
| 1032 |
if ($this->list_level) { |
| 1033 |
$text = preg_replace_callback('{ |
| 1034 |
^ |
| 1035 |
'.$whole_list_re.' |
| 1036 |
}mx', |
| 1037 |
array($this, '_doLists_callback'), $text); |
| 1038 |
} else { |
| 1039 |
$text = preg_replace_callback('{ |
| 1040 |
(?:(?<=\n)\n|\A\n?) # Must eat the newline |
| 1041 |
'.$whole_list_re.' |
| 1042 |
}mx', |
| 1043 |
array($this, '_doLists_callback'), $text); |
| 1044 |
} |
| 1045 |
} |
| 1046 |
|
| 1047 |
return $text; |
| 1048 |
} |
| 1049 |
|
| 1050 |
/** |
| 1051 |
* List parsing callback |
| 1052 |
* @param array $matches |
| 1053 |
* @return string |
| 1054 |
*/ |
| 1055 |
protected function _doLists_callback($matches) { |
| 1056 |
// Re-usable patterns to match list item bullets and number markers: |
| 1057 |
$marker_ul_re = '[*+-]'; |
| 1058 |
$marker_ol_re = '\d+[\.]'; |
| 1059 |
$marker_any_re = "(?:$marker_ul_re|$marker_ol_re)"; |
| 1060 |
$marker_ol_start_re = '[0-9]+'; |
| 1061 |
|
| 1062 |
$list = $matches[1]; |
| 1063 |
$list_type = preg_match("/$marker_ul_re/", $matches[4]) ? "ul" : "ol"; |
| 1064 |
|
| 1065 |
$marker_any_re = ( $list_type == "ul" ? $marker_ul_re : $marker_ol_re ); |
| 1066 |
|
| 1067 |
$list .= "\n"; |
| 1068 |
$result = $this->processListItems($list, $marker_any_re); |
| 1069 |
|
| 1070 |
$ol_start = 1; |
| 1071 |
if ($this->enhanced_ordered_list) { |
| 1072 |
// Get the start number for ordered list. |
| 1073 |
if ($list_type == 'ol') { |
| 1074 |
$ol_start_array = array(); |
| 1075 |
$ol_start_check = preg_match("/$marker_ol_start_re/", $matches[4], $ol_start_array); |
| 1076 |
if ($ol_start_check){ |
| 1077 |
$ol_start = $ol_start_array[0]; |
| 1078 |
} |
| 1079 |
} |
| 1080 |
} |
| 1081 |
|
| 1082 |
if ($ol_start > 1 && $list_type == 'ol'){ |
| 1083 |
$result = $this->hashBlock("<$list_type start=\"$ol_start\">\n" . $result . "</$list_type>"); |
| 1084 |
} else { |
| 1085 |
$result = $this->hashBlock("<$list_type>\n" . $result . "</$list_type>"); |
| 1086 |
} |
| 1087 |
return "\n". $result ."\n\n"; |
| 1088 |
} |
| 1089 |
|
| 1090 |
/** |
| 1091 |
* Nesting tracker for list levels |
| 1092 |
*/ |
| 1093 |
protected int $list_level = 0; |
| 1094 |
|
| 1095 |
/** |
| 1096 |
* Process the contents of a single ordered or unordered list, splitting it |
| 1097 |
* into individual list items. |
| 1098 |
* @param string $list_str |
| 1099 |
* @param string $marker_any_re |
| 1100 |
* @return string |
| 1101 |
*/ |
| 1102 |
protected function processListItems($list_str, $marker_any_re) { |
| 1103 |
/** |
| 1104 |
* The $this->list_level global keeps track of when we're inside a list. |
| 1105 |
* Each time we enter a list, we increment it; when we leave a list, |
| 1106 |
* we decrement. If it's zero, we're not in a list anymore. |
| 1107 |
* |
| 1108 |
* We do this because when we're not inside a list, we want to treat |
| 1109 |
* something like this: |
| 1110 |
* |
| 1111 |
* I recommend upgrading to version |
| 1112 |
* 8. Oops, now this line is treated |
| 1113 |
* as a sub-list. |
| 1114 |
* |
| 1115 |
* As a single paragraph, despite the fact that the second line starts |
| 1116 |
* with a digit-period-space sequence. |
| 1117 |
* |
| 1118 |
* Whereas when we're inside a list (or sub-list), that line will be |
| 1119 |
* treated as the start of a sub-list. What a kludge, huh? This is |
| 1120 |
* an aspect of Markdown's syntax that's hard to parse perfectly |
| 1121 |
* without resorting to mind-reading. Perhaps the solution is to |
| 1122 |
* change the syntax rules such that sub-lists must start with a |
| 1123 |
* starting cardinal number; e.g. "1." or "a.". |
| 1124 |
*/ |
| 1125 |
$this->list_level++; |
| 1126 |
|
| 1127 |
// Trim trailing blank lines: |
| 1128 |
$list_str = preg_replace("/\n{2,}\\z/", "\n", $list_str); |
| 1129 |
|
| 1130 |
$list_str = preg_replace_callback('{ |
| 1131 |
(\n)? # leading line = $1 |
| 1132 |
(^[ ]*) # leading whitespace = $2 |
| 1133 |
('.$marker_any_re.' # list marker and space = $3 |
| 1134 |
(?:[ ]+|(?=\n)) # space only required if item is not empty |
| 1135 |
) |
| 1136 |
((?s:.*?)) # list item text = $4 |
| 1137 |
(?:(\n+(?=\n))|\n) # tailing blank line = $5 |
| 1138 |
(?= \n* (\z | \2 ('.$marker_any_re.') (?:[ ]+|(?=\n)))) |
| 1139 |
}xm', |
| 1140 |
array($this, '_processListItems_callback'), $list_str); |
| 1141 |
|
| 1142 |
$this->list_level--; |
| 1143 |
return $list_str; |
| 1144 |
} |
| 1145 |
|
| 1146 |
/** |
| 1147 |
* List item parsing callback |
| 1148 |
* @param array $matches |
| 1149 |
* @return string |
| 1150 |
*/ |
| 1151 |
protected function _processListItems_callback($matches) { |
| 1152 |
$item = $matches[4]; |
| 1153 |
$leading_line =& $matches[1]; |
| 1154 |
$leading_space =& $matches[2]; |
| 1155 |
$marker_space = $matches[3]; |
| 1156 |
$tailing_blank_line =& $matches[5]; |
| 1157 |
|
| 1158 |
if ($leading_line || $tailing_blank_line || |
| 1159 |
preg_match('/\n{2,}/', $item)) |
| 1160 |
{ |
| 1161 |
// Replace marker with the appropriate whitespace indentation |
| 1162 |
$item = $leading_space . str_repeat(' ', strlen($marker_space)) . $item; |
| 1163 |
$item = $this->runBlockGamut($this->outdent($item)."\n"); |
| 1164 |
} else { |
| 1165 |
// Recursion for sub-lists: |
| 1166 |
$item = $this->doLists($this->outdent($item)); |
| 1167 |
$item = $this->formParagraphs($item, false); |
| 1168 |
} |
| 1169 |
|
| 1170 |
return "<li>" . $item . "</li>\n"; |
| 1171 |
} |
| 1172 |
|
| 1173 |
/** |
| 1174 |
* Process Markdown `<pre><code>` blocks. |
| 1175 |
* @param string $text |
| 1176 |
* @return string |
| 1177 |
*/ |
| 1178 |
protected function doCodeBlocks($text) { |
| 1179 |
$text = preg_replace_callback('{ |
| 1180 |
(?:\n\n|\A\n?) |
| 1181 |
( # $1 = the code block -- one or more lines, starting with a space/tab |
| 1182 |
(?> |
| 1183 |
[ ]{'.$this->tab_width.'} # Lines must start with a tab or a tab-width of spaces |
| 1184 |
.*\n+ |
| 1185 |
)+ |
| 1186 |
) |
| 1187 |
((?=^[ ]{0,'.$this->tab_width.'}\S)|\Z) # Lookahead for non-space at line-start, or end of doc |
| 1188 |
}xm', |
| 1189 |
array($this, '_doCodeBlocks_callback'), $text); |
| 1190 |
|
| 1191 |
return $text; |
| 1192 |
} |
| 1193 |
|
| 1194 |
/** |
| 1195 |
* Code block parsing callback |
| 1196 |
* @param array $matches |
| 1197 |
* @return string |
| 1198 |
*/ |
| 1199 |
protected function _doCodeBlocks_callback($matches) { |
| 1200 |
$codeblock = $matches[1]; |
| 1201 |
|
| 1202 |
$codeblock = $this->outdent($codeblock); |
| 1203 |
if (is_callable($this->code_block_content_func)) { |
| 1204 |
$codeblock = call_user_func($this->code_block_content_func, $codeblock, ""); |
| 1205 |
} else { |
| 1206 |
$codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES); |
| 1207 |
} |
| 1208 |
|
| 1209 |
# trim leading newlines and trailing newlines |
| 1210 |
$codeblock = preg_replace('/\A\n+|\n+\z/', '', $codeblock); |
| 1211 |
|
| 1212 |
$codeblock = "<pre><code>$codeblock\n</code></pre>"; |
| 1213 |
return "\n\n" . $this->hashBlock($codeblock) . "\n\n"; |
| 1214 |
} |
| 1215 |
|
| 1216 |
/** |
| 1217 |
* Create a code span markup for $code. Called from handleSpanToken. |
| 1218 |
* @param string $code |
| 1219 |
* @return string |
| 1220 |
*/ |
| 1221 |
protected function makeCodeSpan($code) { |
| 1222 |
if (is_callable($this->code_span_content_func)) { |
| 1223 |
$code = call_user_func($this->code_span_content_func, $code); |
| 1224 |
} else { |
| 1225 |
$code = htmlspecialchars(trim($code), ENT_NOQUOTES); |
| 1226 |
} |
| 1227 |
return $this->hashPart("<code>$code</code>"); |
| 1228 |
} |
| 1229 |
|
| 1230 |
/** |
| 1231 |
* Define the emphasis operators with their regex matches |
| 1232 |
* @var array |
| 1233 |
* |
| 1234 |
* 2025/03/19 jsmoriss - removed array type definition to avoid PHP error. |
| 1235 |
*/ |
| 1236 |
protected $em_relist = array( |
| 1237 |
'' => '(?:(?<!\*)\*(?!\*)|(?<!_)_(?!_))(?![\.,:;]?\s)', |
| 1238 |
'*' => '(?<![\s*])\*(?!\*)', |
| 1239 |
'_' => '(?<![\s_])_(?!_)', |
| 1240 |
); |
| 1241 |
|
| 1242 |
/** |
| 1243 |
* Define the strong operators with their regex matches |
| 1244 |
* @var array |
| 1245 |
*/ |
| 1246 |
protected $strong_relist = array( |
| 1247 |
'' => '(?:(?<!\*)\*\*(?!\*)|(?<!_)__(?!_))(?![\.,:;]?\s)', |
| 1248 |
'**' => '(?<![\s*])\*\*(?!\*)', |
| 1249 |
'__' => '(?<![\s_])__(?!_)', |
| 1250 |
); |
| 1251 |
|
| 1252 |
/** |
| 1253 |
* Define the emphasis + strong operators with their regex matches |
| 1254 |
* @var array |
| 1255 |
*/ |
| 1256 |
protected $em_strong_relist = array( |
| 1257 |
'' => '(?:(?<!\*)\*\*\*(?!\*)|(?<!_)___(?!_))(?![\.,:;]?\s)', |
| 1258 |
'***' => '(?<![\s*])\*\*\*(?!\*)', |
| 1259 |
'___' => '(?<![\s_])___(?!_)', |
| 1260 |
); |
| 1261 |
|
| 1262 |
/** |
| 1263 |
* Container for prepared regular expressions |
| 1264 |
*/ |
| 1265 |
protected ?array $em_strong_prepared_relist = null; |
| 1266 |
|
| 1267 |
/** |
| 1268 |
* Prepare regular expressions for searching emphasis tokens in any |
| 1269 |
* context. |
| 1270 |
* @return void |
| 1271 |
*/ |
| 1272 |
protected function prepareItalicsAndBold() { |
| 1273 |
foreach ($this->em_relist as $em => $em_re) { |
| 1274 |
foreach ($this->strong_relist as $strong => $strong_re) { |
| 1275 |
// Construct list of allowed token expressions. |
| 1276 |
$token_relist = array(); |
| 1277 |
if (isset($this->em_strong_relist["$em$strong"])) { |
| 1278 |
$token_relist[] = $this->em_strong_relist["$em$strong"]; |
| 1279 |
} |
| 1280 |
$token_relist[] = $em_re; |
| 1281 |
$token_relist[] = $strong_re; |
| 1282 |
|
| 1283 |
// Construct master expression from list. |
| 1284 |
$token_re = '{(' . implode('|', $token_relist) . ')}'; |
| 1285 |
$this->em_strong_prepared_relist["$em$strong"] = $token_re; |
| 1286 |
} |
| 1287 |
} |
| 1288 |
} |
| 1289 |
|
| 1290 |
/** |
| 1291 |
* Convert Markdown italics (emphasis) and bold (strong) to HTML |
| 1292 |
* @param string $text |
| 1293 |
* @return string |
| 1294 |
*/ |
| 1295 |
protected function doItalicsAndBold($text) { |
| 1296 |
if ($this->in_emphasis_processing) { |
| 1297 |
return $text; // avoid reentrency |
| 1298 |
} |
| 1299 |
$this->in_emphasis_processing = true; |
| 1300 |
|
| 1301 |
$token_stack = array(''); |
| 1302 |
$text_stack = array(''); |
| 1303 |
$em = ''; |
| 1304 |
$strong = ''; |
| 1305 |
$tree_char_em = false; |
| 1306 |
|
| 1307 |
while (1) { |
| 1308 |
// Get prepared regular expression for seraching emphasis tokens |
| 1309 |
// in current context. |
| 1310 |
$token_re = $this->em_strong_prepared_relist["$em$strong"]; |
| 1311 |
|
| 1312 |
// Each loop iteration search for the next emphasis token. |
| 1313 |
// Each token is then passed to handleSpanToken. |
| 1314 |
$parts = preg_split($token_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE); |
| 1315 |
$text_stack[0] .= $parts[0]; |
| 1316 |
$token =& $parts[1]; |
| 1317 |
$text =& $parts[2]; |
| 1318 |
|
| 1319 |
if (empty($token)) { |
| 1320 |
// Reached end of text span: empty stack without emitting. |
| 1321 |
// any more emphasis. |
| 1322 |
while ($token_stack[0]) { |
| 1323 |
$text_stack[1] .= array_shift($token_stack); |
| 1324 |
$text_stack[0] .= array_shift($text_stack); |
| 1325 |
} |
| 1326 |
break; |
| 1327 |
} |
| 1328 |
|
| 1329 |
$token_len = strlen($token); |
| 1330 |
if ($tree_char_em) { |
| 1331 |
// Reached closing marker while inside a three-char emphasis. |
| 1332 |
if ($token_len == 3) { |
| 1333 |
// Three-char closing marker, close em and strong. |
| 1334 |
array_shift($token_stack); |
| 1335 |
$span = array_shift($text_stack); |
| 1336 |
$span = $this->runSpanGamut($span); |
| 1337 |
$span = "<strong><em>$span</em></strong>"; |
| 1338 |
$text_stack[0] .= $this->hashPart($span); |
| 1339 |
$em = ''; |
| 1340 |
$strong = ''; |
| 1341 |
} else { |
| 1342 |
// Other closing marker: close one em or strong and |
| 1343 |
// change current token state to match the other |
| 1344 |
$token_stack[0] = str_repeat($token[0], 3-$token_len); |
| 1345 |
$tag = $token_len == 2 ? "strong" : "em"; |
| 1346 |
$span = $text_stack[0]; |
| 1347 |
$span = $this->runSpanGamut($span); |
| 1348 |
$span = "<$tag>$span</$tag>"; |
| 1349 |
$text_stack[0] = $this->hashPart($span); |
| 1350 |
$$tag = ''; // $$tag stands for $em or $strong |
| 1351 |
} |
| 1352 |
$tree_char_em = false; |
| 1353 |
} else if ($token_len == 3) { |
| 1354 |
if ($em) { |
| 1355 |
// Reached closing marker for both em and strong. |
| 1356 |
// Closing strong marker: |
| 1357 |
for ($i = 0; $i < 2; ++$i) { |
| 1358 |
$shifted_token = array_shift($token_stack); |
| 1359 |
$tag = strlen($shifted_token) == 2 ? "strong" : "em"; |
| 1360 |
$span = array_shift($text_stack); |
| 1361 |
$span = $this->runSpanGamut($span); |
| 1362 |
$span = "<$tag>$span</$tag>"; |
| 1363 |
$text_stack[0] .= $this->hashPart($span); |
| 1364 |
$$tag = ''; // $$tag stands for $em or $strong |
| 1365 |
} |
| 1366 |
} else { |
| 1367 |
// Reached opening three-char emphasis marker. Push on token |
| 1368 |
// stack; will be handled by the special condition above. |
| 1369 |
$em = $token[0]; |
| 1370 |
$strong = "$em$em"; |
| 1371 |
array_unshift($token_stack, $token); |
| 1372 |
array_unshift($text_stack, ''); |
| 1373 |
$tree_char_em = true; |
| 1374 |
} |
| 1375 |
} else if ($token_len == 2) { |
| 1376 |
if ($strong) { |
| 1377 |
// Unwind any dangling emphasis marker: |
| 1378 |
if (strlen($token_stack[0]) == 1) { |
| 1379 |
$text_stack[1] .= array_shift($token_stack); |
| 1380 |
$text_stack[0] .= array_shift($text_stack); |
| 1381 |
$em = ''; |
| 1382 |
} |
| 1383 |
// Closing strong marker: |
| 1384 |
array_shift($token_stack); |
| 1385 |
$span = array_shift($text_stack); |
| 1386 |
$span = $this->runSpanGamut($span); |
| 1387 |
$span = "<strong>$span</strong>"; |
| 1388 |
$text_stack[0] .= $this->hashPart($span); |
| 1389 |
$strong = ''; |
| 1390 |
} else { |
| 1391 |
array_unshift($token_stack, $token); |
| 1392 |
array_unshift($text_stack, ''); |
| 1393 |
$strong = $token; |
| 1394 |
} |
| 1395 |
} else { |
| 1396 |
// Here $token_len == 1 |
| 1397 |
if ($em) { |
| 1398 |
if (strlen($token_stack[0]) == 1) { |
| 1399 |
// Closing emphasis marker: |
| 1400 |
array_shift($token_stack); |
| 1401 |
$span = array_shift($text_stack); |
| 1402 |
$span = $this->runSpanGamut($span); |
| 1403 |
$span = "<em>$span</em>"; |
| 1404 |
$text_stack[0] .= $this->hashPart($span); |
| 1405 |
$em = ''; |
| 1406 |
} else { |
| 1407 |
$text_stack[0] .= $token; |
| 1408 |
} |
| 1409 |
} else { |
| 1410 |
array_unshift($token_stack, $token); |
| 1411 |
array_unshift($text_stack, ''); |
| 1412 |
$em = $token; |
| 1413 |
} |
| 1414 |
} |
| 1415 |
} |
| 1416 |
$this->in_emphasis_processing = false; |
| 1417 |
return $text_stack[0]; |
| 1418 |
} |
| 1419 |
|
| 1420 |
/** |
| 1421 |
* Parse Markdown blockquotes to HTML |
| 1422 |
* @param string $text |
| 1423 |
* @return string |
| 1424 |
*/ |
| 1425 |
protected function doBlockQuotes($text) { |
| 1426 |
$text = preg_replace_callback('/ |
| 1427 |
( # Wrap whole match in $1 |
| 1428 |
(?> |
| 1429 |
^[ ]*>[ ]? # ">" at the start of a line |
| 1430 |
.+\n # rest of the first line |
| 1431 |
(.+\n)* # subsequent consecutive lines |
| 1432 |
\n* # blanks |
| 1433 |
)+ |
| 1434 |
) |
| 1435 |
/xm', |
| 1436 |
array($this, '_doBlockQuotes_callback'), $text); |
| 1437 |
|
| 1438 |
return $text; |
| 1439 |
} |
| 1440 |
|
| 1441 |
/** |
| 1442 |
* Blockquote parsing callback |
| 1443 |
* @param array $matches |
| 1444 |
* @return string |
| 1445 |
*/ |
| 1446 |
protected function _doBlockQuotes_callback($matches) { |
| 1447 |
$bq = $matches[1]; |
| 1448 |
// trim one level of quoting - trim whitespace-only lines |
| 1449 |
$bq = preg_replace('/^[ ]*>[ ]?|^[ ]+$/m', '', $bq); |
| 1450 |
$bq = $this->runBlockGamut($bq); // recurse |
| 1451 |
|
| 1452 |
$bq = preg_replace('/^/m', " ", $bq); |
| 1453 |
// These leading spaces cause problem with <pre> content, |
| 1454 |
// so we need to fix that: |
| 1455 |
$bq = preg_replace_callback('{(\s*<pre>.+?</pre>)}sx', |
| 1456 |
array($this, '_doBlockQuotes_callback2'), $bq); |
| 1457 |
|
| 1458 |
return "\n" . $this->hashBlock("<blockquote>\n$bq\n</blockquote>") . "\n\n"; |
| 1459 |
} |
| 1460 |
|
| 1461 |
/** |
| 1462 |
* Blockquote parsing callback |
| 1463 |
* @param array $matches |
| 1464 |
* @return string |
| 1465 |
*/ |
| 1466 |
protected function _doBlockQuotes_callback2($matches) { |
| 1467 |
$pre = $matches[1]; |
| 1468 |
$pre = preg_replace('/^ /m', '', $pre); |
| 1469 |
return $pre; |
| 1470 |
} |
| 1471 |
|
| 1472 |
/** |
| 1473 |
* Parse paragraphs |
| 1474 |
* |
| 1475 |
* @param string $text String to process in paragraphs |
| 1476 |
* @param boolean $wrap_in_p Whether paragraphs should be wrapped in <p> tags |
| 1477 |
* @return string |
| 1478 |
*/ |
| 1479 |
protected function formParagraphs($text, $wrap_in_p = true) { |
| 1480 |
// Strip leading and trailing lines: |
| 1481 |
$text = preg_replace('/\A\n+|\n+\z/', '', $text); |
| 1482 |
|
| 1483 |
$grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY); |
| 1484 |
|
| 1485 |
// Wrap <p> tags and unhashify HTML blocks |
| 1486 |
foreach ($grafs as $key => $value) { |
| 1487 |
if (!preg_match('/^B\x1A[0-9]+B$/', $value)) { |
| 1488 |
// Is a paragraph. |
| 1489 |
$value = $this->runSpanGamut($value); |
| 1490 |
if ($wrap_in_p) { |
| 1491 |
$value = preg_replace('/^([ ]*)/', "<p>", $value); |
| 1492 |
$value .= "</p>"; |
| 1493 |
} |
| 1494 |
$grafs[$key] = $this->unhash($value); |
| 1495 |
} else { |
| 1496 |
// Is a block. |
| 1497 |
// Modify elements of @grafs in-place... |
| 1498 |
$graf = $value; |
| 1499 |
$block = $this->html_hashes[$graf]; |
| 1500 |
$graf = $block; |
| 1501 |
// if (preg_match('{ |
| 1502 |
// \A |
| 1503 |
// ( # $1 = <div> tag |
| 1504 |
// <div \s+ |
| 1505 |
// [^>]* |
| 1506 |
// \b |
| 1507 |
// markdown\s*=\s* ([\'"]) # $2 = attr quote char |
| 1508 |
// 1 |
| 1509 |
// \2 |
| 1510 |
// [^>]* |
| 1511 |
// > |
| 1512 |
// ) |
| 1513 |
// ( # $3 = contents |
| 1514 |
// .* |
| 1515 |
// ) |
| 1516 |
// (</div>) # $4 = closing tag |
| 1517 |
// \z |
| 1518 |
// }xs', $block, $matches)) |
| 1519 |
// { |
| 1520 |
// list(, $div_open, , $div_content, $div_close) = $matches; |
| 1521 |
// |
| 1522 |
// // We can't call Markdown(), because that resets the hash; |
| 1523 |
// // that initialization code should be pulled into its own sub, though. |
| 1524 |
// $div_content = $this->hashHTMLBlocks($div_content); |
| 1525 |
// |
| 1526 |
// // Run document gamut methods on the content. |
| 1527 |
// foreach ($this->document_gamut as $method => $priority) { |
| 1528 |
// $div_content = $this->$method($div_content); |
| 1529 |
// } |
| 1530 |
// |
| 1531 |
// $div_open = preg_replace( |
| 1532 |
// '{\smarkdown\s*=\s*([\'"]).+?\1}', '', $div_open); |
| 1533 |
// |
| 1534 |
// $graf = $div_open . "\n" . $div_content . "\n" . $div_close; |
| 1535 |
// } |
| 1536 |
$grafs[$key] = $graf; |
| 1537 |
} |
| 1538 |
} |
| 1539 |
|
| 1540 |
return implode("\n\n", $grafs); |
| 1541 |
} |
| 1542 |
|
| 1543 |
/** |
| 1544 |
* Encode text for a double-quoted HTML attribute. This function |
| 1545 |
* is *not* suitable for attributes enclosed in single quotes. |
| 1546 |
* @param string $text |
| 1547 |
* @return string |
| 1548 |
*/ |
| 1549 |
protected function encodeAttribute($text) { |
| 1550 |
$text = $this->encodeAmpsAndAngles($text); |
| 1551 |
$text = str_replace('"', '"', $text); |
| 1552 |
return $text; |
| 1553 |
} |
| 1554 |
|
| 1555 |
/** |
| 1556 |
* Encode text for a double-quoted HTML attribute containing a URL, |
| 1557 |
* applying the URL filter if set. Also generates the textual |
| 1558 |
* representation for the URL (removing mailto: or tel:) storing it in $text. |
| 1559 |
* This function is *not* suitable for attributes enclosed in single quotes. |
| 1560 |
* |
| 1561 |
* @param string $url |
| 1562 |
* @param string $text Passed by reference |
| 1563 |
* @return string URL |
| 1564 |
*/ |
| 1565 |
protected function encodeURLAttribute($url, &$text = null) { |
| 1566 |
if (is_callable($this->url_filter_func)) { |
| 1567 |
$url = call_user_func($this->url_filter_func, $url); |
| 1568 |
} |
| 1569 |
|
| 1570 |
if (preg_match('{^mailto:}i', $url)) { |
| 1571 |
$url = $this->encodeEntityObfuscatedAttribute($url, $text, 7); |
| 1572 |
} else if (preg_match('{^tel:}i', $url)) { |
| 1573 |
$url = $this->encodeAttribute($url); |
| 1574 |
$text = substr($url, 4); |
| 1575 |
} else { |
| 1576 |
$url = $this->encodeAttribute($url); |
| 1577 |
$text = $url; |
| 1578 |
} |
| 1579 |
|
| 1580 |
return $url; |
| 1581 |
} |
| 1582 |
|
| 1583 |
/** |
| 1584 |
* Smart processing for ampersands and angle brackets that need to |
| 1585 |
* be encoded. Valid character entities are left alone unless the |
| 1586 |
* no-entities mode is set. |
| 1587 |
* @param string $text |
| 1588 |
* @return string |
| 1589 |
*/ |
| 1590 |
protected function encodeAmpsAndAngles($text) { |
| 1591 |
if ($this->no_entities) { |
| 1592 |
$text = str_replace('&', '&', $text); |
| 1593 |
} else { |
| 1594 |
// Ampersand-encoding based entirely on Nat Irons's Amputator |
| 1595 |
// MT plugin: <http://bumppo.net/projects/amputator/> |
| 1596 |
$text = preg_replace('/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/', |
| 1597 |
'&', $text); |
| 1598 |
} |
| 1599 |
// Encode remaining <'s |
| 1600 |
$text = str_replace('<', '<', $text); |
| 1601 |
|
| 1602 |
return $text; |
| 1603 |
} |
| 1604 |
|
| 1605 |
/** |
| 1606 |
* Parse Markdown automatic links to anchor HTML tags |
| 1607 |
* @param string $text |
| 1608 |
* @return string |
| 1609 |
*/ |
| 1610 |
protected function doAutoLinks($text) { |
| 1611 |
$text = preg_replace_callback('{<((https?|ftp|dict|tel):[^\'">\s]+)>}i', |
| 1612 |
array($this, '_doAutoLinks_url_callback'), $text); |
| 1613 |
|
| 1614 |
// Email addresses: <address@domain.foo> |
| 1615 |
$text = preg_replace_callback('{ |
| 1616 |
< |
| 1617 |
(?:mailto:)? |
| 1618 |
( |
| 1619 |
(?: |
| 1620 |
[-!#$%&\'*+/=?^_`.{|}~\w\x80-\xFF]+ |
| 1621 |
| |
| 1622 |
".*?" |
| 1623 |
) |
| 1624 |
\@ |
| 1625 |
(?: |
| 1626 |
[-a-z0-9\x80-\xFF]+(\.[-a-z0-9\x80-\xFF]+)*\.[a-z]+ |
| 1627 |
| |
| 1628 |
\[[\d.a-fA-F:]+\] # IPv4 & IPv6 |
| 1629 |
) |
| 1630 |
) |
| 1631 |
> |
| 1632 |
}xi', |
| 1633 |
array($this, '_doAutoLinks_email_callback'), $text); |
| 1634 |
|
| 1635 |
return $text; |
| 1636 |
} |
| 1637 |
|
| 1638 |
/** |
| 1639 |
* Parse URL callback |
| 1640 |
* @param array $matches |
| 1641 |
* @return string |
| 1642 |
*/ |
| 1643 |
protected function _doAutoLinks_url_callback($matches) { |
| 1644 |
$url = $this->encodeURLAttribute($matches[1], $text); |
| 1645 |
$link = "<a href=\"$url\">$text</a>"; |
| 1646 |
return $this->hashPart($link); |
| 1647 |
} |
| 1648 |
|
| 1649 |
/** |
| 1650 |
* Parse email address callback |
| 1651 |
* @param array $matches |
| 1652 |
* @return string |
| 1653 |
*/ |
| 1654 |
protected function _doAutoLinks_email_callback($matches) { |
| 1655 |
$addr = $matches[1]; |
| 1656 |
$url = $this->encodeURLAttribute("mailto:$addr", $text); |
| 1657 |
$link = "<a href=\"$url\">$text</a>"; |
| 1658 |
return $this->hashPart($link); |
| 1659 |
} |
| 1660 |
|
| 1661 |
/** |
| 1662 |
* Input: some text to obfuscate, e.g. "mailto:foo@example.com" |
| 1663 |
* |
| 1664 |
* Output: the same text but with most characters encoded as either a |
| 1665 |
* decimal or hex entity, in the hopes of foiling most address |
| 1666 |
* harvesting spam bots. E.g.: |
| 1667 |
* |
| 1668 |
* mailto:foo |
| 1669 |
* @example.co |
| 1670 |
* m |
| 1671 |
* |
| 1672 |
* Note: the additional output $tail is assigned the same value as the |
| 1673 |
* ouput, minus the number of characters specified by $head_length. |
| 1674 |
* |
| 1675 |
* Based by a filter by Matthew Wickline, posted to BBEdit-Talk. |
| 1676 |
* With some optimizations by Milian Wolff. Forced encoding of HTML |
| 1677 |
* attribute special characters by Allan Odgaard. |
| 1678 |
* |
| 1679 |
* @param string $text |
| 1680 |
* @param string $tail Passed by reference |
| 1681 |
* @param integer $head_length |
| 1682 |
* @return string |
| 1683 |
*/ |
| 1684 |
protected function encodeEntityObfuscatedAttribute($text, &$tail = null, $head_length = 0) { |
| 1685 |
if ($text == "") { |
| 1686 |
return $tail = ""; |
| 1687 |
} |
| 1688 |
|
| 1689 |
$chars = preg_split('/(?<!^)(?!$)/', $text); |
| 1690 |
$seed = (int)abs(crc32($text) / strlen($text)); // Deterministic seed. |
| 1691 |
|
| 1692 |
foreach ($chars as $key => $char) { |
| 1693 |
$ord = ord($char); |
| 1694 |
// Ignore non-ascii chars. |
| 1695 |
if ($ord < 128) { |
| 1696 |
$r = ($seed * (1 + $key)) % 100; // Pseudo-random function. |
| 1697 |
// roughly 10% raw, 45% hex, 45% dec |
| 1698 |
// '@' *must* be encoded. I insist. |
| 1699 |
// '"' and '>' have to be encoded inside the attribute |
| 1700 |
if ($r > 90 && strpos('@"&>', $char) === false) { |
| 1701 |
/* do nothing */ |
| 1702 |
} else if ($r < 45) { |
| 1703 |
$chars[$key] = '&#x'.dechex($ord).';'; |
| 1704 |
} else { |
| 1705 |
$chars[$key] = '&#'.$ord.';'; |
| 1706 |
} |
| 1707 |
} |
| 1708 |
} |
| 1709 |
|
| 1710 |
$text = implode('', $chars); |
| 1711 |
$tail = $head_length ? implode('', array_slice($chars, $head_length)) : $text; |
| 1712 |
|
| 1713 |
return $text; |
| 1714 |
} |
| 1715 |
|
| 1716 |
/** |
| 1717 |
* Take the string $str and parse it into tokens, hashing embeded HTML, |
| 1718 |
* escaped characters and handling code spans. |
| 1719 |
* @param string $str |
| 1720 |
* @return string |
| 1721 |
*/ |
| 1722 |
protected function parseSpan($str) { |
| 1723 |
$output = ''; |
| 1724 |
|
| 1725 |
$span_re = '{ |
| 1726 |
( |
| 1727 |
\\\\'.$this->escape_chars_re.' |
| 1728 |
| |
| 1729 |
(?<![`\\\\]) |
| 1730 |
`+ # code span marker |
| 1731 |
'.( $this->no_markup ? '' : ' |
| 1732 |
| |
| 1733 |
<!-- .*? --> # comment |
| 1734 |
| |
| 1735 |
<\?.*?\?> | <%.*?%> # processing instruction |
| 1736 |
| |
| 1737 |
<[!$]?[-a-zA-Z0-9:_]+ # regular tags |
| 1738 |
(?> |
| 1739 |
\s |
| 1740 |
(?>[^"\'>]+|"[^"]*"|\'[^\']*\')* |
| 1741 |
)? |
| 1742 |
> |
| 1743 |
| |
| 1744 |
<[-a-zA-Z0-9:_]+\s*/> # xml-style empty tag |
| 1745 |
| |
| 1746 |
</[-a-zA-Z0-9:_]+\s*> # closing tag |
| 1747 |
').' |
| 1748 |
) |
| 1749 |
}xs'; |
| 1750 |
|
| 1751 |
while (1) { |
| 1752 |
// Each loop iteration seach for either the next tag, the next |
| 1753 |
// openning code span marker, or the next escaped character. |
| 1754 |
// Each token is then passed to handleSpanToken. |
| 1755 |
$parts = preg_split($span_re, $str, 2, PREG_SPLIT_DELIM_CAPTURE); |
| 1756 |
|
| 1757 |
// Create token from text preceding tag. |
| 1758 |
if ($parts[0] != "") { |
| 1759 |
$output .= $parts[0]; |
| 1760 |
} |
| 1761 |
|
| 1762 |
// Check if we reach the end. |
| 1763 |
if (isset($parts[1])) { |
| 1764 |
$output .= $this->handleSpanToken($parts[1], $parts[2]); |
| 1765 |
$str = $parts[2]; |
| 1766 |
} else { |
| 1767 |
break; |
| 1768 |
} |
| 1769 |
} |
| 1770 |
|
| 1771 |
return $output; |
| 1772 |
} |
| 1773 |
|
| 1774 |
/** |
| 1775 |
* Handle $token provided by parseSpan by determining its nature and |
| 1776 |
* returning the corresponding value that should replace it. |
| 1777 |
* @param string $token |
| 1778 |
* @param string $str Passed by reference |
| 1779 |
* @return string |
| 1780 |
*/ |
| 1781 |
protected function handleSpanToken($token, &$str) { |
| 1782 |
switch ($token[0]) { |
| 1783 |
case "\\": |
| 1784 |
return $this->hashPart("&#". ord($token[1]). ";"); |
| 1785 |
case "`": |
| 1786 |
// Search for end marker in remaining text. |
| 1787 |
if (preg_match('/^(.*?[^`])'.preg_quote($token).'(?!`)(.*)$/sm', |
| 1788 |
$str, $matches)) |
| 1789 |
{ |
| 1790 |
$str = $matches[2]; |
| 1791 |
$codespan = $this->makeCodeSpan($matches[1]); |
| 1792 |
return $this->hashPart($codespan); |
| 1793 |
} |
| 1794 |
return $token; // Return as text since no ending marker found. |
| 1795 |
default: |
| 1796 |
return $this->hashPart($token); |
| 1797 |
} |
| 1798 |
} |
| 1799 |
|
| 1800 |
/** |
| 1801 |
* Remove one level of line-leading tabs or spaces |
| 1802 |
* @param string $text |
| 1803 |
* @return string |
| 1804 |
*/ |
| 1805 |
protected function outdent($text) { |
| 1806 |
return preg_replace('/^(\t|[ ]{1,' . $this->tab_width . '})/m', '', $text); |
| 1807 |
} |
| 1808 |
|
| 1809 |
|
| 1810 |
/** |
| 1811 |
* String length function for detab. `_initDetab` will create a function to |
| 1812 |
* handle UTF-8 if the default function does not exist. |
| 1813 |
* can be a string or function |
| 1814 |
*/ |
| 1815 |
protected $utf8_strlen = 'mb_strlen'; |
| 1816 |
|
| 1817 |
/** |
| 1818 |
* Replace tabs with the appropriate amount of spaces. |
| 1819 |
* |
| 1820 |
* For each line we separate the line in blocks delemited by tab characters. |
| 1821 |
* Then we reconstruct every line by adding the appropriate number of space |
| 1822 |
* between each blocks. |
| 1823 |
* |
| 1824 |
* @param string $text |
| 1825 |
* @return string |
| 1826 |
*/ |
| 1827 |
protected function detab($text) { |
| 1828 |
$text = preg_replace_callback('/^.*\t.*$/m', |
| 1829 |
array($this, '_detab_callback'), $text); |
| 1830 |
|
| 1831 |
return $text; |
| 1832 |
} |
| 1833 |
|
| 1834 |
/** |
| 1835 |
* Replace tabs callback |
| 1836 |
* @param string $matches |
| 1837 |
* @return string |
| 1838 |
*/ |
| 1839 |
protected function _detab_callback($matches) { |
| 1840 |
$line = $matches[0]; |
| 1841 |
$strlen = $this->utf8_strlen; // strlen function for UTF-8. |
| 1842 |
|
| 1843 |
// Split in blocks. |
| 1844 |
$blocks = explode("\t", $line); |
| 1845 |
// Add each blocks to the line. |
| 1846 |
$line = $blocks[0]; |
| 1847 |
unset($blocks[0]); // Do not add first block twice. |
| 1848 |
foreach ($blocks as $block) { |
| 1849 |
// Calculate amount of space, insert spaces, insert block. |
| 1850 |
$amount = $this->tab_width - |
| 1851 |
$strlen($line, 'UTF-8') % $this->tab_width; |
| 1852 |
$line .= str_repeat(" ", $amount) . $block; |
| 1853 |
} |
| 1854 |
return $line; |
| 1855 |
} |
| 1856 |
|
| 1857 |
/** |
| 1858 |
* Check for the availability of the function in the `utf8_strlen` property |
| 1859 |
* (initially `mb_strlen`). If the function is not available, create a |
| 1860 |
* function that will loosely count the number of UTF-8 characters with a |
| 1861 |
* regular expression. |
| 1862 |
* @return void |
| 1863 |
*/ |
| 1864 |
protected function _initDetab() { |
| 1865 |
|
| 1866 |
if (function_exists($this->utf8_strlen)) { |
| 1867 |
return; |
| 1868 |
} |
| 1869 |
|
| 1870 |
$this->utf8_strlen = fn($text) => preg_match_all('/[\x00-\xBF]|[\xC0-\xFF][\x80-\xBF]*/', $text, $m); |
| 1871 |
} |
| 1872 |
|
| 1873 |
/** |
| 1874 |
* Swap back in all the tags hashed by _HashHTMLBlocks. |
| 1875 |
* @param string $text |
| 1876 |
* @return string |
| 1877 |
*/ |
| 1878 |
protected function unhash($text) { |
| 1879 |
return preg_replace_callback('/(.)\x1A[0-9]+\1/', |
| 1880 |
array($this, '_unhash_callback'), $text); |
| 1881 |
} |
| 1882 |
|
| 1883 |
/** |
| 1884 |
* Unhashing callback |
| 1885 |
* @param array $matches |
| 1886 |
* @return string |
| 1887 |
*/ |
| 1888 |
protected function _unhash_callback($matches) { |
| 1889 |
return $this->html_hashes[$matches[0]]; |
| 1890 |
} |
| 1891 |
} |
| 1892 |
|
