options = $params; $this->options['toc'] = $this->options['toc'] ?? false; // Marks $state = $this->options['mark'] ?? true; if ($state !== false) { $this->InlineTypes['='][] = 'mark'; $this->inlineMarkerList .= '='; } // Keystrokes $state = $this->options['keystrokes'] ?? true; if ($state !== false) { $this->InlineTypes['['][] = 'Keystrokes'; $this->inlineMarkerList .= '['; } // Inline Math $state = $this->options['math'] ?? false; if ($state !== false) { $this->InlineTypes['\\'][] = 'Math'; $this->inlineMarkerList .= '\\'; $this->InlineTypes['$'][] = 'Math'; $this->inlineMarkerList .= '$'; } // Superscript $state = $this->options['sup'] ?? false; if ($state !== false) { $this->InlineTypes['^'][] = 'Superscript'; $this->inlineMarkerList .= '^'; } // Subscript $state = $this->options['sub'] ?? false; if ($state !== false) { $this->InlineTypes['~'][] = 'Subscript'; } // Emojis $state = $this->options['emojis'] ?? true; if ($state !== false) { $this->InlineTypes[':'][] = 'Emojis'; $this->inlineMarkerList .= ':'; } // Typographer $state = $this->options['typographer'] ?? false; if ($state !== false) { $this->InlineTypes['('][] = 'Typographer'; $this->inlineMarkerList .= '('; $this->InlineTypes['.'][] = 'Typographer'; $this->inlineMarkerList .= '.'; $this->InlineTypes['+'][] = 'Typographer'; $this->inlineMarkerList .= '+'; $this->InlineTypes['!'][] = 'Typographer'; $this->inlineMarkerList .= '!'; $this->InlineTypes['?'][] = 'Typographer'; $this->inlineMarkerList .= '?'; } // Smartypants $state = $this->options['smarty'] ?? false; if ($state !== false) { $this->InlineTypes['<'][] = 'Smartypants'; $this->inlineMarkerList .= '<'; $this->InlineTypes['>'][] = 'Smartypants'; $this->inlineMarkerList .= '>'; $this->InlineTypes['-'][] = 'Smartypants'; $this->inlineMarkerList .= '-'; $this->InlineTypes['.'][] = 'Smartypants'; $this->inlineMarkerList .= '.'; $this->InlineTypes["'"][] = 'Smartypants'; $this->inlineMarkerList .= "'"; $this->InlineTypes['"'][] = 'Smartypants'; $this->inlineMarkerList .= '"'; $this->InlineTypes['`'][] = 'Smartypants'; $this->inlineMarkerList .= '`'; } /* * Blocks * ------------------------------------------------------------------------ */ // Block Math $state = $this->options['math'] ?? false; if ($state !== false) { $this->BlockTypes['\\'][] = 'Math'; $this->BlockTypes['$'][] = 'Math'; } // Task $state = $this->options['lists']['tasks'] ?? true; if ($state !== false) { $this->BlockTypes['['][] = 'Checkbox'; } } public function textParent($text) { $Elements = $this->textElements($text); # convert to markup $markup = $this->elements($Elements); # trim line breaks $markup = \trim($markup, "\n"); # merge consecutive dl elements $markup = \preg_replace('/<\/dl>\s+
'.$tagOrigin.'
'; $replace = "[[TOC]]
" */ protected function encodeTagToHash($text) { $salt = $this->getSalt(); $tagOrigin = $this->getTagToC(); if (\strpos($text, $tagOrigin) === false) { return $text; } $tagHashed = \hash('sha256', $salt.$tagOrigin); return \str_replace($tagOrigin, $tagHashed, $text); } /** * Decodes the hashed ToC tag to an original tag and replaces. * * This is used to avoid parsing user defined ToC tag which includes "_" in * their tag such as "[[_toc_]]". Unless it will be parsed as: * "[[TOC]]
" */ protected function decodeTagFromHash($text) { $salt = $this->getSalt(); $tagOrigin = $this->getTagToC(); $tagHashed = \hash('sha256', $salt.$tagOrigin); if (\strpos($text, $tagHashed) === false) { return $text; } return \str_replace($tagHashed, $tagOrigin, $text); } /** * Unique string to use as a salt value. */ protected function getSalt() { static $salt; if (isset($salt)) { return $salt; } $salt = \hash('md5', (string) \time()); return $salt; } /** * Gets the markdown tag for ToC. */ protected function getTagToC() { return $this->options['toc']['set_toc_tag'] ?? '[toc]'; } /** * Gets the ID attribute of the ToC for HTML tags. */ protected function getIdAttributeToC() { if (!empty($this->idToc)) { return $this->idToc; } return self::ID_ATTRIBUTE_DEFAULT; } /** * Generates an anchor text that are link-able even if the heading is not in * ASCII. */ protected function createAnchorID($str) : string { $optionUrlEncode = $this->options['toc']['urlencode'] ?? false; if ($optionUrlEncode) { // Check AnchorID is unique $str = $this->incrementAnchorId($str); return \urlencode($str); } $charMap = [ // Latin 'ร' => 'A', 'ร' => 'A', 'ร' => 'A', 'ร' => 'A', 'ร' => 'A', 'ร ' => 'AA', 'ร' => 'AE', 'ร' => 'C', 'ร' => 'E', 'ร' => 'E', 'ร' => 'E', 'ร' => 'E', 'ร' => 'I', 'ร' => 'I', 'ร' => 'I', 'ร' => 'I', 'ร' => 'D', 'ร' => 'N', 'ร' => 'O', 'ร' => 'O', 'ร' => 'O', 'ร' => 'O', 'ร' => 'O', 'ล' => 'O', 'ร' => 'OE', 'ร' => 'U', 'ร' => 'U', 'ร' => 'U', 'ร' => 'U', 'ลฐ' => 'U', 'ร' => 'Y', 'ร' => 'TH', 'ร' => 'ss', 'ร ' => 'a', 'รก' => 'a', 'รข' => 'a', 'รฃ' => 'a', 'รค' => 'a', 'รฅ' => 'aa', 'รฆ' => 'ae', 'รง' => 'c', 'รจ' => 'e', 'รฉ' => 'e', 'รช' => 'e', 'รซ' => 'e', 'รฌ' => 'i', 'รญ' => 'i', 'รฎ' => 'i', 'รฏ' => 'i', 'รฐ' => 'd', 'รฑ' => 'n', 'รฒ' => 'o', 'รณ' => 'o', 'รด' => 'o', 'รต' => 'o', 'รถ' => 'o', 'ล' => 'o', 'รธ' => 'oe', 'รน' => 'u', 'รบ' => 'u', 'รป' => 'u', 'รผ' => 'u', 'ลฑ' => 'u', 'รฝ' => 'y', 'รพ' => 'th', 'รฟ' => 'y', // Latin symbols 'ยฉ' => '(c)', 'ยฎ' => '(r)', 'โข' => '(tm)', // Greek 'ฮ' => 'A', 'ฮ' => 'B', 'ฮ' => 'G', 'ฮ' => 'D', 'ฮ' => 'E', 'ฮ' => 'Z', 'ฮ' => 'H', 'ฮ' => '8', 'ฮ' => 'I', 'ฮ' => 'K', 'ฮ' => 'L', 'ฮ' => 'M', 'ฮ' => 'N', 'ฮ' => '3', 'ฮ' => 'O', 'ฮ ' => 'P', 'ฮก' => 'R', 'ฮฃ' => 'S', 'ฮค' => 'T', 'ฮฅ' => 'Y', 'ฮฆ' => 'F', 'ฮง' => 'X', 'ฮจ' => 'PS', 'ฮฉ' => 'W', 'ฮ' => 'A', 'ฮ' => 'E', 'ฮ' => 'I', 'ฮ' => 'O', 'ฮ' => 'Y', 'ฮ' => 'H', 'ฮ' => 'W', 'ฮช' => 'I', 'ฮซ' => 'Y', 'ฮฑ' => 'a', 'ฮฒ' => 'b', 'ฮณ' => 'g', 'ฮด' => 'd', 'ฮต' => 'e', 'ฮถ' => 'z', 'ฮท' => 'h', 'ฮธ' => '8', 'ฮน' => 'i', 'ฮบ' => 'k', 'ฮป' => 'l', 'ฮผ' => 'm', 'ฮฝ' => 'n', 'ฮพ' => '3', 'ฮฟ' => 'o', 'ฯ' => 'p', 'ฯ' => 'r', 'ฯ' => 's', 'ฯ' => 't', 'ฯ ' => 'y', 'ฯ' => 'f', 'ฯ' => 'x', 'ฯ' => 'ps', 'ฯ' => 'w', 'ฮฌ' => 'a', 'ฮญ' => 'e', 'ฮฏ' => 'i', 'ฯ' => 'o', 'ฯ' => 'y', 'ฮฎ' => 'h', 'ฯ' => 'w', 'ฯ' => 's', 'ฯ' => 'i', 'ฮฐ' => 'y', 'ฯ' => 'y', 'ฮ' => 'i', // Turkish 'ล' => 'S', 'ฤฐ' => 'I', 'ฤ' => 'G', 'ล' => 's', 'ฤฑ' => 'i', 'ฤ' => 'g', // Russian 'ะ' => 'A', 'ะ' => 'B', 'ะ' => 'V', 'ะ' => 'G', 'ะ' => 'D', 'ะ' => 'E', 'ะ' => 'Yo', 'ะ' => 'Zh', 'ะ' => 'Z', 'ะ' => 'I', 'ะ' => 'J', 'ะ' => 'K', 'ะ' => 'L', 'ะ' => 'M', 'ะ' => 'N', 'ะ' => 'O', 'ะ' => 'P', 'ะ ' => 'R', 'ะก' => 'S', 'ะข' => 'T', 'ะฃ' => 'U', 'ะค' => 'F', 'ะฅ' => 'H', 'ะฆ' => 'C', 'ะง' => 'Ch', 'ะจ' => 'Sh', 'ะฉ' => 'Sh', 'ะช' => '', 'ะซ' => 'Y', 'ะฌ' => '', 'ะญ' => 'E', 'ะฎ' => 'Yu', 'ะฏ' => 'Ya', 'ะฐ' => 'a', 'ะฑ' => 'b', 'ะฒ' => 'v', 'ะณ' => 'g', 'ะด' => 'd', 'ะต' => 'e', 'ั' => 'yo', 'ะถ' => 'zh', 'ะท' => 'z', 'ะธ' => 'i', 'ะน' => 'j', 'ะบ' => 'k', 'ะป' => 'l', 'ะผ' => 'm', 'ะฝ' => 'n', 'ะพ' => 'o', 'ะฟ' => 'p', 'ั' => 'r', 'ั' => 's', 'ั' => 't', 'ั' => 'u', 'ั' => 'f', 'ั ' => 'h', 'ั' => 'c', 'ั' => 'ch', 'ั' => 'sh', 'ั' => 'sh', 'ั' => '', 'ั' => 'y', 'ั' => '', 'ั' => 'e', 'ั' => 'yu', 'ั' => 'ya', // Ukrainian 'ะ' => 'Ye', 'ะ' => 'I', 'ะ' => 'Yi', 'า' => 'G', 'ั' => 'ye', 'ั' => 'i', 'ั' => 'yi', 'า' => 'g', // Czech 'ฤ' => 'C', 'ฤ' => 'D', 'ฤ' => 'E', 'ล' => 'N', 'ล' => 'R', 'ล ' => 'S', 'ลค' => 'T', 'ลฎ' => 'U', 'ลฝ' => 'Z', 'ฤ' => 'c', 'ฤ' => 'd', 'ฤ' => 'e', 'ล' => 'n', 'ล' => 'r', 'ลก' => 's', 'ลฅ' => 't', 'ลฏ' => 'u', 'ลพ' => 'z', // Polish 'ฤ' => 'A', 'ฤ' => 'C', 'ฤ' => 'e', 'ล' => 'L', 'ล' => 'N', 'ล' => 'S', 'ลน' => 'Z', 'ลป' => 'Z', 'ฤ ' => 'a', 'ฤ' => 'c', 'ฤ' => 'e', 'ล' => 'l', 'ล' => 'n', 'ล' => 's', 'ลบ' => 'z', 'ลผ' => 'z', // Latvian 'ฤ' => 'A', 'ฤ' => 'E', 'ฤข' => 'G', 'ฤช' => 'i', 'ฤถ' => 'k', 'ฤป' => 'L', 'ล ' => 'N', 'ลช' => 'u', 'ฤ' => 'a', 'ฤ' => 'e', 'ฤฃ' => 'g', 'ฤซ' => 'i', 'ฤท' => 'k', 'ฤผ' => 'l', 'ล' => 'n', 'ลซ' => 'u', ]; // Transliterate characters to ASCII $optionTransliterate = $this->options['toc']['transliterate'] ?? false; if ($optionTransliterate) { $str = \str_replace(\array_keys($charMap), $charMap, $str); } // Replace non-alphanumeric characters with our delimiter $optionDelimiter = $this->options['toc']['delimiter'] ?? '-'; $str = \preg_replace('/[^\p{L}\p{Nd}]+/u', $optionDelimiter, $str); // Remove duplicate delimiters $str = \preg_replace('/('.\preg_quote($optionDelimiter, '/').'){2,}/', '$1', $str); // Truncate slug to max. characters $optionLimit = $this->options['toc']['limit'] ?? \mb_strlen($str, 'UTF-8'); $str = \mb_substr($str, 0, $optionLimit, 'UTF-8'); // Remove delimiter from ends $str = \trim($str, $optionDelimiter); $urlLowercase = $this->options['toc']['lowercase'] ?? true; $str = $urlLowercase ? \mb_strtolower($str, 'UTF-8') : $str; return $this->incrementAnchorId($str); } /** * Get only the text from a markdown string. * It parses to HTML once then trims the tags to get the text. */ protected function fetchText($text) : string { return \trim(\strip_tags($this->line($text))); } /** * Set/stores the heading block to ToC list in a string and array format. */ protected function setContentsList(array $Content) : void { // Stores as an array $this->setContentsListAsArray($Content); // Stores as string in markdown list format. $this->setContentsListAsString($Content); } /** * Sets/stores the heading block info as an array. */ protected function setContentsListAsArray(array $Content) : void { $this->contentsListArray[] = $Content; } /** * Sets/stores the heading block info as a list in markdown format. */ protected function setContentsListAsString(array $Content) : void { $text = $this->fetchText($Content['text']); $id = $Content['id']; $level = (int) \trim($Content['level'], 'h'); $link = "[{$text}](#{$id})"; if ($this->firstHeadLevel === 0) { $this->firstHeadLevel = $level; } $cutIndent = $this->firstHeadLevel - 1; $level = $cutIndent > $level ? 1 : $level - $cutIndent; $indent = \str_repeat(' ', $level); // Stores in markdown list format as below: // - [Header1](#Header1) // - [Header2-1](#Header2-1) // - [Header3](#Header3) // - [Header2-2](#Header2-2) // ... $this->contentsListString .= "{$indent}- {$link}".\PHP_EOL; } /** * Collect and count anchors in use to prevent duplicated ids. Return string * with incremental, numeric suffix. Also init optional blacklist of ids. */ protected function incrementAnchorId($str) { // add blacklist to list of used anchors if (!$this->isBlacklistInitialized) { $this->initBlacklist(); } $this->anchorDuplicates[$str] = isset($this->anchorDuplicates[$str]) ? ++$this->anchorDuplicates[$str] : 0; $newStr = $str; if ($count = $this->anchorDuplicates[$str]) { $newStr .= "-{$count}"; // increment until conversion doesn't produce new duplicates anymore if (isset($this->anchorDuplicates[$newStr])) { $newStr = $this->incrementAnchorId($str); } else { $this->anchorDuplicates[$newStr] = 0; } } return $newStr; } /** * Add blacklisted ids to anchor list. */ protected function initBlacklist() : void { if ($this->isBlacklistInitialized) { return; } if (!empty($this->options['headings']['blacklist']) && \is_array($this->options['headings']['blacklist'])) { foreach ($this->options['headings']['blacklist'] as $v) { if (\is_string($v)) { $this->anchorDuplicates[$v] = 0; } } } $this->isBlacklistInitialized = true; } protected function lineElements($text, $nonNestables = []) { $Elements = []; $nonNestables = ( empty($nonNestables) ? [] : \array_combine($nonNestables, $nonNestables) ); // $excerpt is based on the first occurrence of a marker while ($excerpt = \strpbrk($text, $this->inlineMarkerList)) { $marker = $excerpt[0]; $markerPosition = \strlen($text) - \strlen($excerpt); // Get the first char before the marker $beforeMarkerPosition = $markerPosition - 1; $charBeforeMarker = $beforeMarkerPosition >= 0 ? $text[$markerPosition - 1] : ''; $Excerpt = ['text' => $excerpt, 'context' => $text, 'before' => $charBeforeMarker]; foreach ($this->InlineTypes[$marker] as $inlineType) { // check to see if the current inline type is nestable in the current context if (isset($nonNestables[$inlineType])) { continue; } $Inline = $this->{"inline{$inlineType}"}($Excerpt); if (!isset($Inline)) { continue; } // makes sure that the inline belongs to "our" marker if (isset($Inline['position']) && $Inline['position'] > $markerPosition) { continue; } // sets a default inline position if (!isset($Inline['position'])) { $Inline['position'] = $markerPosition; } // cause the new element to 'inherit' our non nestables $Inline['element']['nonNestables'] = isset($Inline['element']['nonNestables']) ? \array_merge($Inline['element']['nonNestables'], $nonNestables) : $nonNestables ; // the text that comes before the inline $unmarkedText = \substr($text, 0, $Inline['position']); // compile the unmarked text $InlineText = $this->inlineText($unmarkedText); $Elements[] = $InlineText['element']; // compile the inline $Elements[] = $this->extractElement($Inline); // remove the examined text $text = \substr($text, $Inline['position'] + $Inline['extent']); continue 2; } // the marker does not belong to an inline $unmarkedText = \substr($text, 0, $markerPosition + 1); $InlineText = $this->inlineText($unmarkedText); $Elements[] = $InlineText['element']; $text = \substr($text, $markerPosition + 1); } $InlineText = $this->inlineText($text); $Elements[] = $InlineText['element']; foreach ($Elements as &$Element) { if (!isset($Element['autobreak'])) { $Element['autobreak'] = false; } } return $Elements; } private function pregReplaceAssoc(array $replace, $subject) { return \preg_replace(\array_keys($replace), \array_values($replace), $subject); } # # Blocks # # # Abbreviation protected function blockAbbreviationBase($Line) { if (\preg_match('/^\*\[(.+?)\]:[ ]*(.+?)[ ]*$/', $Line['text'], $matches)) { $this->DefinitionData['Abbreviation'][$matches[1]] = $matches[2]; return [ 'hidden' => true, ]; } } # # Footnote protected function blockFootnoteBase($Line) { if (\preg_match('/^\[\^(.+?)\]:[ ]?(.*)$/', $Line['text'], $matches)) { return [ 'label' => $matches[1], 'text' => $matches[2], 'hidden' => true, ]; } } protected function blockFootnoteContinue($Line, $Block) { if ($Line['text'][0] === '[' && \preg_match('/^\[\^(.+?)\]:/', $Line['text'])) { return; } if (isset($Block['interrupted'])) { if ($Line['indent'] >= 4) { $Block['text'] .= "\n\n" . $Line['text']; return $Block; } } else { $Block['text'] .= "\n" . $Line['text']; return $Block; } } protected function blockFootnoteComplete($Block) { $this->DefinitionData['Footnote'][$Block['label']] = [ 'text' => $Block['text'], 'count' => null, 'number' => null, ]; return $Block; } # # Definition List protected function blockDefinitionListBase($Line, $Block) { if (!isset($Block) || $Block['type'] !== 'Paragraph') { return; } $Element = [ 'name' => 'dl', 'elements' => [], ]; $terms = \explode("\n", $Block['element']['handler']['argument']); foreach ($terms as $term) { $Element['elements'] []= [ 'name' => 'dt', 'handler' => [ 'function' => 'lineElements', 'argument' => $term, 'destination' => 'elements', ], ]; } $Block['element'] = $Element; return $this->addDdElement($Line, $Block); } protected function blockDefinitionListContinue($Line, array $Block) { if ($Line['text'][0] === ':') { return $this->addDdElement($Line, $Block); } else { if (isset($Block['interrupted']) && $Line['indent'] === 0) { return; } if (isset($Block['interrupted'])) { $Block['dd']['handler']['function'] = 'textElements'; $Block['dd']['handler']['argument'] .= "\n\n"; $Block['dd']['handler']['destination'] = 'elements'; unset($Block['interrupted']); } $text = \substr($Line['body'], \min($Line['indent'], 4)); $Block['dd']['handler']['argument'] .= "\n" . $text; return $Block; } } # # Header protected function blockHeaderBase($Line) { $Block = $this->blockHeaderParent($Line); if ($Block !== null && \preg_match('/[ #]*{('.$this->regexAttribute.'+)}[ ]*$/', $Block['element']['handler']['argument'], $matches, \PREG_OFFSET_CAPTURE)) { $attributeString = $matches[1][0]; $Block['element']['attributes'] = $this->parseAttributeData($attributeString); $Block['element']['handler']['argument'] = \substr($Block['element']['handler']['argument'], 0, $matches[0][1]); } return $Block; } # # Markup protected function blockMarkupBase($Line) { if ($this->markupEscaped || $this->safeMode) { return; } if (\preg_match('/^<(\w[\w-]*)(?:[ ]*'.$this->regexHtmlAttribute.')*[ ]*(\/)?>/', $Line['text'], $matches)) { $element = \strtolower($matches[1]); if (\in_array($element, $this->textLevelElements)) { return; } $Block = [ 'name' => $matches[1], 'depth' => 0, 'element' => [ 'rawHtml' => $Line['text'], 'autobreak' => true, ], ]; $length = \strlen($matches[0]); $remainder = \substr($Line['text'], $length); if (\trim($remainder) === '') { if (isset($matches[2]) || \in_array($matches[1], $this->voidElements)) { $Block['closed'] = true; $Block['void'] = true; } } else { if (isset($matches[2]) || \in_array($matches[1], $this->voidElements)) { return; } if (\preg_match('/<\/'.$matches[1].'>[ ]*$/i', $remainder)) { $Block['closed'] = true; } } return $Block; } } protected function blockMarkupContinue($Line, array $Block) { if (isset($Block['closed'])) { return; } if (\preg_match('/^<'.$Block['name'].'(?:[ ]*'.$this->regexHtmlAttribute.')*[ ]*>/i', $Line['text'])) # open { ++$Block['depth']; } if (\preg_match('/(.*?)<\/'.$Block['name'].'>[ ]*$/i', $Line['text'], $matches)) # close { if ($Block['depth'] > 0) { --$Block['depth']; } else { $Block['closed'] = true; } } if (isset($Block['interrupted'])) { $Block['element']['rawHtml'] .= "\n"; unset($Block['interrupted']); } $Block['element']['rawHtml'] .= "\n".$Line['body']; return $Block; } protected function blockMarkupComplete($Block) { if (!isset($Block['void'])) { $Block['element']['rawHtml'] = $this->processTag($Block['element']['rawHtml']); } return $Block; } # # Setext protected function blockSetextHeaderBase($Line, array $Block = null) { $Block = $this->blockSetextHeaderParent($Line, $Block); if ($Block !== null && \preg_match('/[ ]*{('.$this->regexAttribute.'+)}[ ]*$/', $Block['element']['handler']['argument'], $matches, \PREG_OFFSET_CAPTURE)) { $attributeString = $matches[1][0]; $Block['element']['attributes'] = $this->parseAttributeData($attributeString); $Block['element']['handler']['argument'] = \substr($Block['element']['handler']['argument'], 0, $matches[0][1]); } return $Block; } # # Inline Elements # # # Footnote Marker protected function inlineFootnoteMarker($Excerpt) { if (\preg_match('/^\[\^(.+?)\]/', $Excerpt['text'], $matches)) { $name = $matches[1]; if (!isset($this->DefinitionData['Footnote'][$name])) { return; } ++$this->DefinitionData['Footnote'][$name]['count']; if (!isset($this->DefinitionData['Footnote'][$name]['number'])) { $this->DefinitionData['Footnote'][$name]['number'] = ++ $this->footnoteCount; # ยป & } $Element = [ 'name' => 'sup', 'attributes' => ['id' => 'fnref'.$this->DefinitionData['Footnote'][$name]['count'].':'.$name], 'element' => [ 'name' => 'a', 'attributes' => ['href' => '#fn:'.$name, 'class' => 'footnote-ref'], 'text' => $this->DefinitionData['Footnote'][$name]['number'], ], ]; return [ 'extent' => \strlen($matches[0]), 'element' => $Element, ]; } } private $footnoteCount = 0; # # ~ # private $currentAbreviation; private $currentMeaning; protected function insertAbreviation(array $Element) { if (isset($Element['text'])) { $Element['elements'] = self::pregReplaceElements( '/\b'.\preg_quote($this->currentAbreviation, '/').'\b/', [ [ 'name' => 'abbr', 'attributes' => [ 'title' => $this->currentMeaning, ], 'text' => $this->currentAbreviation, ], ], $Element['text'] ); unset($Element['text']); } return $Element; } protected function inlineText($text) { $Inline = $this->inlineTextParent($text); if (isset($this->DefinitionData['Abbreviation'])) { foreach ($this->DefinitionData['Abbreviation'] as $abbreviation => $meaning) { $this->currentAbreviation = $abbreviation; $this->currentMeaning = $meaning; $Inline['element'] = $this->elementApplyRecursiveDepthFirst( [$this, 'insertAbreviation'], $Inline['element'] ); } } return $Inline; } # # Util Methods # protected function addDdElement(array $Line, array $Block) { $text = \substr($Line['text'], 1); $text = \trim($text); unset($Block['dd']); $Block['dd'] = [ 'name' => 'dd', 'handler' => [ 'function' => 'lineElements', 'argument' => $text, 'destination' => 'elements', ], ]; if (isset($Block['interrupted'])) { $Block['dd']['handler']['function'] = 'textElements'; unset($Block['interrupted']); } $Block['element']['elements'] []= & $Block['dd']; return $Block; } protected function buildFootnoteElement() { $Element = [ 'name' => 'div', 'attributes' => ['class' => 'footnotes'], 'elements' => [ ['name' => 'hr'], [ 'name' => 'ol', 'elements' => [], ], ], ]; \uasort($this->DefinitionData['Footnote'], 'self::sortFootnotes'); foreach ($this->DefinitionData['Footnote'] as $definitionId => $DefinitionData) { if (!isset($DefinitionData['number'])) { continue; } $text = $DefinitionData['text']; $textElements = $this->textElements($text); $numbers = \range(1, $DefinitionData['count']); $backLinkElements = []; foreach ($numbers as $number) { $backLinkElements[] = ['text' => ' ']; $backLinkElements[] = [ 'name' => 'a', 'attributes' => [ 'href' => "#fnref{$number}:{$definitionId}", 'rev' => 'footnote', 'class' => 'footnote-backref', ], 'rawHtml' => '↩', 'allowRawHtmlInSafeMode' => true, 'autobreak' => false, ]; } unset($backLinkElements[0]); $n = \count($textElements) - 1; if ($textElements[$n]['name'] === 'p') { $backLinkElements = \array_merge( [ [ 'rawHtml' => ' ', 'allowRawHtmlInSafeMode' => true, ], ], $backLinkElements ); unset($textElements[$n]['name']); $textElements[$n] = [ 'name' => 'p', 'elements' => \array_merge( [$textElements[$n]], $backLinkElements ), ]; } else { $textElements[] = [ 'name' => 'p', 'elements' => $backLinkElements, ]; } $Element['elements'][1]['elements'] []= [ 'name' => 'li', 'attributes' => ['id' => 'fn:'.$definitionId], 'elements' => \array_merge( $textElements ), ]; } return $Element; } # ~ protected function parseAttributeDataBase($attributeString) { $Data = []; $attributes = \preg_split('/[ ]+/', $attributeString, - 1, \PREG_SPLIT_NO_EMPTY); foreach ($attributes as $attribute) { if ($attribute[0] === '#') { $Data['id'] = \substr($attribute, 1); } else # "." { $classes []= \substr($attribute, 1); } } if (isset($classes)) { $Data['class'] = \implode(' ', $classes); } return $Data; } # ~ protected function processTag($elementMarkup) # recursive { # http://stackoverflow.com/q/1148928/200145 \libxml_use_internal_errors(true); $DOMDocument = new \DOMDocument(); # http://stackoverflow.com/q/11309194/200145 $elementMarkup = \mb_convert_encoding($elementMarkup, 'HTML-ENTITIES', 'UTF-8'); # http://stackoverflow.com/q/4879946/200145 $DOMDocument->loadHTML($elementMarkup); $DOMDocument->removeChild($DOMDocument->doctype); $DOMDocument->replaceChild($DOMDocument->firstChild->firstChild->firstChild, $DOMDocument->firstChild); $elementText = ''; if ($DOMDocument->documentElement->getAttribute('markdown') === '1') { foreach ($DOMDocument->documentElement->childNodes as $Node) { $elementText .= $DOMDocument->saveHTML($Node); } $DOMDocument->documentElement->removeAttribute('markdown'); $elementText = "\n".$this->text($elementText)."\n"; } else { foreach ($DOMDocument->documentElement->childNodes as $Node) { $nodeMarkup = $DOMDocument->saveHTML($Node); if ($Node instanceof \DOMElement && ! \in_array($Node->nodeName, $this->textLevelElements)) { $elementText .= $this->processTag($nodeMarkup); } else { $elementText .= $nodeMarkup; } } } # because we don't want for markup to get encoded $DOMDocument->documentElement->nodeValue = 'placeholder\x1A'; $markup = $DOMDocument->saveHTML($DOMDocument->documentElement); return \str_replace('placeholder\x1A', $elementText, $markup); } # ~ protected function sortFootnotes($A, $B) # callback { return $A['number'] - $B['number']; } # # Fields # protected $regexAttribute = '(?:[#.][-\w]+[ ]*)'; protected function textElements($text) { # make sure no definitions are set $this->DefinitionData = []; # standardize line breaks $text = \str_replace(["\r\n", "\r"], "\n", $text); # remove surrounding line breaks $text = \trim($text, "\n"); # split text into lines $lines = \explode("\n", $text); # iterate through lines to identify blocks return $this->linesElements($lines); } # # Setters # public function setBreaksEnabled($breaksEnabled) { $this->breaksEnabled = $breaksEnabled; return $this; } protected $breaksEnabled; public function setMarkupEscaped($markupEscaped) { $this->markupEscaped = $markupEscaped; return $this; } protected $markupEscaped; public function setUrlsLinked($urlsLinked) { $this->urlsLinked = $urlsLinked; return $this; } protected $urlsLinked = true; public function setSafeMode($safeMode) { $this->safeMode = (bool) $safeMode; return $this; } protected $safeMode; public function setStrictMode($strictMode) { $this->strictMode = (bool) $strictMode; return $this; } protected $strictMode; protected $safeLinksWhitelist = [ 'http://', 'https://', 'ftp://', 'ftps://', 'mailto:', 'tel:', 'data:image/png;base64,', 'data:image/gif;base64,', 'data:image/jpeg;base64,', 'irc:', 'ircs:', 'git:', 'ssh:', 'news:', 'steam:', ]; # # Lines # protected $BlockTypes = [ '#' => ['Header'], '*' => ['Rule', 'List', 'Abbreviation'], '+' => ['List'], '-' => ['SetextHeader', 'Table', 'Rule', 'List'], '0' => ['List'], '1' => ['List'], '2' => ['List'], '3' => ['List'], '4' => ['List'], '5' => ['List'], '6' => ['List'], '7' => ['List'], '8' => ['List'], '9' => ['List'], ':' => ['Table', 'DefinitionList'], '<' => ['Comment', 'Markup'], '=' => ['SetextHeader'], '>' => ['Quote'], '[' => ['Footnote', 'Reference'], '_' => ['Rule'], '`' => ['FencedCode'], '|' => ['Table'], '~' => ['FencedCode'], ]; # ~ protected $unmarkedBlockTypes = [ 'Code', ]; # # Blocks # protected function lines(array $lines) { return $this->elements($this->linesElements($lines)); } protected function linesElements(array $lines) { $Elements = []; $CurrentBlock = null; foreach ($lines as $line) { if (\rtrim($line) === '') { if (isset($CurrentBlock)) { $CurrentBlock['interrupted'] = (isset($CurrentBlock['interrupted']) ? $CurrentBlock['interrupted'] + 1 : 1 ); } continue; } while (($beforeTab = \strstr($line, "\t", true)) !== false) { $shortage = 4 - \mb_strlen($beforeTab, 'utf-8') % 4; $line = $beforeTab . \str_repeat(' ', $shortage) . \substr($line, \strlen($beforeTab) + 1) ; } $indent = \strspn($line, ' '); $text = $indent > 0 ? \substr($line, $indent) : $line; # ~ $Line = ['body' => $line, 'indent' => $indent, 'text' => $text]; # ~ if (isset($CurrentBlock['continuable'])) { $methodName = 'block' . $CurrentBlock['type'] . 'Continue'; $Block = $this->{$methodName}($Line, $CurrentBlock); if (isset($Block)) { $CurrentBlock = $Block; continue; } elseif ($this->isBlockCompletable($CurrentBlock['type'])) { $methodName = 'block' . $CurrentBlock['type'] . 'Complete'; $CurrentBlock = $this->{$methodName}($CurrentBlock); } } # ~ $marker = $text[0]; # ~ $blockTypes = $this->unmarkedBlockTypes; if (isset($this->BlockTypes[$marker])) { foreach ($this->BlockTypes[$marker] as $blockType) { $blockTypes []= $blockType; } } # # ~ foreach ($blockTypes as $blockType) { $Block = $this->{"block{$blockType}"}($Line, $CurrentBlock); if (isset($Block)) { $Block['type'] = $blockType; if (!isset($Block['identified'])) { if (isset($CurrentBlock)) { $Elements[] = $this->extractElement($CurrentBlock); } $Block['identified'] = true; } if ($this->isBlockContinuable($blockType)) { $Block['continuable'] = true; } $CurrentBlock = $Block; continue 2; } } # ~ if (isset($CurrentBlock) && $CurrentBlock['type'] === 'Paragraph') { $Block = $this->paragraphContinue($Line, $CurrentBlock); } if (isset($Block)) { $CurrentBlock = $Block; } else { if (isset($CurrentBlock)) { $Elements[] = $this->extractElement($CurrentBlock); } $CurrentBlock = $this->paragraph($Line); $CurrentBlock['identified'] = true; } } # ~ if (isset($CurrentBlock['continuable']) && $this->isBlockCompletable($CurrentBlock['type'])) { $methodName = 'block' . $CurrentBlock['type'] . 'Complete'; $CurrentBlock = $this->{$methodName}($CurrentBlock); } # ~ if (isset($CurrentBlock)) { $Elements[] = $this->extractElement($CurrentBlock); } # ~ return $Elements; } protected function extractElement(array $Component) { if (!isset($Component['element'])) { if (isset($Component['markup'])) { $Component['element'] = ['rawHtml' => $Component['markup']]; } elseif (isset($Component['hidden'])) { $Component['element'] = []; } } return $Component['element']; } protected function isBlockContinuable($Type) : bool { return \method_exists($this, 'block' . $Type . 'Continue'); } protected function isBlockCompletable($Type) : bool { return \method_exists($this, 'block' . $Type . 'Complete'); } # # Code protected function blockCodeBase($Line, $Block = null) { if (isset($Block) && $Block['type'] === 'Paragraph' && !isset($Block['interrupted'])) { return; } if ($Line['indent'] >= 4) { $text = \substr($Line['body'], 4); return [ 'element' => [ 'name' => 'pre', 'element' => [ 'name' => 'code', 'text' => $text, ], ], ]; } } protected function blockCodeContinue($Line, $Block) { if ($Line['indent'] >= 4) { if (isset($Block['interrupted'])) { $Block['element']['element']['text'] .= \str_repeat("\n", $Block['interrupted']); unset($Block['interrupted']); } $Block['element']['element']['text'] .= "\n"; $text = \substr($Line['body'], 4); $Block['element']['element']['text'] .= $text; return $Block; } } protected function blockCodeComplete($Block) { return $Block; } # # Comment protected function blockCommentBase($Line) { if ($this->markupEscaped || $this->safeMode) { return; } if (\strpos($Line['text'], '') !== false) { $Block['closed'] = true; } return $Block; } } protected function blockCommentContinue($Line, array $Block) { if (isset($Block['closed'])) { return; } $Block['element']['rawHtml'] .= "\n" . $Line['body']; if (\strpos($Line['text'], '-->') !== false) { $Block['closed'] = true; } return $Block; } # # Fenced Code protected function blockFencedCodeBase($Line) { $marker = $Line['text'][0]; $openerLength = \strspn($Line['text'], $marker); if ($openerLength < 3) { return; } $infostring = \trim(\substr($Line['text'], $openerLength), "\t "); if (\strpos($infostring, '`') !== false) { return; } $Element = [ 'name' => 'code', 'text' => '', ]; if ($infostring !== '') { /** * https://www.w3.org/TR/2011/WD-html5-20110525/elements.html#classes * Every HTML element may have a class attribute specified. * The attribute, if specified, must have a value that is a set * of space-separated tokens representing the various classes * that the element belongs to. * [...] * The space characters, for the purposes of this specification, * are U+0020 SPACE, U+0009 CHARACTER TABULATION (tab), * U+000A LINE FEED (LF), U+000C FORM FEED (FF), and * U+000D CARRIAGE RETURN (CR). */ $language = \substr($infostring, 0, \strcspn($infostring, " \t\n\f\r")); $Element['attributes'] = ['class' => "language-{$language}"]; } return [ 'char' => $marker, 'openerLength' => $openerLength, 'element' => [ 'name' => 'pre', 'element' => $Element, ], ]; } protected function blockFencedCodeContinue($Line, $Block) { if (isset($Block['complete'])) { return; } if (isset($Block['interrupted'])) { $Block['element']['element']['text'] .= \str_repeat("\n", $Block['interrupted']); unset($Block['interrupted']); } if (($len = \strspn($Line['text'], $Block['char'])) >= $Block['openerLength'] && \rtrim(\substr($Line['text'], $len), ' ') === '' ) { $Block['element']['element']['text'] = \substr($Block['element']['element']['text'], 1); $Block['complete'] = true; return $Block; } $Block['element']['element']['text'] .= "\n" . $Line['body']; return $Block; } protected function blockFencedCodeComplete($Block) { return $Block; } # # Header protected function blockHeaderParent($Line) { $level = \strspn($Line['text'], '#'); if ($level > 6) { return; } $text = \trim($Line['text'], '#'); if ($this->strictMode && isset($text[0]) && $text[0] !== ' ') { return; } $text = \trim($text, ' '); return [ 'element' => [ 'name' => 'h' . $level, 'handler' => [ 'function' => 'lineElements', 'argument' => $text, 'destination' => 'elements', ], ], ]; } # # List protected function blockListBase($Line, array $CurrentBlock = null) { list($name, $pattern) = $Line['text'][0] <= '-' ? ['ul', '[*+-]'] : ['ol', '[0-9]{1,9}+[.\)]']; if (\preg_match('/^('.$pattern.'([ ]++|$))(.*+)/', $Line['text'], $matches)) { $contentIndent = \strlen($matches[2]); if ($contentIndent >= 5) { --$contentIndent; $matches[1] = \substr($matches[1], 0, -$contentIndent); $matches[3] = \str_repeat(' ', $contentIndent) . $matches[3]; } elseif ($contentIndent === 0) { $matches[1] .= ' '; } $markerWithoutWhitespace = \strstr($matches[1], ' ', true); $Block = [ 'indent' => $Line['indent'], 'pattern' => $pattern, 'data' => [ 'type' => $name, 'marker' => $matches[1], 'markerType' => ($name === 'ul' ? $markerWithoutWhitespace : \substr($markerWithoutWhitespace, -1)), ], 'element' => [ 'name' => $name, 'elements' => [], ], ]; $Block['data']['markerTypeRegex'] = \preg_quote($Block['data']['markerType'], '/'); if ($name === 'ol') { $listStart = \ltrim(\strstr($matches[1], $Block['data']['markerType'], true), '0') ?: '0'; if ($listStart !== '1') { if ( isset($CurrentBlock) && $CurrentBlock['type'] === 'Paragraph' && !isset($CurrentBlock['interrupted']) ) { return; } $Block['element']['attributes'] = ['start' => $listStart]; } } $Block['li'] = [ 'name' => 'li', 'handler' => [ 'function' => 'li', 'argument' => empty($matches[3]) ? [] : [$matches[3]], 'destination' => 'elements', ], ]; $Block['element']['elements'] []= & $Block['li']; return $Block; } } protected function blockListContinue($Line, array $Block) { if (isset($Block['interrupted']) && empty($Block['li']['handler']['argument'])) { return null; } $requiredIndent = ($Block['indent'] + \strlen($Block['data']['marker'])); if ($Line['indent'] < $requiredIndent && ( ( $Block['data']['type'] === 'ol' && \preg_match('/^[0-9]++'.$Block['data']['markerTypeRegex'].'(?:[ ]++(.*)|$)/', $Line['text'], $matches) ) || ( $Block['data']['type'] === 'ul' && \preg_match('/^'.$Block['data']['markerTypeRegex'].'(?:[ ]++(.*)|$)/', $Line['text'], $matches) ) ) ) { if (isset($Block['interrupted'])) { $Block['li']['handler']['argument'] []= ''; $Block['loose'] = true; unset($Block['interrupted']); } unset($Block['li']); $text = isset($matches[1]) ? $matches[1] : ''; $Block['indent'] = $Line['indent']; $Block['li'] = [ 'name' => 'li', 'handler' => [ 'function' => 'li', 'argument' => [$text], 'destination' => 'elements', ], ]; $Block['element']['elements'] []= & $Block['li']; return $Block; } elseif ($Line['indent'] < $requiredIndent && $this->blockList($Line)) { return null; } if ($Line['text'][0] === '[' && $this->blockReference($Line)) { return $Block; } if ($Line['indent'] >= $requiredIndent) { if (isset($Block['interrupted'])) { $Block['li']['handler']['argument'] []= ''; $Block['loose'] = true; unset($Block['interrupted']); } $text = \substr($Line['body'], $requiredIndent); $Block['li']['handler']['argument'] []= $text; return $Block; } if (!isset($Block['interrupted'])) { $text = \preg_replace('/^[ ]{0,'.$requiredIndent.'}+/', '', $Line['body']); $Block['li']['handler']['argument'] []= $text; return $Block; } } protected function blockListComplete(array $Block) { if (isset($Block['loose'])) { foreach ($Block['element']['elements'] as &$li) { if (\end($li['handler']['argument']) !== '') { $li['handler']['argument'] []= ''; } } } return $Block; } # # Quote protected function blockQuoteBase($Line) { if (\preg_match('/^>[ ]?+(.*+)/', $Line['text'], $matches)) { return [ 'element' => [ 'name' => 'blockquote', 'handler' => [ 'function' => 'linesElements', 'argument' => (array) $matches[1], 'destination' => 'elements', ], ], ]; } } protected function blockQuoteContinue($Line, array $Block) { if (isset($Block['interrupted'])) { return; } if ($Line['text'][0] === '>' && \preg_match('/^>[ ]?+(.*+)/', $Line['text'], $matches)) { $Block['element']['handler']['argument'] []= $matches[1]; return $Block; } if (!isset($Block['interrupted'])) { $Block['element']['handler']['argument'] []= $Line['text']; return $Block; } } # # Rule protected function blockRuleBase($Line) { $marker = $Line['text'][0]; if (\substr_count($Line['text'], $marker) >= 3 && \rtrim($Line['text'], " {$marker}") === '') { return [ 'element' => [ 'name' => 'hr', ], ]; } } # # Setext protected function blockSetextHeaderParent($Line, array $Block = null) { if (!isset($Block) || $Block['type'] !== 'Paragraph' || isset($Block['interrupted'])) { return; } if ($Line['indent'] < 4 && \rtrim(\rtrim($Line['text'], ' '), $Line['text'][0]) === '') { $Block['element']['name'] = $Line['text'][0] === '=' ? 'h1' : 'h2'; return $Block; } } # # Reference protected function blockReferenceBase($Line) { if (\strpos($Line['text'], ']') !== false && \preg_match('/^\[(.+?)\]:[ ]*+(\S+?)>?(?:[ ]+["\'(](.+)["\')])?[ ]*+$/', $Line['text'], $matches) ) { $id = \strtolower($matches[1]); $Data = [ 'url' => UriFactory::build($matches[2]), 'title' => isset($matches[3]) ? $matches[3] : null, ]; $this->DefinitionData['Reference'][$id] = $Data; return [ 'element' => [], ]; } } # # Table protected function blockTableBase($Line, array $Block = null) { if (!isset($Block) || $Block['type'] !== 'Paragraph' || isset($Block['interrupted'])) { return; } if ( \strpos($Block['element']['handler']['argument'], '|') === false && \strpos($Line['text'], '|') === false && \strpos($Line['text'], ':') === false || \strpos($Block['element']['handler']['argument'], "\n") !== false ) { return; } if (\rtrim($Line['text'], ' -:|') !== '') { return; } $alignments = []; $divider = $Line['text']; $divider = \trim($divider); $divider = \trim($divider, '|'); $dividerCells = \explode('|', $divider); foreach ($dividerCells as $dividerCell) { $dividerCell = \trim($dividerCell); if ($dividerCell === '') { return; } $alignment = null; if ($dividerCell[0] === ':') { $alignment = 'left'; } if (\substr($dividerCell, - 1) === ':') { $alignment = $alignment === 'left' ? 'center' : 'right'; } $alignments []= $alignment; } # ~ $HeaderElements = []; $header = $Block['element']['handler']['argument']; $header = \trim($header); $header = \trim($header, '|'); $headerCells = \explode('|', $header); if (\count($headerCells) !== \count($alignments)) { return; } foreach ($headerCells as $index => $headerCell) { $headerCell = \trim($headerCell); $HeaderElement = [ 'name' => 'th', 'handler' => [ 'function' => 'lineElements', 'argument' => $headerCell, 'destination' => 'elements', ], ]; if (isset($alignments[$index])) { $alignment = $alignments[$index]; $HeaderElement['attributes'] = [ 'style' => "text-align: {$alignment};", ]; } $HeaderElements []= $HeaderElement; } # ~ $Block = [ 'alignments' => $alignments, 'identified' => true, 'element' => [ 'name' => 'table', 'elements' => [], ], ]; $Block['element']['elements'] []= [ 'name' => 'thead', ]; $Block['element']['elements'] []= [ 'name' => 'tbody', 'elements' => [], ]; $Block['element']['elements'][0]['elements'] []= [ 'name' => 'tr', 'elements' => $HeaderElements, ]; return $Block; } protected function blockTableContinue($Line, array $Block) { if (isset($Block['interrupted'])) { return; } if (\count($Block['alignments']) === 1 || $Line['text'][0] === '|' || \strpos($Line['text'], '|')) { $Elements = []; $row = $Line['text']; $row = \trim($row); $row = \trim($row, '|'); \preg_match_all('/(?:(\\\\[|])|[^|`]|`[^`]++`|`)++/', $row, $matches); $cells = \array_slice($matches[0], 0, \count($Block['alignments'])); foreach ($cells as $index => $cell) { $cell = \trim($cell); $Element = [ 'name' => 'td', 'handler' => [ 'function' => 'lineElements', 'argument' => $cell, 'destination' => 'elements', ], ]; if (isset($Block['alignments'][$index])) { $Element['attributes'] = [ 'style' => 'text-align: ' . $Block['alignments'][$index] . ';', ]; } $Elements []= $Element; } $Element = [ 'name' => 'tr', 'elements' => $Elements, ]; $Block['element']['elements'][1]['elements'] []= $Element; return $Block; } } # # ~ # protected function paragraph($Line) { return [ 'type' => 'Paragraph', 'element' => [ 'name' => 'p', 'handler' => [ 'function' => 'lineElements', 'argument' => $Line['text'], 'destination' => 'elements', ], ], ]; } protected function paragraphContinue($Line, array $Block) { if (isset($Block['interrupted'])) { return; } $Block['element']['handler']['argument'] .= "\n".$Line['text']; return $Block; } # # Inline Elements # protected $InlineTypes = [ '!' => ['Image'], '&' => ['SpecialCharacter'], '*' => ['Emphasis'], ':' => ['Url'], '<' => ['UrlTag', 'EmailTag', 'Markup'], '[' => ['FootnoteMarker', 'Link'], '_' => ['Emphasis'], '`' => ['Code'], '~' => ['Strikethrough'], '\\' => ['EscapeSequence'], ]; # ~ protected $inlineMarkerList = '!*_&[:<`~\\'; # # ~ # public function line($text, $nonNestables = []) { return $this->elements($this->lineElements($text, $nonNestables)); } /* protected function lineElements($text, $nonNestables = array()) { # standardize line breaks $text = str_replace(array("\r\n", "\r"), "\n", $text); $Elements = array(); $nonNestables = (empty($nonNestables) ? array() : array_combine($nonNestables, $nonNestables) ); # $excerpt is based on the first occurrence of a marker while ($excerpt = strpbrk($text, $this->inlineMarkerList)) { $marker = $excerpt[0]; $markerPosition = strlen($text) - strlen($excerpt); $Excerpt = array('text' => $excerpt, 'context' => $text); foreach ($this->InlineTypes[$marker] as $inlineType) { # check to see if the current inline type is nestable in the current context if (isset($nonNestables[$inlineType])) { continue; } $Inline = $this->{"inline$inlineType"}($Excerpt); if ( !isset($Inline)) { continue; } # makes sure that the inline belongs to "our" marker if (isset($Inline['position']) and $Inline['position'] > $markerPosition) { continue; } # sets a default inline position if ( !isset($Inline['position'])) { $Inline['position'] = $markerPosition; } # cause the new element to 'inherit' our non nestables $Inline['element']['nonNestables'] = isset($Inline['element']['nonNestables']) ? array_merge($Inline['element']['nonNestables'], $nonNestables) : $nonNestables ; # the text that comes before the inline $unmarkedText = substr($text, 0, $Inline['position']); # compile the unmarked text $InlineText = $this->inlineText($unmarkedText); $Elements[] = $InlineText['element']; # compile the inline $Elements[] = $this->extractElement($Inline); # remove the examined text $text = substr($text, $Inline['position'] + $Inline['extent']); continue 2; } # the marker does not belong to an inline $unmarkedText = substr($text, 0, $markerPosition + 1); $InlineText = $this->inlineText($unmarkedText); $Elements[] = $InlineText['element']; $text = substr($text, $markerPosition + 1); } $InlineText = $this->inlineText($text); $Elements[] = $InlineText['element']; foreach ($Elements as &$Element) { if ( !isset($Element['autobreak'])) { $Element['autobreak'] = false; } } return $Elements; } */ # # ~ # protected function inlineTextParent($text) { $Inline = [ 'extent' => \strlen($text), 'element' => [], ]; $Inline['element']['elements'] = self::pregReplaceElements( $this->breaksEnabled ? '/[ ]*+\n/' : '/(?:[ ]*+\\\\|[ ]{2,}+)\n/', [ ['name' => 'br'], ['text' => "\n"], ], $text ); return $Inline; } protected function inlineLinkParent($Excerpt) { $Element = [ 'name' => 'a', 'handler' => [ 'function' => 'lineElements', 'argument' => null, 'destination' => 'elements', ], 'nonNestables' => ['Url', 'Link'], 'attributes' => [ 'href' => null, 'title' => null, ], ]; $extent = 0; $remainder = $Excerpt['text']; if (\preg_match('/\[((?:[^][]++|(?R))*+)\]/', $remainder, $matches)) { $Element['handler']['argument'] = $matches[1]; $extent += \strlen($matches[0]); $remainder = \substr($remainder, $extent); } else { return; } if (\preg_match('/^[(]\s*+((?:[^ ()]++|[(][^ )]+[)])++)(?:[ ]+("[^"]*+"|\'[^\']*+\'))?\s*+[)]/', $remainder, $matches)) { $Element['attributes']['href'] = $matches[1]; if (isset($matches[2])) { $Element['attributes']['title'] = \substr($matches[2], 1, - 1); } $extent += \strlen($matches[0]); } else { if (\preg_match('/^\s*\[(.*?)\]/', $remainder, $matches)) { $definition = \strlen($matches[1]) !== 0 ? $matches[1] : $Element['handler']['argument']; $definition = \strtolower($definition); $extent += \strlen($matches[0]); } else { $definition = \strtolower($Element['handler']['argument']); } if (!isset($this->DefinitionData['Reference'][$definition])) { return; } $Definition = $this->DefinitionData['Reference'][$definition]; $Element['attributes']['href'] = $Definition['url']; $Element['attributes']['title'] = $Definition['title']; } return [ 'extent' => $extent, 'element' => $Element, ]; } protected function inlineSpecialCharacter($Excerpt) { if (\substr($Excerpt['text'], 1, 1) !== ' ' && \strpos($Excerpt['text'], ';') !== false && \preg_match('/^&(#?+[0-9a-zA-Z]++);/', $Excerpt['text'], $matches) ) { return [ 'element' => ['rawHtml' => '&' . $matches[1] . ';'], 'extent' => \strlen($matches[0]), ]; } } # ~ protected function unmarkedText($text) { $Inline = $this->inlineText($text); return $this->element($Inline['element']); } # # Handlers # protected function handle(array $Element) { if (isset($Element['handler'])) { if (!isset($Element['nonNestables'])) { $Element['nonNestables'] = []; } if (\is_string($Element['handler'])) { $function = $Element['handler']; $argument = $Element['text']; unset($Element['text']); $destination = 'rawHtml'; } else { $function = $Element['handler']['function']; $argument = $Element['handler']['argument']; $destination = $Element['handler']['destination']; } $Element[$destination] = $this->{$function}($argument, $Element['nonNestables']); if ($destination === 'handler') { $Element = $this->handle($Element); } unset($Element['handler']); } return $Element; } protected function handleElementRecursive(array $Element) { return $this->elementApplyRecursive([$this, 'handle'], $Element); } protected function handleElementsRecursive(array $Elements) { return $this->elementsApplyRecursive([$this, 'handle'], $Elements); } protected function elementApplyRecursive($closure, array $Element) { $Element = $closure($Element); if (isset($Element['elements'])) { $Element['elements'] = $this->elementsApplyRecursive($closure, $Element['elements']); } elseif (isset($Element['element'])) { $Element['element'] = $this->elementApplyRecursive($closure, $Element['element']); } return $Element; } protected function elementApplyRecursiveDepthFirst($closure, array $Element) { if (isset($Element['elements'])) { $Element['elements'] = $this->elementsApplyRecursiveDepthFirst($closure, $Element['elements']); } elseif (isset($Element['element'])) { $Element['element'] = $this->elementsApplyRecursiveDepthFirst($closure, $Element['element']); } return $closure($Element); } protected function elementsApplyRecursive($closure, array $Elements) { foreach ($Elements as &$Element) { $Element = $this->elementApplyRecursive($closure, $Element); } return $Elements; } protected function elementsApplyRecursiveDepthFirst($closure, array $Elements) { foreach ($Elements as &$Element) { $Element = $this->elementApplyRecursiveDepthFirst($closure, $Element); } return $Elements; } protected function element(array $Element) { if ($this->safeMode) { $Element = $this->sanitiseElement($Element); } # identity map if element has no handler $Element = $this->handle($Element); $hasName = isset($Element['name']); $markup = ''; if ($hasName) { $markup .= '<' . $Element['name']; if (isset($Element['attributes'])) { foreach ($Element['attributes'] as $name => $value) { if ($value === null) { continue; } $markup .= " {$name}=\"".self::escape($value).'"'; } } } $permitRawHtml = false; if (isset($Element['text'])) { $text = $Element['text']; } // very strongly consider an alternative if you're writing an // extension elseif (isset($Element['rawHtml'])) { $text = $Element['rawHtml']; $allowRawHtmlInSafeMode = isset($Element['allowRawHtmlInSafeMode']) && $Element['allowRawHtmlInSafeMode']; $permitRawHtml = !$this->safeMode || $allowRawHtmlInSafeMode; } $hasContent = isset($text) || isset($Element['element']) || isset($Element['elements']); if ($hasContent) { $markup .= $hasName ? '>' : ''; if (isset($Element['elements'])) { $markup .= $this->elements($Element['elements']); } elseif (isset($Element['element'])) { $markup .= $this->element($Element['element']); } elseif (!$permitRawHtml) { $markup .= self::escape((string) $text, true); } else { $markup .= $text; } $markup .= $hasName ? '' . $Element['name'] . '>' : ''; } elseif ($hasName) { $markup .= ' />'; } return $markup; } protected function elements(array $Elements) : string { $markup = ''; $autoBreak = true; foreach ($Elements as $Element) { if (empty($Element)) { continue; } $autoBreakNext = (isset($Element['autobreak']) ? $Element['autobreak'] : isset($Element['name']) ); // (autobreak === false) covers both sides of an element $autoBreak = $autoBreak ? $autoBreakNext : $autoBreak; $markup .= ($autoBreak ? "\n" : '') . $this->element($Element); $autoBreak = $autoBreakNext; } return $markup . ($autoBreak ? "\n" : ''); } # ~ protected function li($lines) { $Elements = $this->linesElements($lines); if (! \in_array('', $lines) && isset($Elements[0], $Elements[0]['name']) && $Elements[0]['name'] === 'p' ) { unset($Elements[0]['name']); } return $Elements; } # # AST Convenience # /** * Replace occurrences $regexp with $Elements in $text. Return an array of * elements representing the replacement. */ protected static function pregReplaceElements($regexp, $Elements, $text) { $newElements = []; while (\preg_match($regexp, $text, $matches, \PREG_OFFSET_CAPTURE)) { $offset = $matches[0][1]; $before = \substr($text, 0, $offset); $after = \substr($text, $offset + \strlen($matches[0][0])); $newElements[] = ['text' => $before]; foreach ($Elements as $Element) { $newElements[] = $Element; } $text = $after; } $newElements[] = ['text' => $text]; return $newElements; } # # Deprecated Methods # public static function parse($text) { $parsedown = new self(); return $parsedown->text($text); } protected function sanitiseElement(array $Element) { static $goodAttribute = '/^[a-zA-Z0-9][a-zA-Z0-9-_]*+$/'; static $safeUrlNameToAtt = [ 'a' => 'href', 'img' => 'src', ]; if (!isset($Element['name'])) { unset($Element['attributes']); return $Element; } if (isset($safeUrlNameToAtt[$Element['name']])) { $Element = $this->filterUnsafeUrlInAttribute($Element, $safeUrlNameToAtt[$Element['name']]); } if (! empty($Element['attributes'])) { foreach ($Element['attributes'] as $att => $val) { # filter out badly parsed attribute if (! \preg_match($goodAttribute, $att)) { unset($Element['attributes'][$att]); } # dump onevent attribute elseif (self::striAtStart($att, 'on')) { unset($Element['attributes'][$att]); } } } return $Element; } protected function filterUnsafeUrlInAttribute(array $Element, $attribute) { foreach ($this->safeLinksWhitelist as $scheme) { if (self::striAtStart($Element['attributes'][$attribute], $scheme)) { return $Element; } } $Element['attributes'][$attribute] = \str_replace(':', '%3A', $Element['attributes'][$attribute]); return $Element; } # # Static Methods # protected static function escape(string $text, bool $allowQuotes = false) : string { return \htmlspecialchars($text, $allowQuotes ? \ENT_NOQUOTES : \ENT_QUOTES, 'UTF-8'); } protected static function striAtStart($string, $needle) { $len = \strlen($needle); if ($len > \strlen($string)) { return false; } else { return \strtolower(\substr($string, 0, $len)) === \strtolower($needle); } } public static function instance($name = 'default') { if (isset(self::$instances[$name])) { return self::$instances[$name]; } $instance = new static(); self::$instances[$name] = $instance; return $instance; } private static $instances = []; # # Fields # protected $DefinitionData; public const ID_ATTRIBUTE_DEFAULT = 'toc'; protected $tagToc = '[toc]'; protected $contentsListArray = []; protected $contentsListString = ''; protected $firstHeadLevel = 0; protected $isBlacklistInitialized = false; protected $anchorDuplicates = []; # # Read-Only protected $specialCharacters = [ '\\', '`', '*', '_', '{', '}', '[', ']', '(', ')', '>', '#', '+', '-', '.', '!', '|', '?', '"', "'", '<', ]; protected $StrongRegex = [ '*' => '/^[*]{2}((?:\\\\\*|[^*]|[*][^*]*+[*])+?)[*]{2}(?![*])/s', ]; protected $UnderlineRegex = [ '_' => '/^__((?:\\\\_|[^_]|_[^_]*+_)+?)__(?!_)/us', ]; protected $EmRegex = [ '*' => '/^[*]((?:\\\\\*|[^*]|[*][*][^*]+?[*][*])+?)[*](?![*])/s', '_' => '/^_((?:\\\\_|[^_]|__[^_]*__)+?)_(?!_)\b/us', ]; protected $regexHtmlAttribute = '[a-zA-Z_:][\w:.-]*+(?:\s*+=\s*+(?:[^"\'=<>`\s]+|"[^"]*+"|\'[^\']*+\'))?+'; protected $voidElements = [ 'area', 'base', 'br', 'col', 'command', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', ]; protected $textLevelElements = [ 'a', 'br', 'bdo', 'abbr', 'blink', 'nextid', 'acronym', 'basefont', 'b', 'em', 'big', 'cite', 'small', 'spacer', 'listing', 'i', 'rp', 'del', 'code', 'strike', 'marquee', 'q', 'rt', 'ins', 'font', 'strong', 's', 'tt', 'kbd', 'mark', 'u', 'xm', 'sub', 'nobr', 'sup', 'ruby', 'var', 'span', 'wbr', 'time', ]; }