breaks_enabled = $breaks_enabled; return $this; } # # Fields # private $reference_map = array(); # # Methods # # # Parsedown tries to read Markdown texts the way humans do. First, it breaks # texts into lines. Then, it identifies blocks by looking at how these lines # start and relate to each other. Finally, it identifies inline elements. function parse($text) { # standardize line breaks $text = str_replace("\r\n", "\n", $text); $text = str_replace("\r", "\n", $text); # replace tabs with spaces $text = str_replace("\t", ' ', $text); # remove surrounding line breaks $text = trim($text, "\n"); # split text into lines $lines = explode("\n", $text); # convert lines into html $text = $this->parse_block_elements($lines); # remove trailing line breaks $text = chop($text, "\n"); return $text; } # # Private # private function parse_block_elements(array $lines, $context = '') { $blocks = array(); $block = array( 'type' => '', ); foreach ($lines as $line) { # fenced blocks switch ($block['type']) { case 'fenced block': if ( ! isset($block['closed'])) { if (preg_match('/^[ ]*'.$block['fence'][0].'{3,}[ ]*$/', $line)) { $block['closed'] = true; } else { if ($block['text'] !== '') { $block['text'] .= "\n"; } $block['text'] .= $line; } continue 2; } break; case 'block-level markup': if ( ! isset($block['closed'])) { if (strpos($line, $block['start']) !== false) # opening tag { $block['depth']++; } if (strpos($line, $block['end']) !== false) # closing tag { $block['depth'] > 0 ? $block['depth']-- : $block['closed'] = true; } $block['text'] .= "\n".$line; continue 2; } break; } # * $deindented_line = ltrim($line); if ($deindented_line === '') { $block['interrupted'] = true; continue; } # composite blocks switch ($block['type']) { case 'blockquote': if ( ! isset($block['interrupted'])) { $line = preg_replace('/^[ ]*>[ ]?/', '', $line); $block['lines'] []= $line; continue 2; } break; case 'li': if (preg_match('/^([ ]{0,3})(\d+[.]|[*+-])[ ](.*)/', $line, $matches)) { if ($block['indentation'] !== $matches[1]) { $block['lines'] []= $line; } else { unset($block['last']); $blocks []= $block; unset($block['first']); $block['last'] = true; $block['lines'] = array( preg_replace('/^[ ]{0,4}/', '', $matches[3]), ); } continue 2; } if (isset($block['interrupted'])) { if ($line[0] === ' ') { $block['lines'] []= ''; $line = preg_replace('/^[ ]{0,4}/', '', $line); $block['lines'] []= $line; unset($block['interrupted']); continue 2; } } else { $line = preg_replace('/^[ ]{0,4}/', '', $line); $block['lines'] []= $line; continue 2; } break; } # indentation sensitive types switch ($line[0]) { case ' ': # code block if (isset($line[3]) and $line[3] === ' ' and $line[2] === ' ' and $line[1] === ' ') { $code_line = substr($line, 4); if ($block['type'] === 'code block') { if (isset($block['interrupted'])) { $block['text'] .= "\n"; unset($block['interrupted']); } $block['text'] .= "\n".$code_line; } else { $blocks []= $block; $block = array( 'type' => 'code block', 'text' => $code_line, ); } continue 2; } break; case '#': # atx heading (#) if (isset($line[1])) { $blocks []= $block; $level = 1; while (isset($line[$level]) and $line[$level] === '#') { $level++; } $block = array( 'type' => 'heading', 'text' => trim($line, '# '), 'level' => $level, ); continue 2; } break; case '-': case '=': # setext heading if ($block['type'] === 'paragraph' and isset($block['interrupted']) === false) { $chopped_line = chop($line); $i = 1; while (isset($chopped_line[$i])) { if ($chopped_line[$i] !== $line[0]) { break 2; } $i++; } $block['type'] = 'heading'; $block['level'] = $line[0] === '-' ? 2 : 1; continue 2; } break; } # indentation insensitive types switch ($deindented_line[0]) { case '<': $position = strpos($deindented_line, '>'); if ($position > 1) # tag { $name = substr($deindented_line, 1, $position - 1); $name = chop($name); if (substr($name, -1) === '/') { $self_closing = true; $name = substr($name, 0, -1); } $position = strpos($name, ' '); if ($position) { $name = substr($name, 0, $position); } if ( ! ctype_alpha($name)) { break; } if (in_array($name, $this->inline_tags)) { break; } $blocks []= $block; if (isset($self_closing)) { $block = array( 'type' => 'self-closing tag', 'text' => $deindented_line, ); unset($self_closing); continue 2; } $block = array( 'type' => 'block-level markup', 'text' => $deindented_line, 'start' => '<'.$name.'>', 'end' => ''.$name.'>', 'depth' => 0, ); if (strpos($deindented_line, $block['end'])) { $block['closed'] = true; } continue 2; } break; case '>': # quote if (preg_match('/^>[ ]?(.*)/', $deindented_line, $matches)) { $blocks []= $block; $block = array( 'type' => 'blockquote', 'lines' => array( $matches[1], ), ); continue 2; } break; case '[': # reference if (preg_match('/^\[(.+?)\]:[ ]*(.+?)(?:[ ]+[\'"](.+?)[\'"])?[ ]*$/', $deindented_line, $matches)) { $label = strtolower($matches[1]); $this->reference_map[$label] = array( '»' => trim($matches[2], '<>'), ); if (isset($matches[3])) { $this->reference_map[$label]['#'] = $matches[3]; } continue 2; } break; case '`': case '~': # fenced code block if (preg_match('/^([`]{3,}|[~]{3,})[ ]*(\S+)?[ ]*$/', $deindented_line, $matches)) { $blocks []= $block; $block = array( 'type' => 'fenced block', 'text' => '', 'fence' => $matches[1], ); if (isset($matches[2])) { $block['language'] = $matches[2]; } continue 2; } break; case '*': case '+': case '-': case '_': # hr if (preg_match('/^([-*_])([ ]{0,2}\1){2,}[ ]*$/', $deindented_line)) { $blocks []= $block; $block = array( 'type' => 'hr', ); continue 2; } # li if (preg_match('/^([ ]*)[*+-][ ](.*)/', $line, $matches)) { $blocks []= $block; $block = array( 'type' => 'li', 'ordered' => false, 'indentation' => $matches[1], 'first' => true, 'last' => true, 'lines' => array( preg_replace('/^[ ]{0,4}/', '', $matches[2]), ), ); continue 2; } } # li if ($deindented_line[0] <= '9' and $deindented_line[0] >= '0' and preg_match('/^([ ]*)\d+[.][ ](.*)/', $line, $matches)) { $blocks []= $block; $block = array( 'type' => 'li', 'ordered' => true, 'indentation' => $matches[1], 'first' => true, 'last' => true, 'lines' => array( preg_replace('/^[ ]{0,4}/', '', $matches[2]), ), ); continue; } # paragraph if ($block['type'] === 'paragraph') { if (isset($block['interrupted'])) { $blocks []= $block; $block['text'] = $line; unset($block['interrupted']); } else { if ($this->breaks_enabled) { $block['text'] .= ' '; } $block['text'] .= "\n".$line; } } else { $blocks []= $block; $block = array( 'type' => 'paragraph', 'text' => $line, ); } } $blocks []= $block; unset($blocks[0]); # # ~ # $markup = ''; foreach ($blocks as $block) { switch ($block['type']) { case 'paragraph': $text = $this->parse_span_elements($block['text']); if ($context === 'li' and $markup === '') { if (isset($block['interrupted'])) { $markup .= "\n".'
'.$text.'
'."\n"; } else { $markup .= $text; if (isset($blocks[2])) { $markup .= "\n"; } } } else { $markup .= ''.$text.'
'."\n"; } break; case 'blockquote': $text = $this->parse_block_elements($block['lines']); $markup .= ''."\n".$text.''."\n"; break; case 'code block': $text = htmlspecialchars($block['text'], ENT_NOQUOTES, 'UTF-8'); $markup .= '
'.$text.'
'."\n";
break;
case 'fenced block':
$text = htmlspecialchars($block['text'], ENT_NOQUOTES, 'UTF-8');
$markup .= '
'."\n";
break;
case 'heading':
$text = $this->parse_span_elements($block['text']);
$markup .= ''.$element_text.'
';
$offset = strlen($matches[0]);
}
else
{
$markup .= '`';
$offset = 1;
}
break;
case 'http':
if (preg_match('/^https?:[\/]{2}[^\s]+\b/i', $text, $matches))
{
$element_url = $matches[0];
$element_url = str_replace('&', '&', $element_url);
$element_url = str_replace('<', '<', $element_url);
$markup .= ''.$element_url.'';
$offset = strlen($matches[0]);
}
else
{
$markup .= 'http';
$offset = 4;
}
break;
case '~~':
if (preg_match('/^~~(?=\S)(.+?)(?<=\S)~~/', $text, $matches))
{
$matches[1] = $this->parse_span_elements($matches[1], $markers);
$markup .= '