* Hello World. <-- Is this a Markdown code block or text?
*
<-- Is this a Markdown code block or a real tag?
*
*
* If you don't like this, just don't indent the tag on which
* you apply the markdown="1" attribute.
*
* * If $enclosing_tag_re is not empty, stops at the first unmatched closing
* tag with that name. Nested tags supported.
*
* * If $span is true, text inside must treated as span. So any double
* newline will be replaced by a single newline so that it does not create
* paragraphs.
*
* Returns an array of that form: ( processed text , remaining text )
*
* @param string $text
* @param integer $indent
* @param string $enclosing_tag_re
* @param boolean $span
* @return array
*/
protected function _hashHTMLBlocks_inMarkdown($text, $indent = 0,
$enclosing_tag_re = '', $span = false)
{
if ($text === '') return array('', '');
// Regex to check for the presense of newlines around a block tag.
$newline_before_re = '/(?:^\n?|\n\n)*$/';
$newline_after_re =
'{
^ # Start of text following the tag.
(?>[ ]*)? # Optional comment.
[ ]*\n # Must be followed by newline.
}xs';
// Regex to match any tag.
$block_tag_re =
'{
( # $2: Capture whole tag.
# Tag name.
' . $this->block_tags_re . ' |
' . $this->context_block_tags_re . ' |
' . $this->clean_tags_re . ' |
(?!\s)'.$enclosing_tag_re . '
)
(?:
(?=[\s"\'/a-zA-Z0-9]) # Allowed characters after tag name.
(?>
".*?" | # Double quotes (can contain `>`)
\'.*?\' | # Single quotes (can contain `>`)
.+? # Anything but quotes and `>`.
)*?
)?
> # End of tag.
|
# HTML Comment
|
<\?.*?\?> | <%.*?%> # Processing instruction
|
# CData Block
' . ( !$span ? ' # If not in span.
|
# Indented code block
(?: ^[ ]*\n | ^ | \n[ ]*\n )
[ ]{' . ($indent + 4) . '}[^\n]* \n
(?>
(?: [ ]{' . ($indent + 4) . '}[^\n]* | [ ]* ) \n
)*
|
# Fenced code block marker
(?<= ^ | \n )
[ ]{0,' . ($indent + 3) . '}(?:~{3,}|`{3,})
[ ]*
(?: \.?[-_:a-zA-Z0-9]+ )? # standalone class name
[ ]*
(?: ' . $this->id_class_attr_nocatch_re . ' )? # extra attributes
[ ]*
(?= \n )
' : '' ) . ' # End (if not is span).
|
# Code span marker
# Note, this regex needs to go after backtick fenced
# code blocks but it should also be kept outside of the
# "if not in span" condition adding backticks to the parser
`+
)
}xs';
$depth = 0; // Current depth inside the tag tree.
$parsed = ""; // Parsed text that will be returned.
// Loop through every tag until we find the closing tag of the parent
// or loop until reaching the end of text if no parent tag specified.
do {
// Split the text using the first $tag_match pattern found.
// Text before pattern will be first in the array, text after
// pattern will be at the end, and between will be any catches made
// by the pattern.
$parts = preg_split($block_tag_re, $text, 2,
PREG_SPLIT_DELIM_CAPTURE);
// If in Markdown span mode, add a empty-string span-level hash
// after each newline to prevent triggering any block element.
if ($span) {
$void = $this->hashPart("", ':');
$newline = "\n$void";
$parts[0] = $void . str_replace("\n", $newline, $parts[0]) . $void;
}
$parsed .= $parts[0]; // Text before current tag.
// If end of $text has been reached. Stop loop.
if (count($parts) < 3) {
$text = "";
break;
}
$tag = $parts[1]; // Tag to handle.
$text = $parts[2]; // Remaining text after current tag.
$tag_re = preg_quote($tag); // For use in a regular expression.
// Check for: Fenced code block marker.
// Note: need to recheck the whole tag to disambiguate backtick
// fences from code spans
if (preg_match('{^\n?([ ]{0,' . ($indent + 3) . '})(~{3,}|`{3,})[ ]*(?:\.?[-_:a-zA-Z0-9]+)?[ ]*(?:' . $this->id_class_attr_nocatch_re . ')?[ ]*\n?$}', $tag, $capture)) {
// Fenced code block marker: find matching end marker.
$fence_indent = strlen($capture[1]); // use captured indent in re
$fence_re = $capture[2]; // use captured fence in re
if (preg_match('{^(?>.*\n)*?[ ]{' . ($fence_indent) . '}' . $fence_re . '[ ]*(?:\n|$)}', $text,
$matches))
{
// End marker found: pass text unchanged until marker.
$parsed .= $tag . $matches[0];
$text = substr($text, strlen($matches[0]));
}
else {
// No end marker: just skip it.
$parsed .= $tag;
}
}
// Check for: Indented code block.
else if ($tag{0} == "\n" || $tag{0} == " ") {
// Indented code block: pass it unchanged, will be handled
// later.
$parsed .= $tag;
}
// Check for: Code span marker
// Note: need to check this after backtick fenced code blocks
else if ($tag{0} == "`") {
// Find corresponding end marker.
$tag_re = preg_quote($tag);
if (preg_match('{^(?>.+?|\n(?!\n))*?(?block_tags_re . ')\b}', $tag) ||
( preg_match('{^<(?:' . $this->context_block_tags_re . ')\b}', $tag) &&
preg_match($newline_before_re, $parsed) &&
preg_match($newline_after_re, $text) )
)
{
// Need to parse tag and following text using the HTML parser.
list($block_text, $text) =
$this->_hashHTMLBlocks_inHTML($tag . $text, "hashBlock", true);
// Make sure it stays outside of any paragraph by adding newlines.
$parsed .= "\n\n$block_text\n\n";
}
// Check for: Clean tag (like script, math)
// HTML Comments, processing instructions.
else if (preg_match('{^<(?:' . $this->clean_tags_re . ')\b}', $tag) ||
$tag{1} == '!' || $tag{1} == '?')
{
// Need to parse tag and following text using the HTML parser.
// (don't check for markdown attribute)
list($block_text, $text) =
$this->_hashHTMLBlocks_inHTML($tag . $text, "hashClean", false);
$parsed .= $block_text;
}
// Check for: Tag with same name as enclosing tag.
else if ($enclosing_tag_re !== '' &&
// Same name as enclosing tag.
preg_match('{^= 0);
return array($parsed, $text);
}
/**
* Parse HTML, calling _HashHTMLBlocks_InMarkdown for block tags.
*
* * Calls $hash_method to convert any blocks.
* * Stops when the first opening tag closes.
* * $md_attr indicate if the use of the `markdown="1"` attribute is allowed.
* (it is not inside clean tags)
*
* Returns an array of that form: ( processed text , remaining text )
* @param string $text
* @param string $hash_method
* @param string $md_attr
* @return array
*/
protected function _hashHTMLBlocks_inHTML($text, $hash_method, $md_attr) {
if ($text === '') return array('', '');
// Regex to match `markdown` attribute inside of a tag.
$markdown_attr_re = '
{
\s* # Eat whitespace before the `markdown` attribute
markdown
\s*=\s*
(?>
(["\']) # $1: quote delimiter
(.*?) # $2: attribute value
\1 # matching delimiter
|
([^\s>]*) # $3: unquoted attribute value
)
() # $4: make $3 always defined (avoid warnings)
}xs';
// Regex to match any tag.
$tag_re = '{
( # $2: Capture whole tag.
".*?" | # Double quotes (can contain `>`)
\'.*?\' | # Single quotes (can contain `>`)
.+? # Anything but quotes and `>`.
)*?
)?
> # End of tag.
|
# HTML Comment
|
<\?.*?\?> | <%.*?%> # Processing instruction
|
# CData Block
)
}xs';
$original_text = $text; // Save original text in case of faliure.
$depth = 0; // Current depth inside the tag tree.
$block_text = ""; // Temporary text holder for current text.
$parsed = ""; // Parsed text that will be returned.
// Get the name of the starting tag.
// (This pattern makes $base_tag_name_re safe without quoting.)
if (preg_match('/^<([\w:$]*)\b/', $text, $matches))
$base_tag_name_re = $matches[1];
// Loop through every tag until we find the corresponding closing tag.
do {
// Split the text using the first $tag_match pattern found.
// Text before pattern will be first in the array, text after
// pattern will be at the end, and between will be any catches made
// by the pattern.
$parts = preg_split($tag_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE);
if (count($parts) < 3) {
// End of $text reached with unbalenced tag(s).
// In that case, we return original text unchanged and pass the
// first character as filtered to prevent an infinite loop in the
// parent function.
return array($original_text{0}, substr($original_text, 1));
}
$block_text .= $parts[0]; // Text before current tag.
$tag = $parts[1]; // Tag to handle.
$text = $parts[2]; // Remaining text after current tag.
// Check for: Auto-close tag (like
) // Comments and Processing Instructions. if (preg_match('{^auto_close_tags_re . ')\b}', $tag) || $tag{1} == '!' || $tag{1} == '?') { // Just add the tag to the block as if it was text. $block_text .= $tag; } else { // Increase/decrease nested tag count. Only do so if // the tag's name match base tag's. if (preg_match('{^mode = $attr_m[2] . $attr_m[3]; $span_mode = $this->mode == 'span' || $this->mode != 'block' && preg_match('{^<(?:' . $this->contain_span_tags_re . ')\b}', $tag); // Calculate indent before tag. if (preg_match('/(?:^|\n)( *?)(?! ).*?$/', $block_text, $matches)) { $strlen = $this->utf8_strlen; $indent = $strlen($matches[1], 'UTF-8'); } else { $indent = 0; } // End preceding block with this tag. $block_text .= $tag; $parsed .= $this->$hash_method($block_text); // Get enclosing tag name for the ParseMarkdown function. // (This pattern makes $tag_name_re safe without quoting.) preg_match('/^<([\w:$]*)\b/', $tag, $matches); $tag_name_re = $matches[1]; // Parse the content using the HTML-in-Markdown parser. list ($block_text, $text) = $this->_hashHTMLBlocks_inMarkdown($text, $indent, $tag_name_re, $span_mode); // Outdent markdown text. if ($indent > 0) { $block_text = preg_replace("/^[ ]{1,$indent}/m", "", $block_text); } // Append tag content to parsed text. if (!$span_mode) $parsed .= "\n\n$block_text\n\n"; else $parsed .= "$block_text"; // Start over with a new block. $block_text = ""; } else $block_text .= $tag; } } while ($depth > 0); // Hash last block text that wasn't processed inside the loop. $parsed .= $this->$hash_method($block_text); return array($parsed, $text); } /** * Called whenever a tag must be hashed when a function inserts a "clean" tag * in $text, it passes through this function and is automaticaly escaped, * blocking invalid nested overlap. * @param string $text * @return string */ protected function hashClean($text) { return $this->hashPart($text, 'C'); } /** * Turn Markdown link shortcuts into XHTML tags. * @param string $text * @return string */ protected function doAnchors($text) { if ($this->in_anchor) { return $text; } $this->in_anchor = true; // First, handle reference-style links: [link text] [id] $text = preg_replace_callback('{ ( # wrap whole match in $1 \[ (' . $this->nested_brackets_re . ') # link text = $2 \] [ ]? # one optional space (?:\n[ ]*)? # one optional newline followed by spaces \[ (.*?) # id = $3 \] ) }xs', array($this, '_doAnchors_reference_callback'), $text); // Next, inline-style links: [link text](url "optional title") $text = preg_replace_callback('{ ( # wrap whole match in $1 \[ (' . $this->nested_brackets_re . ') # link text = $2 \] \( # literal paren [ \n]* (?: <(.+?)> # href = $3 | (' . $this->nested_url_parenthesis_re . ') # href = $4 ) [ \n]* ( # $5 ([\'"]) # quote char = $6 (.*?) # Title = $7 \6 # matching quote [ \n]* # ignore any spaces/tabs between closing quote and ) )? # title is optional \) (?:[ ]? ' . $this->id_class_attr_catch_re . ' )? # $8 = id/class attributes ) }xs', array($this, '_doAnchors_inline_callback'), $text); // Last, handle reference-style shortcuts: [link text] // These must come last in case you've also got [link text][1] // or [link text](/foo) $text = preg_replace_callback('{ ( # wrap whole match in $1 \[ ([^\[\]]+) # link text = $2; can\'t contain [ or ] \] ) }xs', array($this, '_doAnchors_reference_callback'), $text); $this->in_anchor = false; return $text; } /** * Callback for reference anchors * @param array $matches * @return string */ protected function _doAnchors_reference_callback($matches) { $whole_match = $matches[1]; $link_text = $matches[2]; $link_id =& $matches[3]; if ($link_id == "") { // for shortcut links like [this][] or [this]. $link_id = $link_text; } // lower-case and turn embedded newlines into spaces $link_id = strtolower($link_id); $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); if (isset($this->urls[$link_id])) { $url = $this->urls[$link_id]; $url = $this->encodeURLAttribute($url); $result = "titles[$link_id] ) ) { $title = $this->titles[$link_id]; $title = $this->encodeAttribute($title); $result .= " title=\"$title\""; } if (isset($this->ref_attr[$link_id])) $result .= $this->ref_attr[$link_id]; $link_text = $this->runSpanGamut($link_text); $result .= ">$link_text"; $result = $this->hashPart($result); } else { $result = $whole_match; } return $result; } /** * Callback for inline anchors * @param array $matches * @return string */ protected function _doAnchors_inline_callback($matches) { $whole_match = $matches[1]; $link_text = $this->runSpanGamut($matches[2]); $url = $matches[3] == '' ? $matches[4] : $matches[3]; $title =& $matches[7]; $attr = $this->doExtraAttributes("a", $dummy =& $matches[8]); // if the URL was of the form it got caught by the HTML
// tag parser and hashed. Need to reverse the process before using the URL.
$unhashed = $this->unhash($url);
if ($unhashed != $url)
$url = preg_replace('/^<(.*)>$/', '\1', $unhashed);
$url = $this->encodeURLAttribute($url);
$result = "encodeAttribute($title);
$result .= " title=\"$title\"";
}
$result .= $attr;
$link_text = $this->runSpanGamut($link_text);
$result .= ">$link_text";
return $this->hashPart($result);
}
/**
* Turn Markdown image shortcuts into ![]()
tags.
* @param string $text
* @return string
*/
protected function doImages($text) {
// First, handle reference-style labeled images: ![alt text][id]
$text = preg_replace_callback('{
( # wrap whole match in $1
!\[
(' . $this->nested_brackets_re . ') # alt text = $2
\]
[ ]? # one optional space
(?:\n[ ]*)? # one optional newline followed by spaces
\[
(.*?) # id = $3
\]
)
}xs',
array($this, '_doImages_reference_callback'), $text);
// Next, handle inline images: 
// Don't forget: encode * and _
$text = preg_replace_callback('{
( # wrap whole match in $1
!\[
(' . $this->nested_brackets_re . ') # alt text = $2
\]
\s? # One optional whitespace character
\( # literal paren
[ \n]*
(?:
<(\S*)> # src url = $3
|
(' . $this->nested_url_parenthesis_re . ') # src url = $4
)
[ \n]*
( # $5
([\'"]) # quote char = $6
(.*?) # title = $7
\6 # matching quote
[ \n]*
)? # title is optional
\)
(?:[ ]? ' . $this->id_class_attr_catch_re . ' )? # $8 = id/class attributes
)
}xs',
array($this, '_doImages_inline_callback'), $text);
return $text;
}
/**
* Callback for referenced images
* @param array $matches
* @return string
*/
protected function _doImages_reference_callback($matches) {
$whole_match = $matches[1];
$alt_text = $matches[2];
$link_id = strtolower($matches[3]);
if ($link_id == "") {
$link_id = strtolower($alt_text); // for shortcut links like ![this][].
}
$alt_text = $this->encodeAttribute($alt_text);
if (isset($this->urls[$link_id])) {
$url = $this->encodeURLAttribute($this->urls[$link_id]);
$result = "![\"$alt_text\"";](\"$url\")
titles[$link_id])) {
$title = $this->titles[$link_id];
$title = $this->encodeAttribute($title);
$result .= " title=\"$title\"";
}
if (isset($this->ref_attr[$link_id]))
$result .= $this->ref_attr[$link_id];
$result .= $this->empty_element_suffix;
$result = $this->hashPart($result);
}
else {
// If there's no such link ID, leave intact:
$result = $whole_match;
}
return $result;
}
/**
* Callback for inline images
* @param array $matches
* @return string
*/
protected function _doImages_inline_callback($matches) {
$whole_match = $matches[1];
$alt_text = $matches[2];
$url = $matches[3] == '' ? $matches[4] : $matches[3];
$title =& $matches[7];
$attr = $this->doExtraAttributes("img", $dummy =& $matches[8]);
$alt_text = $this->encodeAttribute($alt_text);
$url = $this->encodeURLAttribute($url);
$result = "encodeAttribute($title);
$result .= " title=\"$title\""; // $title already quoted
}
$result .= $attr;
$result .= $this->empty_element_suffix;
return $this->hashPart($result);
}
/**
* Process markdown headers. Redefined to add ID and class attribute support.
* @param string $text
* @return string
*/
protected function doHeaders($text) {
// Setext-style headers:
// Header 1 {#header1}
// ========
//
// Header 2 {#header2 .class1 .class2}
// --------
//
$text = preg_replace_callback(
'{
(^.+?) # $1: Header text
(?:[ ]+ ' . $this->id_class_attr_catch_re . ' )? # $3 = id/class attributes
[ ]*\n(=+|-+)[ ]*\n+ # $3: Header footer
}mx',
array($this, '_doHeaders_callback_setext'), $text);
// atx-style headers:
// # Header 1 {#header1}
// ## Header 2 {#header2}
// ## Header 2 with closing hashes ## {#header3.class1.class2}
// ...
// ###### Header 6 {.class2}
//
$text = preg_replace_callback('{
^(\#{1,6}) # $1 = string of #\'s
[ ]*
(.+?) # $2 = Header text
[ ]*
\#* # optional closing #\'s (not counted)
(?:[ ]+ ' . $this->id_class_attr_catch_re . ' )? # $3 = id/class attributes
[ ]*
\n+
}xm',
array($this, '_doHeaders_callback_atx'), $text);
return $text;
}
/**
* Callback for setext headers
* @param array $matches
* @return string
*/
protected function _doHeaders_callback_setext($matches) {
if ($matches[3] == '-' && preg_match('{^- }', $matches[1])) {
return $matches[0];
}
$level = $matches[3]{0} == '=' ? 1 : 2;
$defaultId = is_callable($this->header_id_func) ? call_user_func($this->header_id_func, $matches[1]) : null;
$attr = $this->doExtraAttributes("h$level", $dummy =& $matches[2], $defaultId);
$block = "" . $this->runSpanGamut($matches[1]) . "";
return "\n" . $this->hashBlock($block) . "\n\n";
}
/**
* Callback for atx headers
* @param array $matches
* @return string
*/
protected function _doHeaders_callback_atx($matches) {
$level = strlen($matches[1]);
$defaultId = is_callable($this->header_id_func) ? call_user_func($this->header_id_func, $matches[2]) : null;
$attr = $this->doExtraAttributes("h$level", $dummy =& $matches[3], $defaultId);
$block = "" . $this->runSpanGamut($matches[2]) . "";
return "\n" . $this->hashBlock($block) . "\n\n";
}
/**
* Form HTML tables.
* @param string $text
* @return string
*/
protected function doTables($text) {
$less_than_tab = $this->tab_width - 1;
// Find tables with leading pipe.
//
// | Header 1 | Header 2
// | -------- | --------
// | Cell 1 | Cell 2
// | Cell 3 | Cell 4
$text = preg_replace_callback('
{
^ # Start of a line
[ ]{0,' . $less_than_tab . '} # Allowed whitespace.
[|] # Optional leading pipe (present)
(.+) \n # $1: Header row (at least one pipe)
[ ]{0,' . $less_than_tab . '} # Allowed whitespace.
[|] ([ ]*[-:]+[-| :]*) \n # $2: Header underline
( # $3: Cells
(?>
[ ]* # Allowed whitespace.
[|] .* \n # Row content.
)*
)
(?=\n|\Z) # Stop at final double newline.
}xm',
array($this, '_doTable_leadingPipe_callback'), $text);
// Find tables without leading pipe.
//
// Header 1 | Header 2
// -------- | --------
// Cell 1 | Cell 2
// Cell 3 | Cell 4
$text = preg_replace_callback('
{
^ # Start of a line
[ ]{0,' . $less_than_tab . '} # Allowed whitespace.
(\S.*[|].*) \n # $1: Header row (at least one pipe)
[ ]{0,' . $less_than_tab . '} # Allowed whitespace.
([-:]+[ ]*[|][-| :]*) \n # $2: Header underline
( # $3: Cells
(?>
.* [|] .* \n # Row content
)*
)
(?=\n|\Z) # Stop at final double newline.
}xm',
array($this, '_DoTable_callback'), $text);
return $text;
}
/**
* Callback for removing the leading pipe for each row
* @param array $matches
* @return string
*/
protected function _doTable_leadingPipe_callback($matches) {
$head = $matches[1];
$underline = $matches[2];
$content = $matches[3];
$content = preg_replace('/^ *[|]/m', '', $content);
return $this->_doTable_callback(array($matches[0], $head, $underline, $content));
}
/**
* Make the align attribute in a table
* @param string $alignname
* @return string
*/
protected function _doTable_makeAlignAttr($alignname)
{
if (empty($this->table_align_class_tmpl)) {
return " align=\"$alignname\"";
}
$classname = str_replace('%%', $alignname, $this->table_align_class_tmpl);
return " class=\"$classname\"";
}
/**
* Calback for processing tables
* @param array $matches
* @return string
*/
protected function _doTable_callback($matches) {
$head = $matches[1];
$underline = $matches[2];
$content = $matches[3];
// Remove any tailing pipes for each line.
$head = preg_replace('/[|] *$/m', '', $head);
$underline = preg_replace('/[|] *$/m', '', $underline);
$content = preg_replace('/[|] *$/m', '', $content);
// Reading alignement from header underline.
$separators = preg_split('/ *[|] */', $underline);
foreach ($separators as $n => $s) {
if (preg_match('/^ *-+: *$/', $s))
$attr[$n] = $this->_doTable_makeAlignAttr('right');
else if (preg_match('/^ *:-+: *$/', $s))
$attr[$n] = $this->_doTable_makeAlignAttr('center');
else if (preg_match('/^ *:-+ *$/', $s))
$attr[$n] = $this->_doTable_makeAlignAttr('left');
else
$attr[$n] = '';
}
// Parsing span elements, including code spans, character escapes,
// and inline HTML tags, so that pipes inside those gets ignored.
$head = $this->parseSpan($head);
$headers = preg_split('/ *[|] */', $head);
$col_count = count($headers);
$attr = array_pad($attr, $col_count, '');
// Write column headers.
$text = "\n";
$text .= "\n";
$text .= "
";
return $this->hashBlock($text) . "\n";
}
/**
* Form HTML definition lists.
* @param string $text
* @return string
*/
protected function doDefLists($text) {
$less_than_tab = $this->tab_width - 1;
// Re-usable pattern to match any entire dl list:
$whole_list_re = '(?>
( # $1 = whole list
( # $2
[ ]{0,' . $less_than_tab . '}
((?>.*\S.*\n)+) # $3 = defined term
\n?
[ ]{0,' . $less_than_tab . '}:[ ]+ # colon starting definition
)
(?s:.+?)
( # $4
\z
|
\n{2,}
(?=\S)
(?! # Negative lookahead for another term
[ ]{0,' . $less_than_tab . '}
(?: \S.*\n )+? # defined term
\n?
[ ]{0,' . $less_than_tab . '}:[ ]+ # colon starting definition
)
(?! # Negative lookahead for another definition
[ ]{0,' . $less_than_tab . '}:[ ]+ # colon starting definition
)
)
)
)'; // mx
$text = preg_replace_callback('{
(?>\A\n?|(?<=\n\n))
' . $whole_list_re . '
}mx',
array($this, '_doDefLists_callback'), $text);
return $text;
}
/**
* Callback for processing definition lists
* @param array $matches
* @return string
*/
protected function _doDefLists_callback($matches) {
// Re-usable patterns to match list item bullets and number markers:
$list = $matches[1];
// Turn double returns into triple returns, so that we can make a
// paragraph for the last item in a list, if necessary:
$result = trim($this->processDefListItems($list));
$result = " | \z
)
)
}xm',
array($this, '_processDefListItems_callback_dd'), $list_str);
return $list_str;
}
/**
* Callback for elements in definition lists
* @param array $matches
* @return string
*/
protected function _processDefListItems_callback_dt($matches) {
$terms = explode("\n", trim($matches[1]));
$text = '';
foreach ($terms as $term) {
$term = $this->runSpanGamut(trim($term));
$text .= "\n" . $term . " ";
}
return $text . "\n";
}
/**
* Callback for elements in definition lists
* @param array $matches
* @return string
*/
protected function _processDefListItems_callback_dd($matches) {
$leading_line = $matches[1];
$marker_space = $matches[2];
$def = $matches[3];
if ($leading_line || preg_match('/\n{2,}/', $def)) {
// Replace marker with the appropriate whitespace indentation
$def = str_repeat(' ', strlen($marker_space)) . $def;
$def = $this->runBlockGamut($this->outdent($def . "\n\n"));
$def = "\n". $def ."\n";
}
else {
$def = rtrim($def);
$def = $this->runSpanGamut($this->outdent($def));
}
return "\n" . $def . " \n";
}
/**
* Adding the fenced code block syntax to regular Markdown:
*
* ~~~
* Code block
* ~~~
*
* @param string $text
* @return string
*/
protected function doFencedCodeBlocks($text) {
$less_than_tab = $this->tab_width;
$text = preg_replace_callback('{
(?:\n|\A)
# 1: Opening marker
(
(?:~{3,}|`{3,}) # 3 or more tildes/backticks.
)
[ ]*
(?:
\.?([-_:a-zA-Z0-9]+) # 2: standalone class name
)?
[ ]*
(?:
' . $this->id_class_attr_catch_re . ' # 3: Extra attributes
)?
[ ]* \n # Whitespace and newline following marker.
# 4: Content
(
(?>
(?!\1 [ ]* \n) # Not a closing marker.
.*\n+
)+
)
# Closing marker.
\1 [ ]* (?= \n )
}xm',
array($this, '_doFencedCodeBlocks_callback'), $text);
return $text;
}
/**
* Callback to process fenced code blocks
* @param array $matches
* @return string
*/
protected function _doFencedCodeBlocks_callback($matches) {
$classname =& $matches[2];
$attrs =& $matches[3];
$codeblock = $matches[4];
if ($this->code_block_content_func) {
$codeblock = call_user_func($this->code_block_content_func, $codeblock, $classname);
} else {
$codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES);
}
$codeblock = preg_replace_callback('/^\n+/',
array($this, '_doFencedCodeBlocks_newlines'), $codeblock);
$classes = array();
if ($classname != "") {
if ($classname{0} == '.')
$classname = substr($classname, 1);
$classes[] = $this->code_class_prefix . $classname;
}
$attr_str = $this->doExtraAttributes($this->code_attr_on_pre ? "pre" : "code", $attrs, null, $classes);
$pre_attr_str = $this->code_attr_on_pre ? $attr_str : '';
$code_attr_str = $this->code_attr_on_pre ? '' : $attr_str;
$codeblock = "$codeblock";
return "\n\n".$this->hashBlock($codeblock)."\n\n";
}
/**
* Replace new lines in fenced code blocks
* @param array $matches
* @return string
*/
protected function _doFencedCodeBlocks_newlines($matches) {
return str_repeat("empty_element_suffix",
strlen($matches[0]));
}
/**
* Redefining emphasis markers so that emphasis by underscore does not
* work in the middle of a word.
* @var array
*/
protected $em_relist = array(
'' => '(?:(? '(? '(? '(?:(? '(? '(? '(?:(? '(? '(? tags
* @return string HTML output
*/
protected function formParagraphs($text, $wrap_in_p = true) {
// Strip leading and trailing lines:
$text = preg_replace('/\A\n+|\n+\z/', '', $text);
$grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY);
// Wrap
) // Comments and Processing Instructions. if (preg_match('{^auto_close_tags_re . ')\b}', $tag) || $tag{1} == '!' || $tag{1} == '?') { // Just add the tag to the block as if it was text. $block_text .= $tag; } else { // Increase/decrease nested tag count. Only do so if // the tag's name match base tag's. if (preg_match('{^mode = $attr_m[2] . $attr_m[3]; $span_mode = $this->mode == 'span' || $this->mode != 'block' && preg_match('{^<(?:' . $this->contain_span_tags_re . ')\b}', $tag); // Calculate indent before tag. if (preg_match('/(?:^|\n)( *?)(?! ).*?$/', $block_text, $matches)) { $strlen = $this->utf8_strlen; $indent = $strlen($matches[1], 'UTF-8'); } else { $indent = 0; } // End preceding block with this tag. $block_text .= $tag; $parsed .= $this->$hash_method($block_text); // Get enclosing tag name for the ParseMarkdown function. // (This pattern makes $tag_name_re safe without quoting.) preg_match('/^<([\w:$]*)\b/', $tag, $matches); $tag_name_re = $matches[1]; // Parse the content using the HTML-in-Markdown parser. list ($block_text, $text) = $this->_hashHTMLBlocks_inMarkdown($text, $indent, $tag_name_re, $span_mode); // Outdent markdown text. if ($indent > 0) { $block_text = preg_replace("/^[ ]{1,$indent}/m", "", $block_text); } // Append tag content to parsed text. if (!$span_mode) $parsed .= "\n\n$block_text\n\n"; else $parsed .= "$block_text"; // Start over with a new block. $block_text = ""; } else $block_text .= $tag; } } while ($depth > 0); // Hash last block text that wasn't processed inside the loop. $parsed .= $this->$hash_method($block_text); return array($parsed, $text); } /** * Called whenever a tag must be hashed when a function inserts a "clean" tag * in $text, it passes through this function and is automaticaly escaped, * blocking invalid nested overlap. * @param string $text * @return string */ protected function hashClean($text) { return $this->hashPart($text, 'C'); } /** * Turn Markdown link shortcuts into XHTML tags. * @param string $text * @return string */ protected function doAnchors($text) { if ($this->in_anchor) { return $text; } $this->in_anchor = true; // First, handle reference-style links: [link text] [id] $text = preg_replace_callback('{ ( # wrap whole match in $1 \[ (' . $this->nested_brackets_re . ') # link text = $2 \] [ ]? # one optional space (?:\n[ ]*)? # one optional newline followed by spaces \[ (.*?) # id = $3 \] ) }xs', array($this, '_doAnchors_reference_callback'), $text); // Next, inline-style links: [link text](url "optional title") $text = preg_replace_callback('{ ( # wrap whole match in $1 \[ (' . $this->nested_brackets_re . ') # link text = $2 \] \( # literal paren [ \n]* (?: <(.+?)> # href = $3 | (' . $this->nested_url_parenthesis_re . ') # href = $4 ) [ \n]* ( # $5 ([\'"]) # quote char = $6 (.*?) # Title = $7 \6 # matching quote [ \n]* # ignore any spaces/tabs between closing quote and ) )? # title is optional \) (?:[ ]? ' . $this->id_class_attr_catch_re . ' )? # $8 = id/class attributes ) }xs', array($this, '_doAnchors_inline_callback'), $text); // Last, handle reference-style shortcuts: [link text] // These must come last in case you've also got [link text][1] // or [link text](/foo) $text = preg_replace_callback('{ ( # wrap whole match in $1 \[ ([^\[\]]+) # link text = $2; can\'t contain [ or ] \] ) }xs', array($this, '_doAnchors_reference_callback'), $text); $this->in_anchor = false; return $text; } /** * Callback for reference anchors * @param array $matches * @return string */ protected function _doAnchors_reference_callback($matches) { $whole_match = $matches[1]; $link_text = $matches[2]; $link_id =& $matches[3]; if ($link_id == "") { // for shortcut links like [this][] or [this]. $link_id = $link_text; } // lower-case and turn embedded newlines into spaces $link_id = strtolower($link_id); $link_id = preg_replace('{[ ]?\n}', ' ', $link_id); if (isset($this->urls[$link_id])) { $url = $this->urls[$link_id]; $url = $this->encodeURLAttribute($url); $result = "titles[$link_id] ) ) { $title = $this->titles[$link_id]; $title = $this->encodeAttribute($title); $result .= " title=\"$title\""; } if (isset($this->ref_attr[$link_id])) $result .= $this->ref_attr[$link_id]; $link_text = $this->runSpanGamut($link_text); $result .= ">$link_text"; $result = $this->hashPart($result); } else { $result = $whole_match; } return $result; } /** * Callback for inline anchors * @param array $matches * @return string */ protected function _doAnchors_inline_callback($matches) { $whole_match = $matches[1]; $link_text = $this->runSpanGamut($matches[2]); $url = $matches[3] == '' ? $matches[4] : $matches[3]; $title =& $matches[7]; $attr = $this->doExtraAttributes("a", $dummy =& $matches[8]); // if the URL was of the form
- \n" . $result . "\n
tags and unhashify HTML blocks foreach ($grafs as $key => $value) { $value = trim($this->runSpanGamut($value)); // Check if this should be enclosed in a paragraph. // Clean tag hashes & block tag hashes are left alone. $is_p = $wrap_in_p && !preg_match('/^B\x1A[0-9]+B|^C\x1A[0-9]+C$/', $value); if ($is_p) { $value = "
$value
"; } $grafs[$key] = $value; } // Join grafs in one text, then unhash HTML tags. $text = implode("\n\n", $grafs); // Finish by removing any tag hashes still present in $text. $text = $this->unhash($text); return $text; } /** * Footnotes - Strips link definitions from text, stores the URLs and * titles in hash references. * @param string $text * @return string */ protected function stripFootnotes($text) { $less_than_tab = $this->tab_width - 1; // Link defs are in the form: [^id]: url "optional title" $text = preg_replace_callback('{ ^[ ]{0,' . $less_than_tab . '}\[\^(.+?)\][ ]?: # note_id = $1 [ ]* \n? # maybe *one* newline ( # text = $2 (no blank lines allowed) (?: .+ # actual text | \n # newlines but (?!\[.+?\][ ]?:\s)# negative lookahead for footnote or link definition marker. (?!\n+[ ]{0,3}\S)# ensure line is not blank and followed # by non-indented content )* ) }xm', array($this, '_stripFootnotes_callback'), $text); return $text; } /** * Callback for stripping footnotes * @param array $matches * @return string */ protected function _stripFootnotes_callback($matches) { $note_id = $this->fn_id_prefix . $matches[1]; $this->footnotes[$note_id] = $this->outdent($matches[2]); return ''; // String that will replace the block } /** * Replace footnote references in $text [^id] with a special text-token * which will be replaced by the actual footnote marker in appendFootnotes. * @param string $text * @return string */ protected function doFootnotes($text) { if (!$this->in_anchor) { $text = preg_replace('{\[\^(.+?)\]}', "F\x1Afn:\\1\x1A:", $text); } return $text; } /** * Append footnote list to text * @param string $text * @return string */ protected function appendFootnotes($text) { $text = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}', array($this, '_appendFootnotes_callback'), $text); if (!empty($this->footnotes_ordered)) { $text .= "\n\n"; $text .= "\n";
$text .= "empty_element_suffix . "\n";
$text .= "
";
}
return $text;
}
/**
* Callback for appending footnotes
* @param array $matches
* @return string
*/
protected function _appendFootnotes_callback($matches) {
$node_id = $this->fn_id_prefix . $matches[1];
// Create footnote marker only if it has a corresponding footnote *and*
// the footnote hasn't been used by another marker.
if (isset($this->footnotes[$node_id])) {
$num =& $this->footnotes_numbers[$node_id];
if (!isset($num)) {
// Transfer footnote content to the ordered list and give it its
// number
$this->footnotes_ordered[$node_id] = $this->footnotes[$node_id];
$this->footnotes_ref_count[$node_id] = 1;
$num = $this->footnote_counter++;
$ref_count_mark = '';
} else {
$ref_count_mark = $this->footnotes_ref_count[$node_id] += 1;
}
$attr = "";
if ($this->fn_link_class != "") {
$class = $this->fn_link_class;
$class = $this->encodeAttribute($class);
$attr .= " class=\"$class\"";
}
if ($this->fn_link_title != "") {
$title = $this->fn_link_title;
$title = $this->encodeAttribute($title);
$attr .= " title=\"$title\"";
}
$attr = str_replace("%%", $num, $attr);
$node_id = $this->encodeAttribute($node_id);
return
"".
"$num".
"";
}
return "[^" . $matches[1] . "]";
}
/**
* Abbreviations - strips abbreviations from text, stores titles in hash
* references.
* @param string $text
* @return string
*/
protected function stripAbbreviations($text) {
$less_than_tab = $this->tab_width - 1;
// Link defs are in the form: [id]*: url "optional title"
$text = preg_replace_callback('{
^[ ]{0,' . $less_than_tab . '}\*\[(.+?)\][ ]?: # abbr_id = $1
(.*) # text = $2 (no blank lines allowed)
}xm',
array($this, '_stripAbbreviations_callback'),
$text);
return $text;
}
/**
* Callback for stripping abbreviations
* @param array $matches
* @return string
*/
protected function _stripAbbreviations_callback($matches) {
$abbr_word = $matches[1];
$abbr_desc = $matches[2];
if ($this->abbr_word_re) {
$this->abbr_word_re .= '|';
}
$this->abbr_word_re .= preg_quote($abbr_word);
$this->abbr_desciptions[$abbr_word] = trim($abbr_desc);
return ''; // String that will replace the block
}
/**
* Find defined abbreviations in text and wrap them in elements.
* @param string $text
* @return string
*/
protected function doAbbreviations($text) {
if ($this->abbr_word_re) {
// cannot use the /x modifier because abbr_word_re may
// contain significant spaces:
$text = preg_replace_callback('{' .
'(?abbr_word_re . ')' .
'(?![\w\x1A])' .
'}',
array($this, '_doAbbreviations_callback'), $text);
}
return $text;
}
/**
* Callback for processing abbreviations
* @param array $matches
* @return string
*/
protected function _doAbbreviations_callback($matches) {
$abbr = $matches[0];
if (isset($this->abbr_desciptions[$abbr])) {
$desc = $this->abbr_desciptions[$abbr];
if (empty($desc)) {
return $this->hashPart("$abbr");
} else {
$desc = $this->encodeAttribute($desc);
return $this->hashPart("$abbr");
}
} else {
return $matches[0];
}
}
}
- \n\n";
$attr = "";
if ($this->fn_backlink_class != "") {
$class = $this->fn_backlink_class;
$class = $this->encodeAttribute($class);
$attr .= " class=\"$class\"";
}
if ($this->fn_backlink_title != "") {
$title = $this->fn_backlink_title;
$title = $this->encodeAttribute($title);
$attr .= " title=\"$title\"";
}
$backlink_text = $this->fn_backlink_html;
$num = 0;
while (!empty($this->footnotes_ordered)) {
$footnote = reset($this->footnotes_ordered);
$note_id = key($this->footnotes_ordered);
unset($this->footnotes_ordered[$note_id]);
$ref_count = $this->footnotes_ref_count[$note_id];
unset($this->footnotes_ref_count[$note_id]);
unset($this->footnotes[$note_id]);
$footnote .= "\n"; // Need to append newline before parsing.
$footnote = $this->runBlockGamut("$footnote\n");
$footnote = preg_replace_callback('{F\x1Afn:(.*?)\x1A:}',
array($this, '_appendFootnotes_callback'), $footnote);
$attr = str_replace("%%", ++$num, $attr);
$note_id = $this->encodeAttribute($note_id);
// Prepare backlink, multiple backlinks if multiple references
$backlink = "$backlink_text";
for ($ref_num = 2; $ref_num <= $ref_count; ++$ref_num) {
$backlink .= " $backlink_text";
}
// Add backlink to last paragraph; create new paragraph if needed.
if (preg_match('{$}', $footnote)) {
$footnote = substr($footnote, 0, -4) . " $backlink";
} else {
$footnote .= "\n\n
- \n"; $text .= $footnote . "\n"; $text .= " \n\n"; } $text .= "
$backlink
"; } $text .= "