` blocks.
#
$text = preg_replace_callback('{
(?:\n\n|\A\n?)
( # $1 = the code block -- one or more lines, starting with a space/tab
(?>
[ ]{'.$this->tab_width.'} # Lines must start with a tab or a tab-width of spaces
.*\n+
)+
)
((?=^[ ]{0,'.$this->tab_width.'}\S)|\Z) # Lookahead for non-space at line-start, or end of doc
}xm',
array($this, '_doCodeBlocks_callback'), $text);
return $text;
}
protected function _doCodeBlocks_callback($matches) {
$codeblock = $matches[1];
$codeblock = $this->outdent($codeblock);
$codeblock = htmlspecialchars($codeblock, ENT_NOQUOTES);
# trim leading newlines and trailing newlines
$codeblock = preg_replace('/\A\n+|\n+\z/', '', $codeblock);
$codeblock = "$codeblock\n
";
return "\n\n".$this->hashBlock($codeblock)."\n\n";
}
protected function makeCodeSpan($code) {
#
# Create a code span markup for $code. Called from handleSpanToken.
#
$code = htmlspecialchars(trim($code), ENT_NOQUOTES);
return $this->hashPart("$code");
}
protected $em_relist = array(
'' => '(?:(? '(? '(? '(?:(? '(? '(? '(?:(? '(? '(?em_relist as $em => $em_re) {
foreach ($this->strong_relist as $strong => $strong_re) {
# Construct list of allowed token expressions.
$token_relist = array();
if (isset($this->em_strong_relist["$em$strong"])) {
$token_relist[] = $this->em_strong_relist["$em$strong"];
}
$token_relist[] = $em_re;
$token_relist[] = $strong_re;
# Construct master expression from list.
$token_re = '{('. implode('|', $token_relist) .')}';
$this->em_strong_prepared_relist["$em$strong"] = $token_re;
}
}
}
protected function doItalicsAndBold($text) {
$token_stack = array('');
$text_stack = array('');
$em = '';
$strong = '';
$tree_char_em = false;
while (1) {
#
# Get prepared regular expression for seraching emphasis tokens
# in current context.
#
$token_re = $this->em_strong_prepared_relist["$em$strong"];
#
# Each loop iteration search for the next emphasis token.
# Each token is then passed to handleSpanToken.
#
$parts = preg_split($token_re, $text, 2, PREG_SPLIT_DELIM_CAPTURE);
$text_stack[0] .= $parts[0];
$token =& $parts[1];
$text =& $parts[2];
if (empty($token)) {
# Reached end of text span: empty stack without emitting.
# any more emphasis.
while ($token_stack[0]) {
$text_stack[1] .= array_shift($token_stack);
$text_stack[0] .= array_shift($text_stack);
}
break;
}
$token_len = strlen($token);
if ($tree_char_em) {
# Reached closing marker while inside a three-char emphasis.
if ($token_len == 3) {
# Three-char closing marker, close em and strong.
array_shift($token_stack);
$span = array_shift($text_stack);
$span = $this->runSpanGamut($span);
$span = "$span";
$text_stack[0] .= $this->hashPart($span);
$em = '';
$strong = '';
} else {
# Other closing marker: close one em or strong and
# change current token state to match the other
$token_stack[0] = str_repeat($token{0}, 3-$token_len);
$tag = $token_len == 2 ? "strong" : "em";
$span = $text_stack[0];
$span = $this->runSpanGamut($span);
$span = "<$tag>$span";
$text_stack[0] = $this->hashPart($span);
$$tag = ''; # $$tag stands for $em or $strong
}
$tree_char_em = false;
} else if ($token_len == 3) {
if ($em) {
# Reached closing marker for both em and strong.
# Closing strong marker:
for ($i = 0; $i < 2; ++$i) {
$shifted_token = array_shift($token_stack);
$tag = strlen($shifted_token) == 2 ? "strong" : "em";
$span = array_shift($text_stack);
$span = $this->runSpanGamut($span);
$span = "<$tag>$span";
$text_stack[0] .= $this->hashPart($span);
$$tag = ''; # $$tag stands for $em or $strong
}
} else {
# Reached opening three-char emphasis marker. Push on token
# stack; will be handled by the special condition above.
$em = $token{0};
$strong = "$em$em";
array_unshift($token_stack, $token);
array_unshift($text_stack, '');
$tree_char_em = true;
}
} else if ($token_len == 2) {
if ($strong) {
# Unwind any dangling emphasis marker:
if (strlen($token_stack[0]) == 1) {
$text_stack[1] .= array_shift($token_stack);
$text_stack[0] .= array_shift($text_stack);
}
# Closing strong marker:
array_shift($token_stack);
$span = array_shift($text_stack);
$span = $this->runSpanGamut($span);
$span = "$span";
$text_stack[0] .= $this->hashPart($span);
$strong = '';
} else {
array_unshift($token_stack, $token);
array_unshift($text_stack, '');
$strong = $token;
}
} else {
# Here $token_len == 1
if ($em) {
if (strlen($token_stack[0]) == 1) {
# Closing emphasis marker:
array_shift($token_stack);
$span = array_shift($text_stack);
$span = $this->runSpanGamut($span);
$span = "$span";
$text_stack[0] .= $this->hashPart($span);
$em = '';
} else {
$text_stack[0] .= $token;
}
} else {
array_unshift($token_stack, $token);
array_unshift($text_stack, '');
$em = $token;
}
}
}
return $text_stack[0];
}
protected function doBlockQuotes($text) {
$text = preg_replace_callback('/
( # Wrap whole match in $1
(?>
^[ ]*>[ ]? # ">" at the start of a line
.+\n # rest of the first line
(.+\n)* # subsequent consecutive lines
\n* # blanks
)+
)
/xm',
array($this, '_doBlockQuotes_callback'), $text);
return $text;
}
protected function _doBlockQuotes_callback($matches) {
$bq = $matches[1];
# trim one level of quoting - trim whitespace-only lines
$bq = preg_replace('/^[ ]*>[ ]?|^[ ]+$/m', '', $bq);
$bq = $this->runBlockGamut($bq); # recurse
$bq = preg_replace('/^/m', " ", $bq);
# These leading spaces cause problem with content,
# so we need to fix that:
$bq = preg_replace_callback('{(\s*.+?
)}sx',
array($this, '_doBlockQuotes_callback2'), $bq);
return "\n". $this->hashBlock("\n$bq\n
")."\n\n";
}
protected function _doBlockQuotes_callback2($matches) {
$pre = $matches[1];
$pre = preg_replace('/^ /m', '', $pre);
return $pre;
}
protected function formParagraphs($text) {
#
# Params:
# $text - string to process with html tags
#
# Strip leading and trailing lines:
$text = preg_replace('/\A\n+|\n+\z/', '', $text);
$grafs = preg_split('/\n{2,}/', $text, -1, PREG_SPLIT_NO_EMPTY);
#
# Wrap
tags and unhashify HTML blocks
#
foreach ($grafs as $key => $value) {
if (!preg_match('/^B\x1A[0-9]+B$/', $value)) {
# Is a paragraph.
$value = $this->runSpanGamut($value);
$value = preg_replace('/^([ ]*)/', "
", $value);
$value .= "
";
$grafs[$key] = $this->unhash($value);
}
else {
# Is a block.
# Modify elements of @grafs in-place...
$graf = $value;
$block = $this->html_hashes[$graf];
$graf = $block;
// if (preg_match('{
// \A
// ( # $1 = tag
// ]*
// \b
// markdown\s*=\s* ([\'"]) # $2 = attr quote char
// 1
// \2
// [^>]*
// >
// )
// ( # $3 = contents
// .*
// )
// () # $4 = closing tag
// \z
// }xs', $block, $matches))
// {
// list(, $div_open, , $div_content, $div_close) = $matches;
//
// # We can't call Markdown(), because that resets the hash;
// # that initialization code should be pulled into its own sub, though.
// $div_content = $this->hashHTMLBlocks($div_content);
//
// # Run document gamut methods on the content.
// foreach ($this->document_gamut as $method => $priority) {
// $div_content = $this->$method($div_content);
// }
//
// $div_open = preg_replace(
// '{\smarkdown\s*=\s*([\'"]).+?\1}', '', $div_open);
//
// $graf = $div_open . "\n" . $div_content . "\n" . $div_close;
// }
$grafs[$key] = $graf;
}
}
return implode("\n\n", $grafs);
}
protected function encodeAttribute($text) {
#
# Encode text for a double-quoted HTML attribute. This function
# is *not* suitable for attributes enclosed in single quotes.
#
$text = $this->encodeAmpsAndAngles($text);
$text = str_replace('"', '"', $text);
return $text;
}
protected function encodeURLAttribute($url, &$text = null) {
#
# Encode text for a double-quoted HTML attribute containing a URL,
# applying the URL filter if set. Also generates the textual
# representation for the URL (removing mailto: or tel:) storing it in $text.
# This function is *not* suitable for attributes enclosed in single quotes.
#
if ($this->url_filter_func)
$url = call_user_func($this->url_filter_func, $url);
if (preg_match('{^mailto:}i', $url))
$url = $this->encodeEntityObfuscatedAttribute($url, $text, 7);
else if (preg_match('{^tel:}i', $url))
{
$url = $this->encodeAttribute($url);
$text = substr($url, 4);
}
else
{
$url = $this->encodeAttribute($url);
$text = $url;
}
return $url;
}
protected function encodeAmpsAndAngles($text) {
#
# Smart processing for ampersands and angle brackets that need to
# be encoded. Valid character entities are left alone unless the
# no-entities mode is set.
#
if ($this->no_entities) {
$text = str_replace('&', '&', $text);
} else {
# Ampersand-encoding based entirely on Nat Irons's Amputator
# MT plugin:
$text = preg_replace_callback('{
<
(?:mailto:)?
(
(?:
[-!#$%&\'*+/=?^_`.{|}~\w\x80-\xFF]+
|
".*?"
)
\@
(?:
[-a-z0-9\x80-\xFF]+(\.[-a-z0-9\x80-\xFF]+)*\.[a-z]+
|
\[[\d.a-fA-F:]+\] # IPv4 & IPv6
)
)
>
}xi',
array($this, '_doAutoLinks_email_callback'), $text);
return $text;
}
protected function _doAutoLinks_url_callback($matches) {
$url = $this->encodeURLAttribute($matches[1], $text);
$link = "$text";
return $this->hashPart($link);
}
protected function _doAutoLinks_email_callback($matches) {
$addr = $matches[1];
$url = $this->encodeURLAttribute("mailto:$addr", $text);
$link = "$text";
return $this->hashPart($link);
}
protected function encodeEntityObfuscatedAttribute($text, &$tail = null, $head_length = 0) {
#
# Input: some text to obfuscate, e.g. "mailto:foo@example.com"
#
# Output: the same text but with most characters encoded as either a
# decimal or hex entity, in the hopes of foiling most address
# harvesting spam bots. E.g.:
#
# mailto:foo
# @example.co
# m
#
# Note: the additional output $tail is assigned the same value as the
# ouput, minus the number of characters specified by $head_length.
#
# Based by a filter by Matthew Wickline, posted to BBEdit-Talk.
# With some optimizations by Milian Wolff. Forced encoding of HTML
# attribute special characters by Allan Odgaard.
#
if ($text == "") return $tail = "";
$chars = preg_split('/(? $char) {
$ord = ord($char);
# Ignore non-ascii chars.
if ($ord < 128) {
$r = ($seed * (1 + $key)) % 100; # Pseudo-random function.
# roughly 10% raw, 45% hex, 45% dec
# '@' *must* be encoded. I insist.
# '"' and '>' have to be encoded inside the attribute
if ($r > 90 && strpos('@"&>', $char) === false) /* do nothing */;
else if ($r < 45) $chars[$key] = '&#x'.dechex($ord).';';
else $chars[$key] = '&#'.$ord.';';
}
}
$text = implode('', $chars);
$tail = $head_length ? implode('', array_slice($chars, $head_length)) : $text;
return $text;
}
protected function parseSpan($str) {
#
# Take the string $str and parse it into tokens, hashing embeded HTML,
# escaped characters and handling code spans.
#
$output = '';
$span_re = '{
(
\\\\'.$this->escape_chars_re.'
|
(?no_markup ? '' : '
|
# comment
|
<\?.*?\?> | <%.*?%> # processing instruction
|
<[!$]?[-a-zA-Z0-9:_]+ # regular tags
(?>
\s
(?>[^"\'>]+|"[^"]*"|\'[^\']*\')*
)?
>
|
<[-a-zA-Z0-9:_]+\s*/> # xml-style empty tag
|
# closing tag
').'
)
}xs';
while (1) {
#
# Each loop iteration seach for either the next tag, the next
# openning code span marker, or the next escaped character.
# Each token is then passed to handleSpanToken.
#
$parts = preg_split($span_re, $str, 2, PREG_SPLIT_DELIM_CAPTURE);
# Create token from text preceding tag.
if ($parts[0] != "") {
$output .= $parts[0];
}
# Check if we reach the end.
if (isset($parts[1])) {
$output .= $this->handleSpanToken($parts[1], $parts[2]);
$str = $parts[2];
}
else {
break;
}
}
return $output;
}
protected function handleSpanToken($token, &$str) {
#
# Handle $token provided by parseSpan by determining its nature and
# returning the corresponding value that should replace it.
#
switch ($token{0}) {
case "\\":
return $this->hashPart("&#". ord($token{1}). ";");
case "`":
# Search for end marker in remaining text.
if (preg_match('/^(.*?[^`])'.preg_quote($token).'(?!`)(.*)$/sm',
$str, $matches))
{
$str = $matches[2];
$codespan = $this->makeCodeSpan($matches[1]);
return $this->hashPart($codespan);
}
return $token; // return as text since no ending marker found.
default:
return $this->hashPart($token);
}
}
protected function outdent($text) {
#
# Remove one level of line-leading tabs or spaces
#
return preg_replace('/^(\t|[ ]{1,'.$this->tab_width.'})/m', '', $text);
}
# String length function for detab. `_initDetab` will create a function to
# hanlde UTF-8 if the default function does not exist.
protected $utf8_strlen = 'mb_strlen';
protected function detab($text) {
#
# Replace tabs with the appropriate amount of space.
#
# For each line we separate the line in blocks delemited by
# tab characters. Then we reconstruct every line by adding the
# appropriate number of space between each blocks.
$text = preg_replace_callback('/^.*\t.*$/m',
array($this, '_detab_callback'), $text);
return $text;
}
protected function _detab_callback($matches) {
$line = $matches[0];
$strlen = $this->utf8_strlen; # strlen function for UTF-8.
# Split in blocks.
$blocks = explode("\t", $line);
# Add each blocks to the line.
$line = $blocks[0];
unset($blocks[0]); # Do not add first block twice.
foreach ($blocks as $block) {
# Calculate amount of space, insert spaces, insert block.
$amount = $this->tab_width -
$strlen($line, 'UTF-8') % $this->tab_width;
$line .= str_repeat(" ", $amount) . $block;
}
return $line;
}
protected function _initDetab() {
#
# Check for the availability of the function in the `utf8_strlen` property
# (initially `mb_strlen`). If the function is not available, create a
# function that will loosely count the number of UTF-8 characters with a
# regular expression.
#
if (function_exists($this->utf8_strlen)) return;
$this->utf8_strlen = create_function('$text', 'return preg_match_all(
"/[\\\\x00-\\\\xBF]|[\\\\xC0-\\\\xFF][\\\\x80-\\\\xBF]*/",
$text, $m);');
}
protected function unhash($text) {
#
# Swap back in all the tags hashed by _HashHTMLBlocks.
#
return preg_replace_callback('/(.)\x1A[0-9]+\1/',
array($this, '_unhash_callback'), $text);
}
protected function _unhash_callback($matches) {
return $this->html_hashes[$matches[0]];
}
}