[SOLVED] Automatically anchoring URLs that are *not* inside [code] tags

[Goldeneye]

The background: I have a function that formats text with bold, italics, underlines, images, and automatically anchored URLs. I have a tag (I call it the [raw] tag as in raw-text) for leaving anything inbetween the tags unformatted.

 

The problem: URLs inside the [raw]*[/raw] tags still get automatically anchored.

 

The code:

$strs = array('This is a [i]string[/i] with a link: http://foobar.com/ --  no twix for you.',
	'[raw]This is a [i]string[/i] with a link: http://foobar.com/ -- no twix for you[/raw]'
	);
foreach($strs as $str){
	preg_match_all('/\[raw\](.*?)\[\/raw\]/si', $str, $code);
	foreach($code[1] as $stri) $str = '<span style="font-family: monospace; white-space: pre;">'.str_replace('[', '&#91;', $stri).'</span>';
	// THIS IS WHERE THE RAW-TAG REPLACEMENT ENDS
	// ##################################
	// NOW FOR PARSING THE REGULAR TAGS THAT HAVEN'T HAD THEIR LEFT-SQUARE BRACKETS REPLACED:
	$search = array('/^[\&\#91\;raw\]]((mailto:|(http|ftp|nntp|news):\/\/).*?)(\s|<|\)|"|\\\\|\'|$)^[\&\#91\;\/raw\]]/', '/\[i\](.*?)\[\/i\]/si');
	$replace = array('<a href="$1" rel="nofollow" target="_blank">$1</a>$4', '<i>$1</i>');
	$str = preg_replace($search, $replace, $str);
	echo '<p>'.$str.'</p>';
}

 

What I expect to happen is that the element of the array will be:

This is a <i>string</i> with a link: <a href="http://foobar.com/">http://foobar.com/</a> --  no twix for you.

 

What I expect the second element of the array to look like:

This is a [i]string[/i] with a link: http://foobar.com/ -- no twix for you

 

The breakdown of my automatically anchoring Regular-Expression:

'/^[\&\#91\;raw\]]

((mailto:|(http|ftp|nntp|news):\/\/).*?)(\s|<|\)|"|\\\\|\'|$)

^[\&\#91\;\/raw\]]/'

What this should do is that it looks for URLs not in-between [raw]*[/raw] tags and anchors them. I use ^[\&\#91\;\/raw\]] instead of ^[\[\/raw\]] because the raw-tag works by replacing the first square bracket ([) with it's HTML-entity equivalent ([).

 

A preemptive thanks to any input/advice.

[thebadbad]

Example:

 

<?php
$strings = array(
'This is a [i]string[/i] with a link: http://foobar.com/ --  no twix for you.',
'[raw]This is a [i]string[/i] with a link: http://foobar.com/ -- no twix for you[/raw]'
);
//define BB code
$replace = array(
'~\[i\](.*?)\[/i\]~is' => '<em>$1</em>',
'~\b(?:mailto:|(?:https?|ftp|nntp|news)://)\S+~i' => '<a href="$0">$0</a>'
);
//define callback function for raw content
function raw_convert($str, $reverse = false) {
$raw = array(
	'[' => '&91;',
	']' => '&93;',
	':' => '&58;'
);
if ($reverse) {
	return str_replace($raw, array_keys($raw), $str);
} else {
	return str_replace(array_keys($raw), $raw, $str);
}
}
foreach ($strings as $string) {
//process raw tags
$out = preg_replace_callback(
	'~\[raw\](.*?)\[/raw\]~is',
	create_function(
		'$matches',
		'return \'[raw]\' . raw_convert($matches[1]) . \'[/raw]\';'
	),
	$string
);
//convert BB code
$out = preg_replace(array_keys($replace), $replace, $out);
//reconvert char substitutes in raw content
$out = preg_replace_callback(
	'~\[raw\](.*?)\[/raw\]~s',
	create_function(
		'$matches',
		'return raw_convert($matches[1], true);'
	),
	$out
);
echo "$out<br /><br />";
}
?>

 

Only drawback I can think of, is that if any of the HTML entities from the raw_convert() function are present inside a pair of [raw] tags, they will be replaced by their character equivalent. But to overcome that you could simply use some other 'unique' tokens instead of their HTML entities.

[Goldeneye]

Thank you, thebadbad! I had no idea it was more complex than just writing a regular-expression. This sort of task seems like something a reuglar-expression would be great for. But anyway, thank you, again.