PREG? how to remove useless <FONT> AND </FONT> tags from a string?

sneakyimp

hey everybody:

i have a tricky pattern-matching problem here. i'm terrible at pattern matching (but learning fast). i have successfully used EREG to get rid of some nasty HTML crap that i don't want from a string...however, the results leave some situations like this:

$string = "This is all good and everything. <FONT>But this font tag has got to go<font color=#ff0000>but this font tag has absolutely got to stay</font>as does that end tag</font>but that last /font tag must go for sure";

Do I need to loop a PREG statement? Seems to me I'll need some kind of recursive preg thing going on and i'm frightened by pattern matching...it's so POWERFUL.

any hints?

ALSO: I'm kinda confused about how linebreaks come into play with EREG and PREG...doesn't ^ refer to the 'beginning of the line' or something? My next pattern matching task is to extract the entire BODY of an HTML document minus the <body> and </body> tags and anything before or after.

any help would be much appreciated.

mrhappiness

try

echo preg_replace('%(<font>)(.*)(</font>)%si', '$2', $string);

if you replace font by body it should suit the 2nd problem also

Buddha443556

How about a pattern (PREG) like this:

/<\/?FONT[^>]*>/is

Should get rid of open and close tag.

Yes ^ does refer to the begin of a line but is also negation operator in sets when the first element. [^a] equal everything that's not an letter a.

mrhappiness

won't this remove all font tags?

i thought he only wants to remove font tags with no attributes specified

sneakyimp

i will test these...as food for thought, though, what about this:

<font>get rid of that one
<font color='#ff0000'>but not that one
<font>but definitely that one</font>and that one
</font>but not that one
<font color='#ff0000'>and not that one</font>or that one
</font>but yes that one

that's what i mean by recursion....in theory, there could be font tags within font tags...i certainly hope not though.

Buddha443556

Originally posted by mrhappiness
won't this remove all font tags?

i thought he only wants to remove font tags with no attributes specified

Whoops your right, glade I went to bed might of started trusting user data next.

sidney

 <?php 
$txt="<font>get rid of that one 
<font color='#ff0000'>but not that one 
<font>but definitely that one</font>and that one 
</font>but not that one 
<font color='#ff0000'>and not that one</font>or that one 
</font>but yes that one";
$font= explode("<font",$txt);
foreach($font as $key => $line)
	{
	if(substr($line,0,1)==" ")
		{
		$line="<font".$line;
		$bits=explode("</font>",$line);
		if(!empty($bits[1])){$bits[1]="</font>".$bits[1];}
		$line=implode(" ",$bits);
		}
		else
		{
		$line=substr($line,1);
		$bits=explode("</font>",$line);
		if(!empty($bits[1])){$bits[1].="</font>";}
		$line=implode(" ",$bits);		
		}
	$font[$key] = $line;	
	}
$txt= implode("",$font);
echo $txt;
?>