[Keith]

Ok, this is for a client of mine. This is the way they'd like it, and I'm so-so with regex's...

The program I wrote for them adds text to pages. With each page they are able to specify whether they want carriage returns to insert break tags or whether they want to insert break tags themselves.

Well, I have this client that wants it to be able to add tags that will allow them to turn on break tags, yet add tags that will remove any break tags in between.

Here's waht I'm talking about:

Code:

This is a paragraph. Paragraphs are as old as the hills, and they've never really changed.

This is a new line. Now I'm going to type some HTML:

<nobr>
<div align="center">
<em>This is italicized</em>.
</div>
</nobr>

Now we go back to automatic line breaks?

You see what I mean? They want each new line to add a break tag, but they'd like to have any break tags between the <nobr></nobr> tags stripped out, or not have them inserted to begin with (for exactly that purpose: html tags, etc...).

Any regex experts wanna have a go at it?

[rec0n]

You could try writing a function and replacing tags with the function result. For example:

Code:

function nobr(...) {...}
$mytext = "[nobr]This is a paragraph. Paragraphs are as old as the hills, and they've never really changed.

This is a new line. Now I'm going to type some HTML:

<nobr>
<div align="center">
<em>This is italicized</em>.
</div>
</nobr>[/nobr]

Now we go back to automatic line breaks?";
$mytext = preg_replace("#[nobr](.*?)[/nobr]#ie", "nobr('$1');", $mytext);

Now; I dunno if you get it but there's an idea for ya.

[Keith]

Ehhh, not sure if you understand what I meant, because I don't understand what that's supposed to do.

I got it figured out anyhow, just had to crack my head open and apply myself. Thanks though:

PHP Code:

preg_replace("/(\[)(nobr)(])(\r\n)*(.*)(\[\/nobr\])/esiU", "stripslashes(str_replace('<br />', '', '\\5'))", $text); 

[mikaelf]

Here we go. Consider the tag which disallow line break is [nobr]. Based on your problem, I think we have to do double regex check. First, to strip the line break in the tag and then to strip [nobr] tag from it.

Here is an example code:

PHP Code:

//$myText refers to the data where [nobr] tag may exist



$break = '#<br(\s)?(?(1)\/)>#i';



//Variable $myText points to the text you want to change.

if(preg_match_all("/\[nobr\](.*?)\[\/nobr\]/si",$myText,$match)) {

    for($i=0; $i<count($match[1]); $i++) {

        $replace[] = preg_replace($break,"",$match[1][$i]);    

        //Avoid finding the same identifier in the data

        $match[0][$i] = preg_replace("/#/","",$match[0][$i]);

        $matches[$i] = '#'.$match[0][$i].'#';

        print $replace[$i];

    }

    unset($match);

    $myText = preg_replace($matches,$replace,$myText);

} 

I don't know whether PHP can call function preg_replace within another preg_replace function. If it could, the above code should be optimised.

[mikaelf]

Originally Posted by Keith

Ehhh, not sure if you understand what I men, because I don't understand what that's supposed to do.

I got it figured out anyhow, just had to crack my head open and apply myself. Thanks though:

PHP Code:

preg_replace("/(\[)(nobr)(])(\r\n)*(.*)(\[\/nobr\])/esiU", "stripslashes(str_replace('<br />', '', '\\5'))", $text); 

Oh you've found it. Anyway, str_replace wouldn't substitute <BR>, <BR /> or <br>.

[Keith]

Yep, I use the XHTML break tag anyhow (nl2br() function to be more specific). I wasn't sure either if preg_replace could be used inside preg_replace... but on your suggestion I tried it and it works like a charm for both <br> as well as <br /> tags:

PHP Code:

preg_replace("/(\[)(nobr)(])(\r\n)*(.*)(\[\/nobr\])/esiU", "stripslashes(preg_replace('#<br(\s)?(?(1)\/)>#i', '', '\\5'))", $text); 

[mikaelf]

The optimised code should be:

PHP Code:

$text = preg_replace("/\[nobr\](.*?)\[\/nobr\]/esiU", "stripslashes(preg_replace('/<br(\s)?(?(1)\/)>/i', '', '\\1'))", $text); 



//or



preg_replace("/(\[)(nobr)(])(\r\n)*(.*)(\[\/nobr\])/esiU", "stripslashes(preg_replace('/<br(\s)?(1)\/)>/i', '', '\\5'))", $text); 

[mikaelf]

haha... finally it's done. thank you also. now i know that i can call multiple preg_replace which can save lot of my work... thanks again..

[Keith]

No problem, thanks for the optimization. I still have a lot to learn regarding regex's.

[adoxe]

a good optimization at this point would rather be to put these calls on separate lines since it doesn't accelerate nor lighten the process to put this in a single line. And what you'll get from this ? Tell me what if not a hardened way to maintain your code...