Find and Replace (Dynamically)

xmatthawkx

Hi all,

I've searched the forums, and checked out the Regular Expressions article here on PHPBuilder, but I'm still unable to accomplish my goal, after hours of tearing my hair out. :p. I guess I'll start off by explaining, what I need to do:

I have a string (it is always different - and always a different length). Within this string there are HTML-Like tags ([lu] and [/lu]). Between those tags, there is a word, which is always different. So, the string looks similar to this:

"This is a [lu]test[/lu] of the search and [lu]replace[/lu] function."

I need to find all instances of the [lu]x[/lu] tags, determine what is between them (x) - and replace them with this string:

"<a href=\"glossary.php?w=x\">x</a>"

Where "x" is the word originally enclosed in the [lu] [/lu] tags.

Hopefully that makes some sense :p.

I've found a few examples of regular expressions that will allow me to retrieve the word between the tags, but I have been unable to successfully read the word and replace the specific sections of my string with the modified <a> </a> tags.

If anyone understands what I'm trying to do, and you can help me...pleaseeeee do! 🙂 Thanks in advance, I really appreciate it!

planetsim

This is quite easy when you understand regex. Its hard to comprehend for a while and as its not very well documented its hard to even start to learn

Now a link that taught me a little

http://tutorialforums.com/showthread.php?s=&threadid=16817&highlight=bbcode

You could also look on this site in the articles where there is a quite good explanation of it.

xmatthawkx

Thanks man, that's exactly what I'm looking for 🙂!

I've tried it out though, and here's what happens:

A simple:

 $htmlcode=stripslashes($htmlcode);
 $htmlcode = preg_replace('#[lu](.+?)[/lu]#si', '<a href="glossary.php?w=\1">\1</a>', $htmlcode);
 echo $htmlcode;

(NOTE: vB removed the \ from in front of the 1s, in the example above they show as "1", in my code they are "\1")

Where $htmlcode is a user inputted string, in this case:

"This might [lu]work[/lu] if I am really lucky."

Returns:

"this might [<a href="glossary.php?w=u]work[">u]work[</a><a href="glossary.php?w=u] if i rea">u] if i rea</a><a href="glossary.php?w=y am ">y am </a>ucky."

Any idea what might be causing this sporadic and seemingly random insertion/modification of the code?

xmatthawkx

 $htmlcode=stripslashes($htmlcode);
 $lookups = '/(\[lu\])(.+)(\[\/lu\])/';
 $htmlcode = preg_replace($lookups,'<a href="glossary.php?w=\\2">\\2</a>',$htmlcode);
 echo $htmlcode;

Ended up working, but only to a limited extent. It only succeeds if there is exactly ONE instance of the tags in the string. Greater than 1, and it goes crazy. Any ideas?

Thanks!

xmatthawkx

I understand how and why multiple instances fail, but I need some guidance on how to fix it. The expression looks at the very first and very last tags, it then works its way inward until no tags are remaining. It somehow, needs to be modified to go one set of tags at a time. Any ideas?

xmatthawkx

I figured I'd try something along these lines, to see if it would work, I think it will but there must be a problem in loop, because I think it runs infinetly:

 $lookups = '/(\[lu\])(.+)(\[\/lu\])/';
 $htmlcode=stripslashes($htmlcode);

 $nextlu=strpos($htmlcode, "[/lu]");
 $htmlcode_thispart_original=substr($htmlcode, 0, $nextlu);
 $htmlcode_thispart = preg_replace($lookups,'<a href="glossary.php?w=\\\2">\\\2</a>',$htmlcode_thispart_original, 1);
 $startfrom=$nextlu;
 $nextlu=strpos($htmlcode, "[/lu]", $nextlu);

 while($nextlu)
 {
        $htmlcode_thispart_original=substr($htmlcode, $startfrom, $nextlu);
        $htmlcode_thispart .= preg_replace($lookups,'<a href="glossary.php?w=\\\2">\\\2</a>',$htmlcode_thispart_original, 1);
        $startfrom=$nextlu;
        $nextlu=strpos($htmlcode, "[/lu]", $nextlu);
 }

 $htmlcode=$htmlcode_thispart;

If anyone could help me fix my code and/or suggest a more efficent way to do this, I would really appreciate it. I reallly need to get this done. Thanks 🙂!

xblue

Hi,

the most obvious problem to me seems that [] has special meaning and therefore need to be escaped.

Try it like this:
$htmlcode = preg_replace('#[lu](.+?)[/lu]#si', '<a href="glossary.php?w=$1">$1</a>', $htmlcode);

(the questionmark after .+ is needed to match ungreedy, otherwise it will only find one match that extends from the first [lu] to the last [/lu], which is probably not what you want)

hope this helps...

xmatthawkx

xblue: you da' man 😉!

I used the expression i posted in my second post, and threw a ? after the + like you suggested. That fixed all my problems 🙂.

While I have your attention, I'm just curious if it's possible to trim() the match before replacing it?

Thanks man!

xblue

There may be some solution with the e-modifier (see http://www.php.net/manual/en/pcre.pattern.modifiers.php for details), but I think that would be overkill, I'd prefer to do it right within the regex.

Try adding \s* before and after (.+?).

That would match any whitespace at the beginning and end of the string.

And make sure you put it outside the parenthesis, so it's not caught in the subpattern ($1 in my example) you reuse in your replacement argument. The replacement argument shouldn't need to be changed.