Regular expression trouble

servo

I'm trying to read in an HTML page and wipe out any javascript that might be living on it... but I'm having trouble with the following call:

$nojava = eregi_replace("<SCRIPT.*?</SCRIPT>", " ", $htmlwithjava);

since there could be 1..n script blocks, I obviously want to use the non-greedy (otherwise it would wipe out everything between* two blocks as well). Near as I can tell, PHP does not seem to appreciate the ?.

The above code produces a REG_BADRPT error. Not including the ? allows the code to work with no error, but in "greedy" mode.

Anyone know why the ? throws an error?

Thanks!
-Carl

Anon

why not try this :

eregi_replace("(<SCRIPT>{1})(.*)(</SCRIPT>{1})","\2",$txt);

Anon

Try this. I stay away from the ereg functions because they are slower and less powerful than the preg functions.
preg_replace("/<SCRIPT.*?<\/SCRIPT>/ims", " ", $htmlwithjava);

servo

If I'm reading that correctly, LM, it's replacing the SCRIPT block with everything that's inside the SCRIPT block... in effect, stripping out the <SCRIPT> tags. Not quite what I was after. But

eregi_replace("(<SCRIPT>{1})(.*)(</SCRIPT>{1})"," ",$txt);

doesn't work either... because it operates in "greedy" mode (matches as much as it can get). Meaning, if there are, for example, two script blocks on the page with valuable HTML in between:

<HEAD>

<SCRIPT>
function blahblah...
</SCRIPT>

</HEADgt;

<BODY>
<P>Hello World. This is the content of my page.</P>

<SCRIPT>
document.write('Some inline JS stuff');
</SCRIPT>

</BODY>

eregi_replace would replace everything from the beginning tag of the first block to the ending tag of the second block... wiping out everything including the closing HEAD tag, the opening BODY tag and the content of the page.

I need to be able to throw the non-greedy switch (like '?' in PERL). And that's what seems to be causing the problem.

Thanks.
-Carl

servo

Thanks PMan! That worked great! I'm not sure what the difference truly is between what I did, and what you showed... but yours works and mine doesn't.

One curiousity... why the double escape before the /SCRIPT ? I'm sure it makes perfect sense, I'm probably just too tired to realize why.

Thanks again!
-Carl

P.S. - I never WAS terribly clear as to the difference between ereg and preg. Reading the man... they appear to be the same thing??

servo

Shoot.

Nevermind... I caught on. I'm more tired than I thought.

Of course that would work, LM. Sorry for doubting you!

Thanks!
-C