Regex question

[Enigmabomb]

Hey Guys,

I'm getting better with regex, but this query is giving me a mother of a time. I want to capture the information between two arbitrary html tags, with no back references. (so it'll parse less shit)

in this case, the tag is named <Features> and </Features>

preg_match("<Features>(.*?)<//Features>",$description,$urlarray);


Warning: preg_match() [function.preg-match]: Unknown modifier '(' in


I know I'm close, and it's just missing some stupid piece of something. Anyone see where Im completely fucking this up?

thanks.

josh

 

[ambirex]

You are missing the /'s to define the pattern

<?php

$html = '<feature someatt="foo" otheratt="bar">blah <b>Blah</b> blah</feature>';

$regex = '/<feature(.*)>(.*)<\/feature>/';

preg_match($regex, $html, $matches);

// what you want is in matches[0]
echo "<pre>"; print_r($matches); echo "</pre>";

?>

 

[ChuckleFuck]

preg_match("/<Features>(.*?)<\/Features>/", $description, $urlarray);

preg_match("/<Features>(.*?)<\/Features>/i", $description, $urlarray); to make it case-insenstive

 

[argh01]

I prefer to use | for the sides of my regexes; that way you don't have to escape the slash in the close tag.

But yeah, what he said.

 

[Enigmabomb]

Damnit, I was so close. Thanks guys!

Josh

 

[Enigmabomb]

I'm still having problems even with the above Regexs.

Here's what I'm after:

<FEATURES>Ass</FEATURES>

I want it to return Ass

 

[ambirex]

I'm guessing you might be trying to match several instance of ' <FEATURES>Ass</FEATURES>' with the same document, if this is the case then you want to use preg_match_all, also you might want to add the greedy flag ( 'U' ) to the regex. So for instance:

<?php

$html = '<FEATURES>Ass</FEATURES><FEATURES>Ass2</FEATURES>';

$regex = '/<FEATURES>(.*)<\/FEATURES>/Ui';

preg_match_all($regex, $html, $matches);

echo "<pre>"; print_r($matches); echo "</pre>";

?>

without the greedy flag it will match the broadest sense of the pattern

ungreedy match ( 1 match )
<FEATURES>Ass</FEATURES><FEATURES>Ass2</FEATURES>

greedy match ( 1 match )
<FEATURES>Ass</FEATURES><FEATURES>Ass2</FEATURES>

greedy match with preg_match_all ( 2 matches )
<FEATURES>Ass</FEATURES><FEATURES>Ass2</FEATURES>

 

[nis]

Didn't know about the U-thing. I just add a ? after my "how many of these do you want to find". Like this:

$regex = '/<FEATURES>(.*?)<\/FEATURES>/i';

 

[Enigmabomb]

The greedy tag is interesting, I also didn't know about that. The pattern matches the hwole thing, Tags INCLUDED. Now that I think about it, I can use str_Replace to filter out the tags rather than making a regex to pluck between the tags.

thanks for the interesting feedback

josh