[regex] Why doesn't preg_replace work?

[Gilles Ganault]

Hello

I went through some examples, tried a bunch of things... but still
can't figure out why I can't extract the TITLE section of a web page
using preg_replace():

-----------
<?php

$url = "http://www.cnn.com";

$response = file_get_conten ts($url);

$output=preg_re place("|<title> (.+?)</title>|smiU",
"TITLE=$1",
$response);

$fp = fopen ("output.htm l", "w");
fputs ($fp,$output);
fclose($fp);
-----------

Any idea?

Thanks!

[klenwell]

Re: Why doesn't preg_replace work?

Hi Gilles,

I'm not a regex guru, but I can see a spot a couple problem areas in
your expression:

1. The core syntax could probably be simplified using something like
this:

|^<title>([^<]+)</title>$|i

I hope I got that right -- I usually have to test my expression a few
times before I get all the nuances right. :)

2. smiU - That's modifier overkill. The U here and the ? in your
expression are probably reacting to each other in unexpected ways. If
you don't know about this page, it can help:

http://www.php.net/manual/en/reference.pcre.pattern.modifiers.php

I have a prefab function I've used for this very thing, but
unfortunately I don't have access to it that moment. Hopefully,
someone will be along shortly with the proper syntax. In the
meantime, I hope this helps in a more general sense.

Regards,
Tom



On May 7, 4:30 pm, Gilles Ganault <nos...@nospam. comwrote:

Hello
>
I went through some examples, tried a bunch of things... but still
can't figure out why I can't extract the TITLE section of a web page
using preg_replace():
>
-----------
<?php
>
$url = "http://www.cnn.com";
>
$response = file_get_conten ts($url);
>
$output=preg_re place("|<title> (.+?)</title>|smiU",
"TITLE=$1",
$response);
>
$fp = fopen ("output.htm l", "w");
fputs ($fp,$output);
fclose($fp);
-----------
>
Any idea?
>
Thanks!

[Gilles Ganault]

Re: Why doesn't preg_replace work?

On 7 May 2007 16:46:27 -0700, klenwell <klenwell@gmail .comwrote:

>2. smiU - That's modifier overkill. The U here and the ? in your
>expression are probably reacting to each other in unexpected ways.

Ah, ah... Indeed, it seems like it's either using the U switch to make
Preg non-greedy, or use the ? limiter (eg. ".+?"). Thanks for pointing
it out.

Found it: To extract bits, I shouldn't use preg_replace() but
preg_match():

--------------
$url = "http://www.cnn.com";
$response = file_get_conten ts($url);

preg_match("|<t itle>(.+?)</title>|smi",$re sponse,$matches );
$response = $matches[1];

$fp = fopen ("output.htm l", "w");
fputs ($fp,$response) ;
fclose($fp);
--------------

Thank you.