[dannzyx]

Good day!

I am currently learning the whole preg_match abacadabra.
To do this I decided to write a small app. to search within html tags of a page;
However I stumbled upon a problem when trying to search within alt tags:

I only want to search within the alt tags of images.
Using this code:

preg_match_all('/(<img\s) .* alt="(\w*)" .* (>)/ismxU', $code, $result);

It does not work the way I want it to, it does not seem to search only within the image tags...

Can anyone point me in the right direction?

Thanks in advance!

[grimpirate]

Try changing this
/(<img\s) .* alt="(\w*)" .* (>)/ismxU
To this
/(<img\\s) .* alt="(\\w*)" [^>]* (>)/ismxU

This post has been edited by grimpirate: 27 March 2011 - 10:18 AM

[dannzyx]

Thanks for your reply, however it still only gives me the alt content outside of the img tag.

[codeprada]

PCRE functions are crazy stuff

But however this retrieves the alt text in IMG links. It's hard to explain but I'm using (.+?) to match all the characters within the quotes in an <img /> tag.

$str = '<a href="link.html" alt="Link">Link Here</a> <img src="image.jpg" alt="Image" />';
preg_match_all("#<img.+?alt=\"(.+?)\" />#", $str, $matches);
var_dump($matches);

[grimpirate]

I would suggest the following changes to codeprada's expression:
'#<img[^>]+?alt="([^>]+?)"[^>]+?>#'
The reasons are as follows:
It's easier to read if you don't have to escape the double quotations.
As I understand it a negated character class matches more efficiently than a . multi-character, and in order for the . multi-character to match newlines you have to add the 's' flag to your expression.
It's possible the img tag may not close elegantly with a space, a slash, and a greater than sign, but we do know it definitely terminates with a greater than sign.

This post has been edited by grimpirate: 27 March 2011 - 10:47 PM

[dannzyx]

Still didnt do the job :<
Thanks for the additional explanation grim.
it's very helpfull in understanding the whole matter.

Ill let you know if I find a solution