PHP Regular expression to identify a link from html

[deane034]

Hi,

 

I'm developing a wordpress theme, and I want to extract all the HTML Links (hrefs) with the extension of MP3 into my side bar. I know the wordpress functions to get posts, etc. What I want to know is how to use PHP regular expressions to get a link for an .mp3 that exist as a link from a string variable.

 

For example suppose the string is,

$mystring= "<p> Hello this is a WP post. <a href="somefile.mp3">Here is a link to mp3<a/> </p> ";

 

I want to be able to have a function which returns "somefile.mp3"

 

I'm a bit of a noob when it comes to PHP regular expressions, Any help would be much appreciated.

 

Thanks,

Deane.

 

 

[premiso]

May not be the best way, but will match it as long as there is only 1 mp3 link per line.

 

preg_match('~href=[\'|"](.*\.mp3)[\'|"]~i', $mystring, $match);

 

[cags]

Using the alternation bar inside a character class will simply allow the 'bar' character to match in the pattern, I think you were either thinking of using a capture group with an alternation or a character class. I'd personally opt for something along the lines of...

 

~href=([\'"])([^\1 ]+\.mp3)\1~i

 

It uses back referencing to make sure the string delimiters of the href match. It will also cut out a bit sooner on hrefs that don't match since it will only match until it finds a space or the opening delimiter again. I believe technically an attribute doesn't require delimiters if it doesn't have spaces in it (which a URL obviously won't have), but I'm not entirely sure how you could accurately add that to the pattern, most people avoid that system anyway, so hopefully it's not a requirement.

[cags]

Thanks to salathe for pointing out (via IRC) my, erm... deliberate mistake. That's what you get for posting before your morning coffee. A back reference can't be used in a character class. The \1 inside the character class actually represents the ascii character with the octal number 1 (SOH a control character), based on that it would perhaps make as much sense to simply use...

 

~href=([\'"])(\S+\.mp3)\1~i