[RESOLVED] Regular Expression to match section of link. Very stuck.

The-Master

Hi.

I am terrible at regular expressions so was wondering if you could help me make one that would match the bit between the 2nd and 3rd underscore. The numbers are variable. Here is an example of the parts it will match.:

"http://photos-e.ak.fbcdn.net/hphotos-ak-snc1/hs132.snc1/5653_504358248751_308030211_139556_1421132_s.jpg"
"http://photos-d.ak.fbcdn.net/hphotos-ak-snc1/hs132.snc1/5653_504158233761_308030211_139555_4325138_s.jpg"

The part it should extract is "308030211".

Thanks in advance.

🙂

dagon

if its always in the same position explode() on the underscore

The-Master

Sounds good! Do you know if that will that be slower than using regular expressions because I have a lot of these to separate so speed will be quite a big factor?

Thanks.

dagon

The-Master;10927333 wrote:
Sounds good! Do you know if that will that be slower than using regular expressions because I have a lot of these to separate so speed will be quite a big factor?

Thanks.

benchmark it and find out.

The-Master

I don't have a regular expression though! Other than this - '/(.*?)_/' - which only works because it matches the underscore after everything. I don't know how to get it to match 2 underscores before the quantifier.

Kudose

Just for fun and a slap stick regex ...

<?php

    $string = 'http://photos-e.ak.fbcdn.net/hphotos-ak-snc1/hs132.snc1/5653_504358248751_308030211_139556_1421132_s.jpg';

    $start_regex = microtime(true);
    for($i = 0; $i < 5000; $i++)
            preg_match('#.*_([^_]+)_.*_.*_.*#', $string);
    echo 'Regex took ' . (microtime(true) - $start_regex) . ' seconds.' . "\n";

    $start_explode = microtime(true);
    for($i = 0; $i < 5000; $i++)
            explode('_', $string);
    echo 'Explode took ' . (microtime(true) - $start_explode) . ' seconds.' . "\n";

    $start_split = microtime(true);
    for($i = 0; $i < 5000; $i++)
            split('_', $string);
    echo 'Split took ' . (microtime(true) - $start_split) . ' seconds.' . "\n";

root@server [/home/lamppdev/phptests]# php -f split_explode_test.php
Regex took 0.0194711685181 seconds.
Explode took 0.0091540813446 seconds.
Split took 0.0346639156342 seconds.
root@server [/home/lamppdev/phptests]# php -f split_explode_test.php
Regex took 0.0202159881592 seconds.
Explode took 0.00892806053162 seconds.
Split took 0.0341391563416 seconds.
root@server [/home/lamppdev/phptests]# php -f split_explode_test.php
Regex took 0.0192739963531 seconds.
Explode took 0.0103480815887 seconds.
Split took 0.0351750850677 seconds.
root@server [/home/lamppdev/phptests]# php -f split_explode_test.php
Regex took 0.0201539993286 seconds.
Explode took 0.00944900512695 seconds.
Split took 0.0367538928986 seconds.
root@server [/home/lamppdev/phptests]#

I have always found explode to be fastest for these simple tasks.

Weedpacket

Kudose wrote:
I have always found explode to be fastest for these simple tasks.

Which makes sense of course; explode() doesn't need to parse its argument to find out what ti split on, nor build and run a finite automaton to traverse the string.

The-Master

Excellent! Thanks very much for doing that for me Kudose. I Appreciate it! 🙂