preg_match_all

[dreamwest]

Ive almost got this but i cant seem to get it to stop at the first /

 

$link = "site.com/bfdgnh/oc8687/funny_names.jpg";
if (preg_match_all("/\/(.*)\.jpg/i", $link, $titles)) {

print_r($titles);

}

 

This outputs:

 

Array ( [0] => Array ( [0] => /bfdgnh/oc8687/funny_names.mp3 ) [1] => Array ( [0] => bfdgnh/oc8687/funny_names ) )

 

 

 

but i just need the last part "funny_names"

[Garethp]

This really belongs in the regex section. Try

 

"~/(.*?)\.jpg~s" as your pattern

[knsito]

Start your string with "http://"

and use parse_url() and pathinfo()

 

$link = "http://site.com/bfdgnh/oc8687/funny_names.jpg";
$p = parse_url($link);
var_dump($p);
$p = pathinfo($p['path']);
var_dump($p);
echo $p['filename'];

[MadTechie]

I would probably do this,

 

if (preg_match('%([^/.]+)\.?[^/.]{0,3}$%', $link, $regs)) {
$titles= $regs[1];
} 

 

EDIT: typo

[thebadbad]

basename() is made for this:

 

$link = 'site.com/bfdgnh/oc8687/funny_names.jpg';
echo basename($link, '.jpg');

[MadTechie]

basename is the best option if it's always .jpg, but I assumed the extension would be anything.

[thebadbad]

basename is the best option if it's always .jpg, but I assumed the extension would be anything.

 

Could alternatively be taken care of by using substr() and strrpos():

 

$link = 'site.com/bfdgnh/oc8687/funny_names.jpg';
$filename = basename($link);
$lastdot = strrpos($filename, '.');
if ($lastdot !== false) {$filename = substr($filename, 0, $lastdot);}
echo $filename;

[MadTechie]

You could but that's would probably be slower than the RegEx.

[thebadbad]

Probably. But then again, your pattern fails when the filename contains a dot

[MadTechie]

that's easy to fix, just drop the greediness

 

$link = "site.com/bfdgnh/oc8687/funny_n.ames.jpg";
if (preg_match('%([^/]*?).?[^/.]{0,3}$%', $link, $regs)) {
$titles= $regs[1];
} 
echo $titles;

[thebadbad]

Yeah, I know, I'm just mocking you Fun fact: My method using the built in string functions is ~ 38 times faster than your method using regex (0.052 seconds versus 1.965 seconds when run 10000 times in a loop).

 

Edit: And actually, your pattern cuts the last three chars off of the filename if it hasn't got any extension. Might not be relevant in the OP's case, but could be.

 

But in the big picture, I prefer to write and use code that is easily readable, even at the possible (slight) cost of efficiency. And when the simpler method in this case is faster than the more complicated, harder to read regex, I find it an easy choice. But each man has his own preferences.

[MadTechie]

I ran some tests here and yes its quicker but i had different results

(i fixed the no dot problem)

only real issue with the basefile is if you have a dot but no extention

ie hello.world would be dropped the world, so the RegEx has a little more control but slightly slower.. i hashed the names to stop the caching.

 

<?php
//0.255233
$start = microtime();
for($i=1;$i<5000;$i++)
{
$h = md5($i);
$link = "site.com/bfdgnh/$h/$h.jpg";
$filename = basename($link);
$lastdot = strrpos($filename, '.');
if ($lastdot !== false) {$filename = substr($filename, 0, $lastdot);}
}
echo microtime() - $start."\n";
?>

<?php
//0.306391
$start = microtime();
for($i=1;$i<5000;$i++)
{
$h = md5($i);
$link = "site.com/bfdgnh/$h/$h.jpg";
preg_match('/(.*?)(?:\.[^.]{0,3})?$/', $link,$regs);
$titles= $regs[1];
}
echo microtime() - $start."\n";
?>

 

as we have taken over this thread.. wanna run the scripts at your side see if you get the same..

[thebadbad]

only real issue with the basefile is if you have a dot but no extention

ie hello.world would be dropped the world

 

Technically there is no such thing. world would be the extension in that example. Extensions can be longer than 3 chars, so it doesn't make sense to keep them based on length.

 

And good idea with the hashing. When I run your code the string function method is still over 4 times faster than the regex method (0.045751 seconds versus 0.200152 seconds). I'm testing with PHP 5.2.6 @ Apache 2.2.9 on a quad-core Vista box.