regex nightmare: anyone want to check it? - GoFuckYourself.com

fris · 09-15-2009, 08:35 AM

My youtube wordpress plugin works via

[!youtube]L1BDM1oBRJ8[/youtube!]

I also wanted to be able to do it via

https://youtube.com/watch?v=L1BDM1oBRJ8
https://youtube.com/watch?v=L1BDM1oBRJ8
www.youtube.com/watch?v=L1BDM1oBRJ8
youtube.com/watch?v=L1BDM1oBRJ8
watch?v=L1BDM1oBRJ8
v=L1BDM1oBRJ8
L1BDM1oBRJ8
https://youtube.com/v/L1BDM1oBRJ8

Code:

$regex = "/\[(?:(?:http:\/\/)?(?:www\.)?youtube\.com\/)?(?:(?:watch\?)?v=|v\/)?([a-zA-Z0-9\-\_]{11})(?:&[a-zA-Z0-9\-\_]+=[a-zA-Z0-9\-\_]+)*\]/";

borked · 09-15-2009, 08:41 AM

well, the simplest is to optionally allow each component:

(http:\/\/)?(www\.)?(youtube\.com)?(\/)?(watch\?)?(v=)?([a-zA-Z0-9]{11})

but then this means that any optional component can be left out eg:
http://v=L1BDM1oBRJ8

is valid.

BestXXXPorn · 09-15-2009, 08:42 AM

That should work, it might be a little overkill with all the valid character checking though :P

You could also:

$arr = explode('=', $youTubeURL);

and validate just the portion you want:

$linkCode = $arr[count($arr)-1];

While that wouldn't work with the last case you posted you could just do a:

$youTubeUrl = str_replace('/', '=', $youTubeUrl);

before hand since you don't care what the domain is anyway... It would definitely be much faster than using regex, parsing wise :P

BestXXXPorn · 09-15-2009, 08:51 AM

Running benchmarks on regex (using borked's simplified regex) vs explode (since we're trying to strip the actual value for v...

I show explode method as twice as fast as regex ;)

09-15-2009, 08:35 AM	#1
fris Too lazy to set a custom title Industry Role: Join Date: Aug 2002 Posts: 55,372	regex nightmare: anyone want to check it? My youtube wordpress plugin works via [!youtube]L1BDM1oBRJ8[/youtube!] I also wanted to be able to do it via https://youtube.com/watch?v=L1BDM1oBRJ8 https://youtube.com/watch?v=L1BDM1oBRJ8 www.youtube.com/watch?v=L1BDM1oBRJ8 youtube.com/watch?v=L1BDM1oBRJ8 watch?v=L1BDM1oBRJ8 v=L1BDM1oBRJ8 L1BDM1oBRJ8 https://youtube.com/v/L1BDM1oBRJ8 Code: $regex = "/\[(?:(?:http:\/\/)?(?:www\.)?youtube\.com\/)?(?:(?:watch\?)?v=\|v\/)?([a-zA-Z0-9\-\_]{11})(?:&[a-zA-Z0-9\-\_]+=[a-zA-Z0-9\-\_]+)*\]/"; __________________ Since 1999: 69 Adult Industry awards for Best Hosting Company and professional excellence. WP Stuff

09-15-2009, 08:41 AM	#2
borked Totally Borked Industry Role: Join Date: Feb 2005 Posts: 6,284	well, the simplest is to optionally allow each component: (http:\/\/)?(www\.)?(youtube\.com)?(\/)?(watch\?)?(v=)?([a-zA-Z0-9]{11}) but then this means that any optional component can be left out eg: http://v=L1BDM1oBRJ8 is valid. __________________ For coding work - hit me up on andy // borkedcoder // com (consider figuring out the email as test #1) All models are wrong, but some are useful. George E.P. Box. p202

09-15-2009, 08:42 AM	#3
BestXXXPorn Confirmed User Join Date: Jun 2009 Location: Asheville, NC Posts: 2,277	That should work, it might be a little overkill with all the valid character checking though :P You could also: $arr = explode('=', $youTubeURL); and validate just the portion you want: $linkCode = $arr[count($arr)-1]; While that wouldn't work with the last case you posted you could just do a: $youTubeUrl = str_replace('/', '=', $youTubeUrl); before hand since you don't care what the domain is anyway... It would definitely be much faster than using regex, parsing wise :P __________________ ICQ: 258-202-811 \| Email: eric{at}bestxxxporn.com

09-15-2009, 08:51 AM	#4
BestXXXPorn Confirmed User Join Date: Jun 2009 Location: Asheville, NC Posts: 2,277	Running benchmarks on regex (using borked's simplified regex) vs explode (since we're trying to strip the actual value for v... I show explode method as twice as fast as regex ;) __________________ ICQ: 258-202-811 \| Email: eric{at}bestxxxporn.com