YouTube video ID regex

gistfile1.rb

# Parses YouTube URLs directly or from iframe code.  Handles:
# * Address bar on YouTube url (ex: http://www.youtube.com/watch?v=ZFqlHhCNBOI)
# * Direct http://youtu.be/ url (ex: http://youtu.be/ZFqlHhCNBOI)
# * Full iframe embed code (ex: <iframe src="http://www.youtube.com/embed/ZFqlHhCNBOI">)
# * Old <object> tag embed code (ex: <object><param name="movie" value="http://www.youtube.com/v/ZFqlHhCNBOI">...)

/(youtu\.be\/|youtube\.com\/(watch\?(.*&)?v=|(embed|v)\/))([^\?&"'>]+)/
$5 #=> the video ID

# test it on Rubular: http://rubular.com/r/eaJeSMkJvo

Thanks! Just one comment for other people coming across this, you may need to escape the "' symbols:
/(youtu.be/|youtube.com/(watch?(.*&)?v=|(embed|v)/))([^\?&\"\'>]+)/

This seems to be a better regex (ig flags) - It is much more flexible to the input text:

Match is in 1st match group, all other groups are non-capturing:
(?:youtube(?:-nocookie)?\.com\/(?:[^\/\n\s]+\/\S+\/|(?:v|e(?:mbed)?)\/|\S*?[?&]v=)|youtu\.be\/)([a-zA-Z0-9_-]{11})\W  

Tested against this text at regex101.com:

 var url = "http://www.youtube.com/user/Scobleizer#p/u/1/1p3vONEsYGo
url = "http://youtu.be/NLqAFTWOVbY
var url = "http://www.youtube.com/embed/NLqTHREEVbY https://www.youtube.com/embed/NLqAFOURVbY
var url = "http://www.youtube.com/v/NLqAFFIVEbY?fs=1&hl=en_US
var url = "http://www.youtube.com/watch?v=NLqASIXrVbY
var url = "http://www.youtube.com/user/Scobleizer#p/u/1/1pSEVENsYGo
var url = "http://www.youtube.com/ytscreeningroom?v=NRHEIGHTx8I
var url = "http://www.youtube.com/user/Scobleizer#p/u/1/1p3NINEsYGo
var url = "http://www.youtube.com/watch?v=JYATEN_TzhA&feature=featured
# Parses YouTube URLs directly or from iframe code. Handles:
# * Address bar on YouTube url (ex: http://www.youtube.com/watch?v=ZFELEVEN-OI)
# * Direct http://youtu.be/ url (ex: http://youtu.be/ZFTWELVEBOI)
# * Full iframe embed code (ex: <iframe src="http://www.youtube.com/embed/13_lHhCNBOI">)
# * Old <object> tag embed code (ex: <object><param name="movie" value="http://www.youtube.com/v/FOURTEEN_--">...)
http://www.youtube.com/user/Scobleizer#p/u/1/1p3vLASTYGo

This last regexp does not work with ruby:

> url = "http://www.youtube.com/watch?v=NLqASIXrVbY"
=> "http://www.youtube.com/watch?v=NLqASIXrVbY"
> vid_regex = /(?:youtube(?:-nocookie)?\.com\/(?:[^\/\n\s]+\/\S+\/|(?:v|e(?:mbed)?)\/|\S*?[?&]v=)|youtu\.be\/)([a-zA-Z0-9_-]{11})\W/
=> /(?:youtube(?:-nocookie)?\.com\/(?:[^\/\n\s]+\/\S+\/|(?:v|e(?:mbed)?)\/|\S*?[?&]v=)|youtu\.be\/)([a-zA-Z0-9_-]{11})\W/
> url =~ vid_regex
=> nil
> $1
=> nil

@toots it works for me removing the last \W

Same thing for php - must remove trailing \W. I also had to enclose it in slashes /.

Thanks, I use it in PHP:

function getYoutubeId($url)
{
    // original regex source: https://gist.github.com/afeld/1254889#gistcomment-1253992
    $regex = '/(?:youtube(?:-nocookie)?\.com\/(?:[^\/\n\s]+\/\S+\/|(?:v|e(?:mbed)?)\/|\S*?[?&]v=)|youtu\.be\/)([a-zA-Z0-9_-]{11})/mi';
    preg_match($regex, $url, $matches);
    return isset($matches[1]) ? $matches[1] : null;
}

$url = 'https://www.youtube.com/watch?v=SN102svZHQg';
echo getYoutubeId($url); // SN102svZHQg

private const MAGIC_REGEX = '#^(?:https?://|//)?(?:www\.|m\.|.+\.)?(?:youtu\.be/|youtube\.com/(?:embed/|v/|shorts/|feeds/api/videos/|watch\?v=|watch\?.+&v=))([\w-]{11})(?![\w-])#';

added support for shorts :)