getting the youtube id from a link

Question

I got this code to get the youtube id from the links like www.youtube.com/watch?v=xxxxxxx

  URL youtubeURL = new URL(link);
  youtubeURL.getQuery();

basically this will get me the id easily v=xxxxxxxx

but I noticed sometime youtube links will be like this

http://gdata.youtube.com/feeds/api/videos/xxxxxx

I am getting the links from a feed so do I need to build a regex for that or theres a parser to get that for me ?

Solution 1

Tried the other ones but failed in my case - adjusted the regex to fit for my urls

String pattern = "(?<=watch\\?v=|/videos/|embed\\/)[^#\\&\\?]*";
    
    Pattern compiledPattern = Pattern.compile(pattern);
    Matcher matcher = compiledPattern.matcher(url);

    if(matcher.find()){
        return matcher.group();
    }

This one works for: (you could also implement a security check youtubeid length = 11 )

http://www.youtube.com/embed/Woq5iX9XQhA?html5=1

http://www.youtube.com/watch?v=384IUU43bfQ

http://gdata.youtube.com/feeds/api/videos/xTmi7zzUa-M&whatever

Woq5iX9XQhA

384IUU43bfQ

xTmi7zzUa-M

Solution 2

public static String getYoutubeVideoId(String youtubeUrl)
 {
 String video_id="";
  if (youtubeUrl != null && youtubeUrl.trim().length() > 0 && youtubeUrl.startsWith("http"))
 {

String expression = "^.*((youtu.be"+ "\\/)" + "|(v\\/)|(\\/u\\/w\\/)|(embed\\/)|(watch\\?))\\??v?=?([^#\\&\\?]*).*"; // var regExp = /^.*((youtu.be\/)|(v\/)|(\/u\/\w\/)|(embed\/)|(watch\?))\??v?=?([^#\&\?]*).*/;
 CharSequence input = youtubeUrl;
 Pattern pattern = Pattern.compile(expression,Pattern.CASE_INSENSITIVE);
 Matcher matcher = pattern.matcher(input);
 if (matcher.matches())
 {
String groupIndex1 = matcher.group(7);
 if(groupIndex1!=null && groupIndex1.length()==11)
 video_id = groupIndex1;
 }
 }
 return video_id;
 }

Solution 3

This regex would do the trick:

(?<=videos\/|v=)([\w-]+)

This means that we're first looking for video/ or v= then captures all the following characters that can be in word (letters, digits, and underscores) and hyphens.

Example in java:

public static void main(String[] args) {

    String link = "http://gdata.youtube.com/feeds/api/videos/xTmi7zzUa-M&whatever";
    String pattern = "(?:videos\\/|v=)([\\w-]+)";

    Pattern compiledPattern = Pattern.compile(pattern);
    Matcher matcher = compiledPattern.matcher(link);

    if(matcher.find()){
        System.out.println(matcher.group());
    }
}

Output:

xTmi7zzUa-M

Solution 4

Got a better solution from this link.

Use the following method to get the videoId from the link.

YoutubeHelper.java

import com.google.inject.Singleton; 

import java.util.regex.Matcher;
import java.util.regex.Pattern;

@Singleton 
public class YouTubeHelper { 

    final String youTubeUrlRegEx = "^(https?)?(://)?(www.)?(m.)?((youtube.com)|(youtu.be))/";
    final String[] videoIdRegex = { "\\?vi?=([^&]*)","watch\\?.*v=([^&]*)", "(?:embed|vi?)/([^/?]*)", "^([A-Za-z0-9\\-]*)"};

    public String extractVideoIdFromUrl(String url) {
        String youTubeLinkWithoutProtocolAndDomain = youTubeLinkWithoutProtocolAndDomain(url);

        for(String regex : videoIdRegex) {
            Pattern compiledPattern = Pattern.compile(regex);
            Matcher matcher = compiledPattern.matcher(youTubeLinkWithoutProtocolAndDomain);

            if(matcher.find()){
                return matcher.group(1);
            } 
        } 

        return null; 
    } 

    private String youTubeLinkWithoutProtocolAndDomain(String url) {
        Pattern compiledPattern = Pattern.compile(youTubeUrlRegEx);
        Matcher matcher = compiledPattern.matcher(url);

        if(matcher.find()){
            return url.replace(matcher.group(), "");
        } 
        return url;
    } 
} 

Hope this helps.

Solution 5

That pattern worked for me:

"http(?:s?)://(?:www\.)?youtu(?:be\.com/watch\?v=|\.be/)([\w\-]+)(&(amp;)?[\w\?=‌​]*)?"

source: Regular expression for youtube links

View more solutions

Comments

[ridgerunner]

You may want to look at my answer to a very similar question. It extracts the video-id from a variety of YouTube URL formats.

[Peril]

@ridgerunner thanks but it misses the gdata links

[ridgerunner]

Thanks for pointing that out. I've updatated my YouTube ID matching expression so that it now correctly matches your gdata subdomain example.

[Peril]

this will not get me the id only

[Marcus]

Using http://gdata.youtube.com/feeds/api/videos/xxxxxx as indata I get xTmi7zzUa-M as output. Did I misread your question and you were asking for a regex that would allow you to parse both v=xxxxxxxx and the other one?

[Marcus]

Well actually. That regex works in both cases. What result do you get?

[Peril]

but the regex you wrote will get me the id and other stuff check here gskinner.com/RegExr

[Derek Springer]

It's the '$' at the end that makes it only capture everything beyond .../videos/. It wouldn't work in the first example if there if there were any other parameters on the link (i.e. v=xxxxxxx&other_param=yyyyy...)

[Marcus]

Salah have a look at my updated answer now. @Derek you're right about that, though that wasn't specified in the question I didn't take that into account.

[Code Jockey]

A lookbehind would be better in certain cases than a non-capturing group. If used instead, the entire match (that is matcher.group()) would be the id the asker is looking for, rather than the first capturing group (matcher.group(1))

[Marcus]

@Jockey you're right about that. I replaced my non capturing group with a positive lookbehind. Ty

[Pkmmte]

Thank you soooooooooooooooo much!!

[kritzikratzi]

it's not pretty, but here you go: (?<=watch\?v=|/videos/|embed\/|youtu.be\/|\/v\/|watch\?v%3D|‌​%2Fvideos%2F|embed%2‌​F|youtu.be%2F|%2Fv%2‌​F)[^#\&\?\n]*

[Code Jockey]

@krishan indeed not (that wasn't even a valid URL when this question was asked, to my knowledge...) but if you're interested, this should work for the youtu.be as well as /embed/ links: (?<=watch\?v=|/videos/|/embed/|youtu.be/)[^&#?]* (I'll update my answer) - please let me know if it doesn't, as I made that up just now by hand...

[Er KK Chopra]

i got a solution about this my question stackoverflow.com/questions/25718304/…

[JPM]

@Peril This should be the answer please mark it so.