How To Download YouTube Videos with PHP

[Update #1: In a next post you can find the full Youtube Download PHP scripts. However browsing through this article will help you understand the technique hidden in the source code. your choice!]

As everybody else I suppose, I have used some software but also sites to download YouTube videos. I used to made a wrapper software that used this techcrunch tool. There is a vast collection of tools available that could do the job as easily as this. However, there is some time now (1 or 2 months) that all these have...stopped working. You can see for yourself. Try searching for 'youtube video download tool' and roam on the results for a while. You will find out that the vast majority of these sites are basically 'out-of-service'. Perhaps because of legal issues.

Anyway, to hell with 'cloud computing', I said, and decided to write my own code for downloading videos from YouTube. You may have noticed a small iframe in the right bottom corner of the blog with the name 'DOL Gadgets' (under the categories section). It basically hosts some PHP code that runs in my machine and can provide download links for YouTube videos. Regardless, the language you use to write the code, one has to overcome a small problem. The initial YouTube video link does not contain explicitly a link where the video actually is hosted. So, there are some small tricks you have to employ.

First thing is to discover how your browser finds the content to play. As an example let's say you click on the Eric Prydz's Armed video with the URL being:
http://www.youtube.com/watch?v=T0XIMkuo81Q

In Firefox, I used the Live Http Headers add-on to make a little spying. Using this tool we can watch in real time the http communication that is going on while you are browsing (GET/POST requests, responses, headers etc) The first thing to notice is that the browser finally makes a request like this:
GET /get_video?video_id=T0XIMkuo81Q&t=OEgsToPDskJW2DY1pibC1Tu7D7tVOZrE
If you put it directly into the browser this will open a Save File Dialog for you to download it. Saving the file with an flv extension finishes our job. But to create a request like this?

There are 2 parameters in this: the video_id and t. The first is what you already know from the initial URL. The latter, which I tend to call it ticket, is not available but somehow the browser locates it (or the player). So the plan is clear: get the video_id from the initial URL and then try to find the ticket.

Let's begin from the harder one. Since there is no other alternative the ticket should be communicated in the initial request or some subsequent. Opening the source of the page that is returned from the initial video URL we can find that the ticket is actually there. To get the video_id we simply process the initial URL. Now we are done. The scenario is complete:

1. Get the page returned from the initial URL
2. Retrieve the ticket from this page
2. Retrieve the video_id from the initial URL
4. Create the GET request using the video_id and the ticket

This is exactly what the PHP code does. (Excuse me for the poor PHP syntax and coding, but I am far from being comfortable with the language) If $location has the initial URL then :

To execute step 1. we can use the following PHP snippet:
$handle = fopen($location, "r");
if($handle) { $contents = '';
while (!feof($handle)) { $contents .= fread($handle, 8192); }
fclose($handle);
With this, we read the stream to the variable $contents which now has the YouTube landing page.

To execute step 2.
$result1 = preg_match("/&t=([\w]*)&/",$contents,$tickets);
$ticket = $tickets[1];
The preg_match is a regular expression matching function in PHP. Basically we match a pattern that can extract the ticket against the YouTube page. It returns true when there is a match and the matched groups (if any) are returned in the $tickets. After this snippet the $ticket variable has the needed string to construct the GET request.

To execute step 3.
$result2 = preg_match("/v=(\w*)/",$location,$video_id);
$id = $video_id[1];
The same reasoning holds here. $id has the video id(first parameter)

To create the GET request and execute the last step we can use the php snippet
echo "<"; echo "a href = \"http://www.youtube.com/get_video?video_id="; echo $id; echo "&t="; echo $ticket; echo "\"Download link.";
This creates a hyperlink that when followed prompts to download the video file.

That's all! One networking function, two simple regular expressions and ...done, we have a YouTube video downloading on our PC. Now, it is easy to expand the service. In subsequent posts, I will try to give code for the Javascript version, but also other enhancements like entering the video name rather than the video UR etc. Till then, feel free to use the tool in the blog site or adjust the code to your needs.