Fallback for live streams when web scraping fails

[dagwieers]

This implements vualto live stream support (by video_id) for streams
that do not rely on web scraping, but also implements a fallback
mechanism in case web scraping fails.

This is the last piece to get rid of any reliance on web scraping.
Users everywhere rejoice !

This fixes #57

[dagwieers]

@mediaminister Please test!

I tested this by enabling (non-working) stubru and mnm livestreams. And it correctly falls back to using the provided video_id.

[dagwieers]

I am going to merge this, so I can test everything (including Radio support) in a single build.
Feedback and improvements welcome !

[mediaminister]

This code can't work. You can't return streamurls before we check roaming, etc...
I'll provide a pull request with a fix.

[dagwieers]

@mediaminister OK, this affects only the fallback for the original streams. We can fix this.

[mediaminister]

Maybe it affects more than that. I'm testing at the moment.

[dagwieers]

No, it worked fine (as a fallback) for the MNM and Studio Brussel streams, and only is used as a fallback for the original streams. So yes, we need to fix this, it definitely fails to fallback.

[mediaminister]

We never used web scraping for MNM and Studio Brussel, it never executed your fall back code.

[dagwieers]

Well, for testing I added the live URL that obviously fails to be scraped correctly.
That is how I tested where and how to inject it, but it is insufficient for the original streams.

[dagwieers]

I'll be back later this evening.

[mediaminister]

I found out what was wrong and created a fix: #170

[dagwieers]

I wondered what the _geo streams were, apparently they are the solution ;-)