Fix MPEG-2 TS packet scanning regression

[robwalch]

This PR will...

Restore TS demuxer code that scans for Transport Stream packets.

Why is this Pull Request needed?

Users reporting issues with streaming containing TS segments that do not start with mp2t packet sync byte. These bytes should be skipped.

Are there any points in the code the reviewer needs to double check?

Resolves issues:

Fixes #4922 #5002
Address comment #5001 (comment)

Checklist

• changes have been done against master branch, and PR does not conflict
• new unit / functional tests have been added (whenever applicable)
• API or design changes are documented in API.md

[philcluff]

I think this is the right change to make given the reports of this change causing issues, as much as I don't like supporting clearly malformed streams.

I think we should add a specific test stream for this honestly, if it's something that we're committing to supporting long term.

Would we be open to being more strict on this under a major version revision do you think?

[itsjamie]

The general rule @philcluff that seems to be the guiding principle for "should we do this", is if it works in the <video> element natively on Safari.

A user shouldn't have content that functions playing it natively, go to a browser that uses MSE with hls.js, and have it fail.

I'm all for next major verison if it fails in Safari, we can fail.