Exit gracefully when switching to an audio group with no tracks

[robwalch]

This PR will...

Exit gracefully when switching to an audio group with no (supported) tracks

Why is this Pull Request needed?

Users report receiving a fatal error in streams with no audio tracks (but audio groups?)

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

Resolves issues:

Related to #4602

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

[robwalch]

cc @connected-salmouaswas @Conceptiks @TanakritBenz @lp-contentplus

[mtoczko]

Hi @robwalch
Were you able to reproduce this error? Because on my computer, the audio controller didn't return such an error. (#4602 )

This PR will fix the issue, but nonetheless, I'd like to investigate the cause independently because from the description, it seems to occur randomly.

I can repro on the demo page as well. It doesn't happen 100% of the time, but if you retry enough (like, re-clicking the "Apply" button) it will happen eventually. Demo link

[robwalch]

Hi @mtoczko,

I was not able to reproduce this issue.

I have seen similar failures where the track selection fallback was broken before this change was made:

https://github.com/video-dev/hls.js/pull/5562/files#diff-fce83d85205f6441ddd410c5aca54bf4215dcbe6c16b6184af98ae438f930626L212-R215

It is not clear how the audio track controller would get a level switch to have a level with audio group ids but no tracks - especially with an asset that has no audio groups. Maybe by (guesses):

1. Calling hls.audioTrack = N while playing an asset that has no tracks to select?
2. Reusing the player instance?
3. Manipulating level and track data?
4. Forcing a level switch in a player event loop before group tracks are reset?
5. Asset issue not captured in the issue thread - not caught by m3u8 parser?

Either way, I wanted to provide a safer exit path than what was being suggested as a workaround in #4602.

I changed the description to "Related to #4602", since this PR does not conclusively resolve #4602. This PR will prevent the escapes users were reporting later in the thread.