HLS Parser Handling Aborted Operations as "No Partial HLS Segment Support"

[vaage]

This is an off-shoot of #1785.

What version of Shaka Player are you using?
2.5.0-beta2

Can you reproduce the issue with our latest release version?
Reproduced with 2.5.0-beta2

Can you reproduce the issue with the latest code from master?
Reproduced with 2.5.0-beta2

Are you using the demo app or your own custom app?
Custom App

If custom app, can you reproduce the issue using our demo app?
The current demo app does not have the mechanism to change streams fast enough to easily reproduce the issue.

What browser and OS are you using?
Chrome 71.0.3578.98 on OSX

For embedded devices (smart TVs, etc.), what model and firmware version are you using?

What are the manifest and license server URIs?
Custom manifest. However this can be reproduced with any HLS content.

What did you do?

1. Load HLS manifest (manifest A)
2. Load HLS manifest (manifest B)
3. Repeat steps 1 and 2 multiple times.

In order for this to work, there needs to be some delay between calls to load.

What did you expect to happen?
Calls to load should interrupt previous calls to load with no errors and the last call to load should produce playback.

What actually happened?
When the HLS Parser is shutting down and aborting its in-progress networking requests, it will misinterrupt the aborted operation as "Unable to fetch a partial HLS segment! Falling back to a full segment request, which is expensive!"

[vaage]

After talking it over with @joeyparrish the current plan of action is:

For 2.5:

1. Create a unit test for the HLS parser to show this behaviour. This test will be used as a regression test once the fix is created.
2. Implement a short-term (less invasive change) that will allow the test mentioned above to pass.

For 2.7

1. As part of the greater HLS changes, this fix should be built into the design of the parser in a more sustainable way than was done for 2.5.

[vaage]

Updates on this issue:

• A test for reproducing the issue has been created. The test is in code review.
• A fix for the issue has been created. The fix is in code review.

Many members of the team have been out-of-office lately, so reviews have been taking longer than normal. Thank you to everyone for their patience.

[vaage]

We have completed all the work for 2.5. The only work remaining is the work listed for 2.7. I am going to remove this from the 2.5 milestone so that it does not block the release. When we scope out 2.7 it should be added to that milestone.

@joeyparrish I am tagging you here since this has to do with release planning for 2.5 and 2.7.