VTT regions are not correctly implemented

[joeyparrish]

Have you read the FAQ and checked for duplicate open issues?
Yes

What version of Shaka Player are you using?
master

Can you reproduce the issue with our latest release version?
yes

Can you reproduce the issue with the latest code from master?
yes

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

What browser and OS are you using?
ChromeOS 81 stable

What are the manifest and license server URIs?
Don't have test content for this yet

Looking closely at cue rendering, we use viewportAnchor{X,Y}, but not regionAnchor{X,Y}. It seems that regionAnchor attributes are only set for VTT, not TTML, but our renderer doesn't use them ever.

The VTT spec says that viewportAnchor is the position in the viewport where the region anchor is attached, and that regionAnchor is the position in the region where the region anchor attaches to the viewportAnchor.

Meanwhile, in TTML, we assign the location of the region's origin to viewportAnchor.

So we're not using all the VTT information, and we're using the attributes to mean something slightly different for TTML.

[gkatsev]

WebVTT regions are more to create 608-like sliding captions rather than provide areas to display caption like in TTML.
I've recently mostly gotten regions working in the Video.js fork of vtt.js videojs/vtt.js#50

[dk00]

The parser is also needed to be updated,

It was implemented for Region: id=fred width=50% lines=3 regionanchor=0%,100% viewportanchor=10%,90% scroll=up,
but now regions looks like:

REGION
id:fred
width:40%
lines:3
regionanchor:0%,100%
viewportanchor:10%,90%
scroll:up

[avelad]

To clarify, the parsing is working correctly, and NativeTextDisplayer is doing it correctly; and UITextDisplayer handles regionAnchor{X,Y}, so I think this issue should already be resolved... @joeyparrish can you review it? Thanks!