feat: expose CEA708 window position in the cue's region (#5924)

CEA708 captions have positioning data available in their windows.
However, this isn't currently translated and exposed by shaka though it
is parsed from the bitstream.

Translates the windows into WebVTT regions and uses the mappings
outlined
https://dvcs.w3.org/hg/text-tracks/raw-file/default/608toVTT/608toVTT.html#positioning-in-cea-708

This is also partially implements #2583.

test/cea/cea_decoder_unit.js

@@ -565,7 +565,8 @@ describe('CeaDecoder', () => {
       decoder.extract(hideWindow, endTime);
 
       const text = 'test';
-      const topLevelCue = new shaka.text.Cue(startTime, endTime, '');
+      const topLevelCue = CeaUtils.createWindowedCue(startTime, endTime, '',
+          serviceNumber, 0, 11, 11);
       topLevelCue.nestedCues = [
         CeaUtils.createDefaultCue(startTime, endTime, /* payload= */ text),
       ];