Parse tags that span cues

[kmahelona]

We need to be able to parse tags that span cues. I don't believe this is implemented based on the tests.

20
00:01:53.870 --> 00:01:59.164
He kōrero nui hoki tērā

21
00:01:59.164 --> 00:02:06.966
taua wāhi rā, nā ōku mātua, ā, <c.ingoatupuna>Ngārama Te Maru</c> me tana wahine <c.ingoatupuna>Matire

22
00:02:06.966 --> 00:02:12.353
Rapihana</c>, i tuku mai hei tūnga whare mō te kāinga

Here, <c.ingoatupuna>Matire Rapihana</c> spans cues 21 and 22.

[kmahelona]

I'd write a test but I'm not sure how to do that as it seems like the parsing converts the webvtt to a list of cues. Does a new line in the tests represent a joined list? E.g. could I write a test,

test_string = "this is a <c.tag>wonderful" /
              "tag</c> that spans a line break."

[gregplaysguitar]

Hmm. This is going to be pretty tricky because I'm using a 3rd party lib to parse the webvtt file (pyvtt) and that's what creates the list of cues. I'll have to give it some thought.

To test it, I think you could test the parse_transcript function, possibly passing a file-like object containing the sample webvtt cues rather than an actual file reference? I already do something like that here https://github.com/TeHikuMedia/django-colloquial/blob/master/colloquial/colloquialisms/tests/test_parser.py#L224

[gregplaysguitar]

This should be fixed now in 4356c77