Microsoft Teams generates .VTT transcript files for all recorded meetings. Unfortunately, it appears those files are not spec-conforming (shocker!), because they don't 0-pad any of the fields to their respective sizes (3 for milliseconds, 2 for everything else).
Pragmatically speaking, it would help for webvtt-py to support this, since it doesn't harm its ability to correctly and safely parse conforming ones. This patch does that, and adds a testcase for a representative Teams VTT.
Microsoft Teams generates .VTT transcript files for all recorded meetings. Unfortunately, it appears those files are not spec-conforming (shocker!), because they don't 0-pad any of the fields to their respective sizes (3 for milliseconds, 2 for everything else).
Pragmatically speaking, it would help for webvtt-py to support this, since it doesn't harm its ability to correctly and safely parse conforming ones. This patch does that, and adds a testcase for a representative Teams VTT.
Closes #44 among many others, I'm sure.