Fix for Bug #877 Extended cue breakage in vtt files #1236
Conversation
|
Thanks for the PR! What exactly do you mean when you say the tests no longer apply? The point of those tests was that anyone submitting a fix for #877 wouldn't need to write them, so those definitely shouldn't be removed. If the PR is ready, those should be uncommented and passing. |
|
Hello! I removed them because my change was to prevent parseCueTime from receiving extra data in the date string. Right now those tests would still fail, though thinking on it, perhaps parseCueTime should still sanity check it's string before proceeding. |
|
I think this is a fine way to handle it. At some point we need to actually store the extra flags that would come after the end time, but at the moment we don't do anything with them. And either way it's probably better to sanitize the end time before passing it to parseCueTime. Any chance you want to squash #183 at the same time? And maybe write a test to test this new approach? |
one or more tabs and/or spaces fixes videojs#183 Uncommented tests, added new ones for multi and mixed whitespace cases
|
Sure! instead of splitting on just one space, now split on one or more tab or space (which appears to be covered in the spec). I added a test for multiple tabs / spaces and can add more if you can think of something that doesn't have coverage. Nothing was immediately popping out though, at least within what it can currently handle. |
|
Thanks again for this! Tests look good to me. I just merged it in and it will go out with v4.7.0. |
Only send actual time components to parseCueTime (no extended info). I also some commented out tests since they no longer apply.