GitLab

Check for both automatic captions and subtitles with options `--write-sub` and `--write-auto-sub` (fixes #1224)

[youtube] Support automatic captions with original language different from English (fixes #1225) and download in multiple languages.

It takes too much to follow the redirection.

It raises ExtractorError for the same exceptions we have to catch.

[subtitles] Simplify the extraction of subtitles in subclasses and remove NoAutoSubtitlesInfoExtractor

Subclasses just need to call the method extract_subtitles, which will call _extract_subtitles and _request_automatic_caption
Now the default implementation of _request_automatic_caption returns {}.

Otherwise it can be automatically detected as a IE ready for use.

It uses the same system as Pluzz, create a base class for both extractors.

Wrong property name

It printed for the minutes the result of (seconds // 60)

In urls like http://www.youtube.com/watch?v=BaW_jenozKcsharePLED17F32AD9753930 you can't split the query string and ids always have that length.

They have undo the change

This reverts commit 0638ad99.

As mentioned in the pull request, I forgot to include this changes.
https://github.com/rg3/youtube-dl/commit/aa6a10c44a8e2e86f709c5301f9ea6ac3f01f002

They were being caught by YoutubeUserIE, now it only extracts a url if the rest of extractors aren't suitable.
Now the url tests check that the urls can only be extracted with an specific extractor.

Urls like https://www.youtube.com/feed/watch_later were being as users (before the last changes to YoutubeUserIE, as videos)