This functionality is intended to eventually encompass the current format filtering.
Reported at https://bugzilla.redhat.com/show_bug.cgi?id=1093517 .
For now, we just wrap some common errors. More may follow. We do not want to catch actual programming errors in the extractors, such as 1 // 0.
Also, update the testcase, detect private/country restricted videos and modernize a bit.
Nobody wants to know what the worst possible format is. And if they do, they can still provide -f worst.
This makes nationalarchives.gov.uk work (Fixes #4907, fixes #4876)
XPath is used for extracting the video url and the thumbnail
Was supposed to be deleted by 67c2bcd
67c2bcd
Todo: URL and Thumbnail should be extracted with XPath
The test now doesn't fail anymore. It just checks the string for having "vimeo" in it, instead of using the method for URL-checking, since it's returns an error. The tests don't fail, and the extractor works fine now.