This commit moved the HLS parsing task to a separate method, did little performance improvements and used final where possible in the SoundcloudStreamExtractor file.
If a progressive stream is present in the transcodings, it's unnecessary to show twice an MP3 128kbps stream so if this is the case, the MP3 HLS stream will be not added to the audioStreams, else it will.
This commit also catch fetching errors in HLS manifests parsing and don't add the corresponding stream if an error occurs.
This commit tries to support SoundCloud HLS streams by parsing M3U manifests, get the last segment URL (in order to get track length) and request a segment URL equals to track's duration so it's a single URL.
* Test for bandcamp footer instead of meta tag (which is not present on
all pages)
* Accept links to /music, not just /releases
* Correctly handle uppercase URLs
The ContentNotSupportedException is thrown because no supported audio streams where extracted. However, SoundCLoud does not check, whether there are any streams available.
This commit should be reverted in #526
These exceptions are thrown on a test with the error messages text, because YouTube returns only "UNPLAYABLE" status in most error cases.
Tests are based with English strings, so changing the lang used by
extractor will throw the generic exception (ContentNotAvailableException).
This commit adds two new exceptions in NewPipe Extractor: GeographicRestrictionException and SoundCloudGoPlusException (which extend to ContentNotAvailableException). These exceptions allow showing different error messages to user when a content isn't available in his/her/its country (only used for now by SoundCloudStreamExtractor) or when the content is a SoundCloud Go+ track.
- do not parse responseBody twice for continuation
instead try to get commentsTokenInside with the new pattern ("sectionListRenderer")
and try again with the old pattern ("commentSectionRenderer") on failure
- do not unescape responseBody multiple times
-> parse responseBody less times
SoundCloud changed their Javascript structure. That caused the client_id to be in the middle of the file (at byte ~43000 ). To have a little buffer, we now fetch the first 50000 bytes.
Requires sending a POST request instead of GET.
clientName and clientVersion, which were required as headers previously now need to be part of the request payload.
continuation id also needs to be part of request body.
quick and dirty solution.
