They bring nothing of value (we never use import * anyway) and introduce
a maintenance burden.
The __all__ that I left might not be very useful either but at least
they are not a pain to maintain; they are also used by pyflakes in some
cases.
* comment out or remove some debug prints
* try to cope with non-UTF-8 title passed (1013483).
Couldn't get it to be recoded yet.
* handle boggus url to xml config file (http:/...) (1015020)
* handle iframes with page name different than 'player.html' (1013798)
* handle xml config filenames with spaces (1441)
* catch xml config names with 'smil' as hostname, which means speakervideo
points to a smil file describing the streams, and take the file with the
highest bitrate
* account for xml config with only valid slidesvideo (1016627) and use it
as fallback
* handle configs with 'Invalid Date' as date text (1016634)
We can now dump the entire gdcvault video files urls with a few exceptions
(403 HTTP errors), and many /mediaProxy.php urls failing to redirect, which
are still to be investigated.
* Implement search functionality, using POST to get JSON data
* Fix download for most items even for non-free ones.
For now only missing are non-free mp3 files it seems (like 769),
/mediaProxy.php returns 2bytes html crap.
For now it only fetches the speaker video, but each page can have
both a speaker and slides video feed.
TODO: search
Signed-off-by: François Revol <revol@free.fr>
Signed-off-by: Romain Bignon <romain@symlink.me>
