Automatic annotation of video using external collections

[xavierfav]

Currently we rely on the subliminal parser to extract informations about the videos. It allows us to identify the type of content (movie or series), and extracts the episode and season number for series. This is purely based on the name of the video file. Sometimes, the parser makes mistakes and is not always robust to any type of character in the filenames. It would be nice to have a better way to extract the video informations for instance by using external collections such as IMDb.

In the code, get_video_type_and_info() takes care of that. The idea would be to improve it by using library for retrieving and managing the data of the IMDb database. IMDbPY seems like a good candidate to start with. We could store additional information such as the genre of the movies, and store it in our database, for improving our search engine.