3 -> Implement a system to support an "on-demand" scan. The goal is
4 to retrieve in priority the metadata for a specific file.
5 -> Improve the log system to send the messages in a callback
6 instead of the stderr (if the callback is defined).
9 -> Add a new SQLite table to store some informations related to the
10 library and to the version of the db file.
11 -> Rewrite the whole structure of the database in order to handle
12 generic metadata and new filetypes (video, pictures, ...).
13 -> Add a public function to get the data for only one file.
16 -> Implement a system to add filters (to ignore some directories)
18 -> Rewrite a bit the scanner in order to support several scanners
19 in parallel (for the same valhalla object). It will be useful
20 when some paths are not on the same device because it can
21 increase a lot the speed.
22 It is already possible to do that but it needs to create a
23 valhalla object for each path.
29 -> Implement a structure to add grabbers in Valhalla. The idea is to
30 have 3 steps for a file.
32 #1 the file is parsed by FFmpeg and saved in the database.
33 #2 the file is sent to the grabbers to retrieve new informations
34 exactly like the grabbers of Enna (with the metadata retrieved
36 #3 the files (like images) retrieved by some grabbers are downloaded
37 and saved somewhere on the disk.
39 A first draft of this new architecture is visible at this link:
40 http://www.geexbox.org/~schroeterm/valhalla/new_valhalla.png
42 The goal is to move the grabbers of Enna in Valhalla and to fix
43 definitively the problem of concurrency. Of course, it must be easy