> so we would need to spend some time in discussing the api, I think.
OK, even when it is an extra project, we should discuss the api here. I
was thinking about:
1. Get all the info possible. This includes
I was thinking to build a structure called table that is
mostly a list of key-value pairs and a type id. An
EXIF table would i.e. include exif keys and the corresponding
values. The keys themselves should remain the ones used in
the file and be translated to a human readable format by
gettext so they can be i18n-ed. From all tables that were
build by the parser, a Metadata-Object is build. I also
like the idea of a generic object that has some generic
data like title, artist, description, keywords. This way
we can access id3 and ogg metadata from one interface
without caring whats really inside the audio file.
CDDB and perhaps VCD, DVD are a little special here since
they generate information about more than one file. We
need to work something out for that. I must admid that I
don't have a keen idea, yet. :-(
2. Caching the data
- Cache data with pickle to disc. This could be done with one file
for each directory. Discs could get an extra cache file, some for
CDDB infos. Inserting a DVD again could give you all the infos by
just loading the media-id.
- Rememeber mtime for files and disc id for discs to check if you
have to reread the file or take the cache. Make it possible to
set ROM drives to find the cache entry when you insert a mp3 disc
in a different drive with a different mount point
- A script to cache a directory. You could run that for all Freevo
dirs in a cron to speed up Freevo
At the moment mmpython does not do caching of any kind.
> So far I can parse avi, ogg, mp3, jpg, tiff without external libs. My
> personal main focus is on pictures and video but a general package is
> what I really would like to have.
I would love to use that lib. I'm working on the DVD parser right now.
But it depends on libdvdread, I can't think of a way without it.
No problem. I just made all this in python because I found
it interesting to understand the file formats and build the
parsers. Well ... For some formats the libs really are a
big overhead. OGG libs are especially huge if you just want
to extract some metadata from this very simple format.
Can we include the dvd parser lib in the cvs or where can
we get it.
Small bugreport: your lib always reports 2 audiochannels. Do you mean 2
as in stereo or two as in two audio streams to select from? And for mpg
you lib report 4096x4096 as widthxheight -- nope, that's wrong.
Yep... Still a lot of work to do. I think 2 audiochannels
means stereo but I just hacked down the parser from the
docs I found so 2 may also be some sort of define for
stereo. Will investigate further. As you already found out
mpeg parsing is far from being perfect...
l8r...
Thomas
-------------------------------------------------------
This SF.net email is sponsored by: Etnus, makers of TotalView, The best
thread debugger on the planet. Designed with thread debugging features
you've never dreamed of, try TotalView 6 free at www.etnus.com.
_______________________________________________
Freevo-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/freevo-devel
