[haiku] Re: Need Some GSoC Advice
- From: "François Revol" <revol@xxxxxxx>
- To: haiku@xxxxxxxxxxxxx
- Date: Fri, 27 Mar 2009 14:45:29 +0100 CET
> Ryan Leavengood <leavengood@xxxxxxxxx> wrote:
> > On Fri, Mar 27, 2009 at 2:14 AM, Ankur Sethi <
> > get.me.ankur@xxxxxxxxx>
> > wrote:
> > > I have decided to go with a two pronged approach, the same one
> > > that
> > > the SkyOS Index Feeder uses. A database for full-text indexing,
> > > and
> > > filesystem attributes for files such as MP3s. I submitted my
> > > proposal
> > > just a few minutes ago. Thanks to everyone who took an interest
> > > :)
> > From what I have read in this very long thread, mimicking the SkyOS
> > Index Feeder is probably a good idea.
>
> Yes, definitely.
We can still come up with a better idea using BFSv2 for R2 anyway :)
>
> > Also you may recall earlier in this thread I mentioned Army Knife
> > in
> > relation to your media player idea. Well I think it might also
> > prove
> > useful for mp3 attribute extraction for this project. In other
> > words
> > maybe code from Army Knife could be usable in the mp3 search
> > plugin.
>
> Army Knife is using TagLib (http://developer.kde.org/~wheeler/taglib.html)
> > for this. I also have ID3v2 code that I have written some time ago
> that
> could be put under the MIT if the need arises.
IIRC many BeOS apps fail at reading recent ID3 (some also fail at
reading the mp3 itself due to the new ID3 stuff...)
Also interesting is the OGG metadata stuff.
ffmpeg also recently gained a new metadata API for demuxers, it could
be intersting to export it from the BMediaFile class.
I'm just writing something on the BaseTranslator class to see how it
could reuse translators for this. I'll be posting a diff later on.
François.
Other related posts:
