[FFmpeg-devel] [PATCH] reduce score returned by FLIC format

[Michael Niedermayer]

On Wed, Mar 26, 2008 at 07:14:52AM -0700, Mike Melanson wrote:
> Mike Melanson wrote:
> > Aurelien Jacobs wrote:
> >>> The first 4 bytes of the file should match the total size of the file.
> >> But you can't check this unless the stream is seekable.
> > 
> > And it also doesn't work for the bastardized FLIC variant found on a
> > particular game. There are maybe 7 files in existence that depend on
> > that hack. But one of them is a FATE test, so we can't break it. :)
> > 
> >>> The width and height could also be sanity-checked (this is part of
> >>> how the id CIN demuxer detects a file).
> >> This might be enough...
> >>
> >> Still this test would be a bit week, and maybe it would be better to
> >> make it return 9*AVPROBE_SCORE_MAX/10 or something like that.
> > 
> > I'll take that into account.
> 
> Here's a patch. Sane enough? I declared that each dimension must be
> between 4..2048. That should be generous enough. And I also used Aurel's
> suggestion (90% sure).

didnt you say =0 was occiring in some files? If so you should allow this
too
Also you could check the "speed" variable and the chunk size of the first
chunk (megabyte or larger being a little unlikely ...)

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Thouse who are best at talking, realize last or never when they are wrong.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20080326/bbac53d7/attachment.pgp>