[FFmpeg-devel] [PATCH] Use file protocol to deal with bogus protocol strings

[Michael Niedermayer]

On Sat, Sep 29, 2007 at 05:08:56PM +0200, Stefano Sabatini wrote:
> On date Friday 2007-09-28 12:04:06 +0200, Stefano Sabatini encoded:
> > On date Friday 2007-09-28 10:48:26 +0200, Reimar D?ffinger encoded:
> > > Hello,
> > > On Thu, Sep 27, 2007 at 11:13:32PM +0200, Luca Barbato wrote:
> > > > Stefano Sabatini wrote:
> > > > > Another (maybe cleaner) solution would be to request the user to
> > > > > specify the protocol string for ambiguos files such as foo:movie.mpeg,
> > > > > which should then be specified as file:foo:movie.mpeg. Anyway in this
> > > > > case ffmpeg should say something if it can't recognize the protocol
> > > > > (for example: unrecognized protocol: foo), rather than simply negate
> > > > > the existence of the file.
> > > > 
> > > > what about enforcing the "://" separator?
> > > 
> > > I must somehow have missed the whole issue... Why wouldn't someone who
> > > want to use such filenames specify them as
> > > file://file...
> > > for relative and
> > > file:///path...
> > > for absolute path like it is done almost everywhere else?
> > > Or is that not supported by ffmpeg yet?
> > 
> > Hi, 
> > 
> > no, actually in libavutil/file.c there is:
> > 
> > static int file_open(URLContext *h, const char *filename, int flags)
> > {
> >     int access;
> >     int fd;
> > 
> >     av_strstart(filename, "file:", &filename);
> > ...
> > 
> > so file://foo.avi
> > 
> > doesn't make ffmpeg use the file protocol, while
> > file:foo.avi
> > 
> > does it.
> > 
> > With the attached patch I'm changing this behaviour, favouring the
> > file://foo.avi approach.
> > 
> > Note that in this way a file such as:
> > foo:movie.avi
> > 
> > is still interpreted using the file protocol (while foo://movie.avi fails).
> [...]
> 
> I'm reattaching the patch: yes it should be splitted, at first
> introducing the av_find_protocol_by_name function.
> 
> If this approach is acceptable I can proceed with the first one.

rejected, leave the code as it is!

if you think theres a bug FIRST explain VERY precissely what you think
is buggy dont just randomly complicate the code and break existing
syntax!

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

The greatest way to live with honor in this world is to be what we pretend
to be. -- Socrates
-------------- 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/20070929/7740775f/attachment.pgp>