[FFmpeg-devel] [PATCH] Private options for url protocols

[Michael Niedermayer]

On Tue, Nov 15, 2011 at 01:42:46AM +0100, Stefano Sabatini wrote:
> On date Thursday 2011-11-10 17:56:11 +0100, Michael Niedermayer encoded:
> > Hi
> > 
> > Attached patchset implemets subj
> > Comments welcome, ill probably push it tomorrow if there are no
> > objections
> > 
> > An example of how useage would look is:
> > ffplay http,user-agent=JustTesting://samples.multimedia.cx/MPEG2/test422.m2v
> > 
> > the syntax would also work for nested protocols ...
> > 
> > -- 
> > Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
> > 
> > No human being will ever know the Truth, for even if they happen to say it
> > by chance, they would not even known they had done so. -- Xenophanes
> 
> > From d300a9c6a787827dbe2d86a1b330b8aba830bcc7 Mon Sep 17 00:00:00 2001
> > From: Michael Niedermayer <michaelni at gmx.at>
> > Date: Thu, 10 Nov 2011 17:02:02 +0100
> > Subject: [PATCH 1/2] avio: allow any chars in protocols
> > 
> > Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
> > ---
> >  libavformat/avio.c |    6 +++---
> >  1 files changed, 3 insertions(+), 3 deletions(-)
> > 
> > diff --git a/libavformat/avio.c b/libavformat/avio.c
> > index 47b2d51..f08cbab 100644
> > --- a/libavformat/avio.c
> > +++ b/libavformat/avio.c
> > @@ -221,17 +221,17 @@ int ffurl_alloc(URLContext **puc, const char *filename, int flags)
> >  {
> >      URLProtocol *up;
> >      char proto_str[128], proto_nested[128], *ptr;
> > -    size_t proto_len = strspn(filename, URL_SCHEME_CHARS);
> > +    const char *proto_end = strchr(filename, ':');
> 
> uhmm... what if you have ":" in the options? e.g.:
> ffplay http,user-agent=::the_bloody_user_agent::://samples.multimedia.cx/MPEG2/test422.m2v
> 
> I still find the lack of defined namespaces for the options a big
> issue, but I can't say what would be a good way of setting this,
> maybe:
> -proto_opts=user_agent=... http://...
> 
> yes that would be inconsistent...

also theres the problem that protocols can be nested like
applehttp+http:...
we should be able to transfer options specifically to one


> 
> >      if (!first_protocol) {
> >          av_log(NULL, AV_LOG_WARNING, "No URL Protocols are registered. "
> >                                       "Missing call to av_register_all()?\n");
> >      }
> >  
> > -    if (filename[proto_len] != ':' || is_dos_path(filename))
> > +    if (!proto_end || is_dos_path(filename))
> >          strcpy(proto_str, "file");
> >      else
> > -        av_strlcpy(proto_str, filename, FFMIN(proto_len+1, sizeof(proto_str)));
> > +        av_strlcpy(proto_str, filename, FFMIN(proto_end-filename+1, sizeof(proto_str)));
> >  
> >      av_strlcpy(proto_nested, proto_str, sizeof(proto_nested));
> >      if ((ptr = strchr(proto_nested, '+')))
> > -- 
> > 1.7.4.1
> > 
> 
> > From fbee6fa7eb68c87f5df51849e046904c9113a23d Mon Sep 17 00:00:00 2001
> > From: Michael Niedermayer <michaelni at gmx.at>
> > Date: Thu, 10 Nov 2011 04:34:35 +0100
> > Subject: [PATCH 2/2] avio: Support private options in URLProtocols
> > 
> > Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
> > ---
> >  libavformat/avio.c |   10 ++++++++++
> >  1 files changed, 10 insertions(+), 0 deletions(-)
> > 
> > diff --git a/libavformat/avio.c b/libavformat/avio.c
> > index f08cbab..bb7660a 100644
> > --- a/libavformat/avio.c
> > +++ b/libavformat/avio.c
> > @@ -108,8 +108,16 @@ static int url_alloc_for_protocol (URLContext **puc, struct URLProtocol *up,
> >      if (up->priv_data_size) {
> >          uc->priv_data = av_mallocz(up->priv_data_size);
> >          if (up->priv_data_class) {
> > +            char *p, *p2;
> >              *(const AVClass**)uc->priv_data = up->priv_data_class;
> >              av_opt_set_defaults(uc->priv_data);
> > +            if((p2=uc->filename+strcspn(uc->filename, ":+")) && *p2 && (p=strchr(uc->filename, ',')) && p<p2){
> > +                int term= *p2;
> > +                *p2++= 0;
> > +                av_set_options_string(uc->priv_data, p+1, "=", ",");
> 
> maybe you should check&abort for errors here

locally fixed

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

Asymptotically faster algorithms should always be preferred if you have
asymptotical amounts of data
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20111115/7d0c6bff/attachment.asc>