[FFmpeg-devel] [PATCH] Put the list of the regression tests in dedicated files

Stefano Sabatini stefano.sabatini-lala
Tue Jan 12 01:22:21 CET 2010


On date Saturday 2009-12-26 13:02:25 +0100, Stefano Sabatini encoded:
> On date Saturday 2009-12-26 12:41:32 +0100, Diego Biurrun encoded:
> > On Sat, Dec 26, 2009 at 12:52:25AM +0100, Stefano Sabatini wrote:
> > > Hi, as in subject.
> > 
> > I like the general idea.
> > 
> > > --- ffmpeg.orig/Makefile	2009-12-26 00:23:05.000000000 +0100
> > > +++ ffmpeg/Makefile	2009-12-26 00:42:14.000000000 +0100
> > > @@ -152,99 +152,9 @@
> > > +CODEC_TESTS = $(addprefix regtest-, $(shell cat $(SRC_PATH)/tests/codec.regression.list | grep -v '^\#'))
> > > +LAVF_TESTS  = $(addprefix regtest-, $(shell cat $(SRC_PATH)/tests/lavf.regression.list  | grep -v '^\#'))
> > > +# LAVFI_TESTS = $(addprefix regtest-, $(shell cat $(SRC_PATH)/tests/lavfi.regression.list | grep -v '^\#'))
> > 
> > Are any tests commented out?  If not, then the grep could be left out.
> 
> That was meant to make easier to comment/decomment things on the fly.
> 
> > Also, the transformation could possibly be refactored into a common
> > function.
> 
> Yes, I'd say to do this in a separate step.
> 
> > > --- ffmpeg.orig/tests/codec-regression.sh	2009-12-26 00:23:05.000000000 +0100
> > > +++ ffmpeg/tests/codec-regression.sh	2009-12-26 00:40:59.000000000 +0100
> > > @@ -3,7 +3,7 @@
> > >  # automatic regression test for ffmpeg
> > >  #
> > >  #
> > > -set -x
> > > +#set -x
> > 
> > Unrelated?
> 
> Yes.
>  
> > > --- /dev/null	1970-01-01 00:00:00.000000000 +0000
> > > +++ ffmpeg/tests/codec.regression.list	2009-12-26 00:36:53.000000000 +0100
> > > @@ -0,0 +1,49 @@
> > > +mpeg
> > > +mpeg2
> > > +mpeg2thread
> > > +msmpeg4v2
> > > +msmpeg4
> > > +wmv1
> > > +wmv2
> > > +h261
> > > +h263
> > > +h263p
> > > +mpeg4
> > > +huffyuv
> > > +rc
> > 
> > IMO it's high time these are sorted into alphabetical order.
> 
> With the "modular test system patch" order is not anymore relevant, so
> a simple sort-lines in the regression lists should be enough (rather
> than re-arranging all the tests in the regression.ref files).

Mans?

BTW I also suggest to replace TEST.regression.list ->
TEST.regression.tests which seems more meaningful and doesn't cause
emacs to use lisp mode for the file.

Regards.
-- 
FFmpeg = Fantastic and Foolish Maxi Peaceful Elastic Genius



More information about the ffmpeg-devel mailing list