[FFmpeg-cvslog] r24827 - in trunk: libavcore/avcore.h libavcore/imgutils.c libavcore/imgutils.h libavfilter/vf_crop.c libavfilter/vf_hflip.c

[Stefano Sabatini]

On date Thursday 2010-08-19 15:03:59 +0200, Stefano Sabatini wrote:
> On date Thursday 2010-08-19 10:30:14 +0200, Stefano Sabatini wrote:
> > On date Wednesday 2010-08-18 23:02:38 +0200, stefano wrote:
> > > Author: stefano
> > > Date: Wed Aug 18 23:02:38 2010
> > > New Revision: 24827
> > > 
> > > Log:
> > > Implement inline function av_fill_image_max_pixstep() and use it for
> > > factorizing code.
> > 
> > If possible I'd like to change the name of the function, from
> > av_fill_image_max_pixstep() to av_get_image_max_pixstep().
> > 
> > "fill" sounds like a setter so it was a bad choice, "get" is also
> > shorter, and is also consistent with av_get_image_linesize().
> > 
> > It is OK just do it, or do I have to add a backward-compatibility
> > layer? Since there is no risk of breaking apps then I'd just bump
> > micro and consider it like an API bugfix.
> 
> I'll apply the change tonight if no one is against (and I won't change
> my mind again).

Other ideas:

static inline
void av_{get,fill}_image_max_pixsteps(int max_pixsteps[4], int max_pixstep_comps[4],
                                      const AVPixFmtDescriptor *pixdesc);

I prefer the plural form as we're getting/filling an array of
elements, also that would be consistent with:
av_fill_image_pointers
av_fill_image_linesizes

as for the fill/get dilemma after more thought I don't think it is
such important, "fill" makes sense since we're filling an array with
the image pixsteps, since the term is already used then it should not
be an issue.

Regards.