MJPEG Tools Svn

Dammit --- I DID IT AGAIN: type "cvs commit yuv4mpeg.h yuv4mpeg.h" instead
of "cvs commit yuv4mpeg.h yuv4mpeg.c".

GOT IT RIGHT NOW.
Must investigate why smack to forehead failed.
Kicking self in butt in meantime.

(ps: Finally succeeded in committing code for y4m_chroma_ss_[xy]_ratio().)

o Remove sms's addition of y4m_get_chroma_ss().
o Actually commit the *code* for the y4m_chroma_ss_[xy]_ratio() functions.
o Smack myself in forehead twice (for good measure).

Note to those who read notes:

y4m_chroma_ss_[xy]_ratio() returns the scaling ratio "n:d" by which the
chroma planes are subsampled/scaled from full resolution. This means,
more or less, that any regions defined on a frame need to be aligned to
a multiple of "d" pixels --- since to a zeroth-order approximation, "d"
luma pixels correspond to "n" chroma pixels.

-matto

Add y4m_get_chroma_ss() function. Returns the subsampling factors (i.e. for
411 those are 4 and 1, for 422 those are 2 and 2, and so on). Useful for
programs that need to make sure counts are a proper multiple of the chroma
subsampling factor.

support YCbCr4:2:2 software recording/software playback

Added y4m_chroma_ss_[xy]_ratio() functions, to lookup information about
a Y4M_CHROMA_* mode.

(Go get 'em, sms.)

Tweaked Y4MPipeReader class to compile with upgraded YUV4MPEG2 API
(and fixed two potential memory-leaks in the process).

As an interim, transitionary measure:
If no 'C' chroma tag is found (or if library is running at old-skool
feature-level of 0), check for an X-tag for chroma (YSCSS=...).

patch from Nicolas Boos

Er, "y4m_accept_extensions()" is not "y4m_feature_level()".

A rose by any other name... generates a linker error.

Fixed ridiculously dumb cut'n'paste error in y4m_si_get_plane_height().
(Thank you, Steven Schultz, you must have gotten more sleep than me.)