svn commit: r229081 - stable/9/usr.bin/grep

[John Baldwin]

On Saturday, December 31, 2011 9:06:24 am Gábor Kövesdán wrote:
> On 2011.12.31. 15:00, Gleb Kurtsou wrote:
> >>   	${BINDIR}/grep ${BINDIR}/fgrep \
> >> >    	${BINDIR}/grep ${BINDIR}/zgrep \
> >> >    	${BINDIR}/grep ${BINDIR}/zegrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/zfgrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/xzgrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/xzegrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/xzfgrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/lzgrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/lzegrep \
> >> >  -	${BINDIR}/grep ${BINDIR}/lzfgrep
> >> >  +	${BINDIR}/grep ${BINDIR}/zfgrep
> > Is there a reason we need all this mess is the first place? 12 grep links 
under
> > /usr/bin. Why not to make zgrep handle all compression types. Tranparent
> > compression handling in tar was a great success.
> >
> 1, xzgrep comes from archivers/xz so scripts may be using it
> 2, the rest to follow the naming convention
> 3, auto-detection is not always trivial

Presumably the auto-detection is more trivial than that in tar that handles 
the combination of various archive formats and various compression formats.  I 
would definitely prefer to have a single "smart" zgrep that DTRT.

-- 
John Baldwin