On Fri, Sep 30, 2011 at 10:55 PM, Dan McGee <dpmcgee@gmail.com> wrote:
On Fri, Sep 30, 2011 at 1:34 AM, lolilolicon <lolilolicon@gmail.com> wrote:
On Fri, Sep 30, 2011 at 2:20 PM, Allan McRae <allan@archlinux.org> wrote:
On 30/09/11 15:23, lolilolicon wrote:
bsdtar -cf - $comp_files * | case "$PKGEXT" in - *tar.gz) gzip -c -f -n ;; - *tar.bz2) bzip2 -c -f ;; - *tar.xz) xz -c -z - ;; - *tar.Z) compress -c -f ;; - *tar) cat ;; + *tar.gz) gzip -c -f -n ;; + *tar.bz2) bzip2 -c -f ;; + *tar.xz) xz -c -z - ;; + *tar.Z) compress -c -f ;; + *tar) cat ;; + *) warning "$(gettext "'%s' is not a valid archive extension." <&-)" \ + "$PKGEXT" <&- >&2; cat ;;
I have no idea what all the <&- etc does here...
<&- closes stdin for the command. Not really needed, but explicitly ensures the stream piped from bsdtar is passed through to cat. Let's not fix (and make more complex!) what's not broken. -1.
Honestly I thought this warning was the reason the $EXT was not removed in the first place... Oh well, let's implement this when something breaks for not doing it :P
Also, why did you reindent everything here? Totally unnecessary and just makes the noise > signal here.
It was leading TAB mixed spaces. We should fix it.
For example
$ echo a | { tr a A; cat; } A $ echo a | { tr a A <&-; cat; } tr: read error: Bad file descriptor a
warning and gettext should not read the pipe, but explicitly closing their stdin doesn't hurt. At least, that's my theory...