On Fri 14 Aug 2009 at 12:15AM, Danek Duvall wrote:
> On Mon, Aug 03, 2009 at 02:20:17PM +0100, Darren J Moffat wrote:
>
> > http://cr.opensolaris.org/~darrenm/pkg/
>
> I can't speak to the design; Dan will have to comment on that, but for the
> specific problem you're trying to fix, it seems reasonable to me.
>
> Otherwise, the change looks fine to me, though you could have simply added
> the "exit 0" to the existing awk command. Also, though you didn't
> introduce this, awk will always exit 0 in this case, so the test for the
> return code is useless. You could add an "END {exit 1}" clause, though
> since we're already assuming some stability of the output format, perhaps
> we don't have to do that. :) If what we care about is the exit status of
> the pkg command, then we might have to rely on ksh93's pipefail option
> (though that doesn't seem to work for me the way I'd expect).
>
> Similarly, the way get_pub_secinfo is used, the return value of the
> function is meaningless. It should really print "None None" and return on
> failure, and its callers should test for that.
>
> Another nit: there's no need for a backslash inside $().
>
> Also, please preserve the style of the rest of the script -- no spaces
> inside "{}" or before semicolons in awk, and no spaces inside the parens in
> the $() constructs.
I looked at this. I have nothing to add to Danek's feedback.
-dp
--
Daniel Price, Solaris Kernel Engineering http://blogs.sun.com/dp
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
