On 02/16/10 04:07 PM, Brock Pytlik wrote:
On 02/16/10 11:10 AM, Shawn Walker wrote:
Greetings,
The following webrev contains fixes for the following issue:
14292 pkg list doesn't exit with partial failure when some names are
unmatched
webrev:
http://cr.opensolaris.org/~swalker/pkg-14292/
In short, this changes the client to perform its own additional input
checking as the results are returned by the API to determine if *all*
of the patterns provided matched at least one of the packages listed.
The client already correctly handled the single pattern case, or the
case where all of the patterns specified didn't match.
Finally, I would note that (on my system) this appears to have no
measurable impact on the performance of pkg list.
Cheers,
client.py:
447-451: the comment here suggests there is an impact on performance,
but what you wrote in the email suggests there's not. Assuming the
email's right, could this comment either be removed, or a more detailed
description of the impact given?
I'll just remove the comment. It's logical that there would be some
amount of performance impact given the extra work that would be done,
but I personally can't measure it on *my* own system.
346-389: couldn't this just be a call to a new api function so that this
code isn't copied from the code in api.py? That would solve the hard
coding of brelease wouldn't it? It looks to me like this code could be a
function that takes in pats, and returns illegals and pat_tubles. Am I
missing something?
Actually, the code was originally in pkg list and then put into the API
;) However, I'll see if I can come up with something generic enough
that I'm comfortable offering it as part of the API. If the change is
big enough to the API, I'll probably skip it given our release status.
--
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
