On 03/24/10 08:44 PM, Shawn Walker wrote:
On 03/24/10 10:42 PM, Liane Praza wrote:
On 03/24/10 08:36 PM, Shawn Walker wrote:
On 03/24/10 06:40 PM, Liane Praza wrote:
Here's my review for
15346 need a not_these_pkgs to handle skipping unincorporated packages
in an effort to exhaust more of the option letter space for the
importer.
http://cr.opensolaris.org/~lianep/pkg-re/
Comments appreciated, I'm offline for a few hours and will get back to
them later this evening.
I think -J is fine, it's similar to what we do for pkg(1), (-m to add
mirror, -M to remove mirror, etc.).
importer.py:
line 1349: extra newline?
lines 1404-1405: This could traceback if they specified a package that
doesn't actually exist. I know that's unlikely since you're hardcoding
the package names, but, just in case... something like the following?
unknown = []
for pkg in not_these_pkgs:
try:
del pkgdict[pkg]
except KeyError:
unknown.append(pkg)
if unknown:
emsg("Unknown package(s) specified using -J:\n%s" % \
"\n".join(p for p in unknown))
sys.exit(2)
Hrm. The importer isn't terribly resilient currently to tracebacks, and
I'm unconvinced that the friendlier error is worth the retest cycle at
this point, as opposed to getting the bits to Comay sooner.
Can I defer this until after 136 closes?
Sure, as you said, the importer has a limited lifespan anyway.
And I wasn't aware (as you pointed out in another reply) that you
weren't expecting it to be used for anything else.
OK, thanks. I'll file a bug to capture the comment and take ownership
of it, though.
liane
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
