On 06/13/12 02:27, Saurabh Vyas wrote:
Hi All,
Please review these trivial fix
CR 7140808 pkgsend traces back on a manifest with a versionless pkg.fmri
webrev :
https://cr.opensolaris.org/action/browse/pkg/saurabhv/7140808-rev0/webrev/
src/publish.py:
line 367: This must be done *after* pfmri.version.timestr = None. So
move this to after line 377.
lines 370-376: This version check shouldn't be done inside of
"allow-timestamp" and it could be simplified.
diff --git a/src/publish.py b/src/publish.py
--- a/src/publish.py
+++ b/src/publish.py
@@ -21,7 +21,7 @@
#
#
-# Copyright (c) 2007, 2011, Oracle and/or its affiliates. All rights
reserved.
+# Copyright (c) 2007, 2012, Oracle and/or its affiliates. All rights
reserved.
#
import fnmatch
@@ -364,9 +364,16 @@ def trans_publish(repo_uri, fargs):
try:
pfmri = pkg.fmri.PkgFmri(m["pkg.fmri"], "5.11")
+ if not pfmri.version:
+ # Cannot have a FMRI without version
+ error(_("The pkg.fmri '%s' in the package
manifest "
+ "must include a version.") % pfmri,
cmd="publish")
+ return 1
if not DebugValues["allow-timestamp"]:
# If not debugging, timestamps are ignored.
- pfmri.version.timestr = None
+ if pfmri.version:
+ pfmri.version.timestr = None
+
pkg_name = pfmri.get_fmri()
except KeyError:
error(_("Manifest does not set pkg.fmri"))
lines 385-389: I don't think this change is necessary; especially if
you change things as I suggested above. If it still is, then I'd like
to understand the failure case better and it doesn't make sense to me
right now.
-Shawn
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
