Merge lp:~sil2100/ubuntu-archive-tools/sru-review-invalid-bugs into lp:ubuntu-archive-tools

Proposed by Łukasz Zemczak
Status: Merged
Merged at revision: 1159
Proposed branch: lp:~sil2100/ubuntu-archive-tools/sru-review-invalid-bugs
Merge into: lp:ubuntu-archive-tools
Diff against target: 18 lines (+5/-3)
1 file modified
sru_workflow.py (+5/-3)
To merge this branch: bzr merge lp:~sil2100/ubuntu-archive-tools/sru-review-invalid-bugs
Reviewer Review Type Date Requested Status
Steve Langasek Approve
Brian Murray Approve
Review via email: mp+340754@code.launchpad.net

Commit message

Propose not excluding tasks that are in 'invalid/released' states when handling sru-review bugs as otherwise the tool can crash trying to re-add a task for an existing suite.

Description of the change

Propose not excluding tasks that are in 'invalid/released' states when handling sru-review bugs as otherwise the tool can crash trying to re-add a task for an existing suite.

I have decided that printing a warning is enough as this step happens after the SRU has been already accepted from the queue, so prompting for reviewer input is at this moment meaningless. A reviewer knows what she/he is doing and otherwise the tool crash can just cause the SRU bugs not being handled at all.

To post a comment you must log in.
Revision history for this message
Steve Langasek (vorlon) wrote :

I don't like the idea of having the review tool automatically setting such bug tasks back to fix-committed; this seems like an error on the input side which should be resolved prior to accepting the SRU, because the correct resolution varies. (Sometimes this means the bug task should be reopened, but sometimes it indicates a wrong bug is linked from the SRU and the tool shouldn't presume.)

review: Needs Fixing
Revision history for this message
Brian Murray (brian-murray) wrote :

This seems fine to me.

review: Approve
Revision history for this message
Łukasz Zemczak (sil2100) wrote :

Steve: I agree it might somehow be a problem, but leaving it as is right now is breaks the whole tool and makes the reviewer's life really hard (as sru-review crashes after accepting and no bugs get updated and need to be updated manually). The best solution would be of course to check the bugs before actually accepting the SRU, but this means a bigger rewrite of the code. My assumption is that the SRU reviewers responsibility is to make sure the package is good for -proposed promotion before actually accepting, during the review - so the status is just a formality that was left untouched.

Revision history for this message
Steve Langasek (vorlon) wrote :

ok, objections withdrawn

review: Approve
Revision history for this message
Brian Murray (brian-murray) wrote :

Given Steve's concerns the print statement should be more informative about the fact that the reviewer should double check the bug for any errors.

1160. By Łukasz Zemczak

Make the warning more verbose.

Revision history for this message
Łukasz Zemczak (sil2100) wrote :

Made more informative. Thanks!

Preview Diff

[H/L] Next/Prev Comment, [J/K] Next/Prev File, [N/P] Next/Prev Hunk
1=== modified file 'sru_workflow.py'
2--- sru_workflow.py 2017-12-07 16:28:15 +0000
3+++ sru_workflow.py 2018-03-06 08:52:48 +0000
4@@ -38,9 +38,11 @@
5 print("Ignoring task %s in bug %s" % (task.web_link, num))
6 continue
7 sourcepkg_match = True
8- if (match.group('suite') == release and
9- task.status not in ("Invalid", "Won't Fix",
10- "Fix Released")):
11+ if match.group('suite') == release:
12+ if task.status in ("Invalid", "Won't Fix", "Fix Released"):
13+ print("Matching task was set to %s before accepting the SRU, "
14+ "please double-check if this bug is still liable for "
15+ "fixing. Switching to Fix Committed." % task.status)
16 task.status = "Fix Committed"
17 task.lp_save()
18 print("Success: task %s in bug %s" % (task.web_link, num))

Subscribers

People subscribed via source and target branches