Merge lp:~sinzui/launchpad/merge-non-active-person into lp:launchpad
| Status: | Merged |
|---|---|
| Approved by: | Brad Crittenden on 2012-09-28 |
| Approved revision: | no longer in the source branch. |
| Merged at revision: | 16055 |
| Proposed branch: | lp:~sinzui/launchpad/merge-non-active-person |
| Merge into: | lp:launchpad |
| Diff against target: |
82 lines (+49/-1) 2 files modified
lib/lp/services/verification/browser/logintoken.py (+2/-1) lib/lp/services/verification/browser/tests/test_logintoken.py (+47/-0) |
| To merge this branch: | bzr merge lp:~sinzui/launchpad/merge-non-active-person |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Brad Crittenden (community) | code | 2012-09-28 | Approve on 2012-09-28 |
|
Review via email:
|
|||
Commit Message
Allow users to merge non-active accounts.
Description of the Change
MergePeopleView
won't find people by a NEW email address, causing merges of unactivated
accounts to fail like OOPS-e2f5d3b6769c012de049daf4a60c0941.
This is basically the same as bug #1019975, just at a later stage.
The fix was to pass filter_status=False to the method so that NEW
addresses can be used to look up the IPerson.
-------
RULES
Pre-
* Add a test for the current case with active accounts and validated
email addresses.
* Setup the same test with a NEW email address and non-active account.
Expect the LocationError: (None, 'name') result.
* Add filter_status=False to the view's call to PersonSet.
and verify the test passes.
QA
* Visit https:/
* Follow the "Are you Ddw" link and Continue to send an confirmation email.
* Look in qastaging db for the token associated with your address:
select * from logintoken where created > '2012-09-28';
* Visit https:/
* Verify the page loads.
* Confirm
* Verify the page loads and you see that the merge is queued.
LINT
lib/
lib/
LoC
I have a 3000 line credit at the moment.
TEST
./bin/test -vvc -t MergePeopleView lp.services.
IMPLEMENTATION
This is a rare case where the implementation plan was exactly as expected.
lib/
lib/

Looks good to me. I'd prefer sticking to 'example.com' for fake addresses but I'm an old fuddy-duddy.