Merge lp:~benji/launchpad/bug-597324 into lp:launchpad
| Status: | Merged |
|---|---|
| Approved by: | Gary Poster on 2010-07-23 |
| Approved revision: | no longer in the source branch. |
| Merged at revision: | 11221 |
| Proposed branch: | lp:~benji/launchpad/bug-597324 |
| Merge into: | lp:launchpad |
| Diff against target: |
81 lines (+57/-3) 2 files modified
lib/canonical/launchpad/webapp/login.py (+15/-3) lib/canonical/launchpad/webapp/tests/test_login.py (+42/-0) |
| To merge this branch: | bzr merge lp:~benji/launchpad/bug-597324 |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Gary Poster (community) | 2010-07-23 | Approve on 2010-07-23 | |
|
Review via email:
|
|||
Description of the Change
Fix a unicode sin: use the decoded query parameters instead of creating
new, undecoded copies.
This /should/ fix the OOPS introduced in on edge by this branch (see bug
609029).
I also added a test to ensure this particular failure doesn't reoccur.
| Benji York (benji) wrote : | # |
On Fri, Jul 23, 2010 at 5:52 PM, Gary Poster <email address hidden> wrote:
> - Please file a bug against canonical-
> the additional "csrfmiddleware
> bug at the top of the pertinent comment below (e.g., something like
> ``# XXX benji 2010-07-23 bug=12345``)
https:/
> - I don't love the assert error for multi-valued fields. It feels
> like it ought to be a real exception that has a chance of being
> handled by one of our error views. However, the right error doesn't
> jump out at me, so I'll let it go unless you see where I'm coming from
> and have a good idea.
Switched to a ValueError in r11111.
--
Benji York

merge-conditional
Wow, great find!
Also, your cover letter neglected to mention that this fixes the problems discovered in QA for the original bug, but your commit messages clarify it. :-)
- Please file a bug against canonical- openid- provider about removing the additional "csrfmiddleware token" field and add a reference to the bug at the top of the pertinent comment below (e.g., something like ``# XXX benji 2010-07-23 bug=12345``)
- I don't love the assert error for multi-valued fields. It feels like it ought to be a real exception that has a chance of being handled by one of our error views. However, the right error doesn't jump out at me, so I'll let it go unless you see where I'm coming from and have a good idea.
Thank you!
Gary