Merge lp:~leonardr/lazr.restfulclient/test-long-etag into lp:lazr.restfulclient
| Status: | Merged |
|---|---|
| Merged at revision: | not available |
| Proposed branch: | lp:~leonardr/lazr.restfulclient/test-long-etag |
| Merge into: | lp:lazr.restfulclient |
| Diff against target: |
112 lines (+17/-24) 4 files modified
src/lazr/restfulclient/NEWS.txt (+6/-0) src/lazr/restfulclient/_browser.py (+1/-7) src/lazr/restfulclient/docs/authorizer.txt (+9/-16) src/lazr/restfulclient/version.txt (+1/-1) |
| To merge this branch: | bzr merge lp:~leonardr/lazr.restfulclient/test-long-etag |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Brad Crittenden (community) | code | 2010-03-16 | Approve on 2010-03-16 |
|
Review via email:
|
|||
Description of the Change
Ignore the name of this branch; the thing I was testing didn't need any changes, but I found some test failures while I was testing.
A new version of simplejson breaks some lazr.restfulclient tests: what used to be a ValueError is now a JSONDecodeError. I had code that created a fake web service and interpreted a ValueError as a sign that a client had managed to authenticate with the service--since the web service didn't actually serve any JSON, the client's next step after authenticating ended in failure.
I decided to fix the tests by making the fake web service a little better, making it serve some (fake) JSON, so that if the client authenticates with the fake web service there will be no error at all.
To keep the fake web service simple, I removed some no-longer-needed code that was used to fix bug 457146. Since Launchpad now runs a version of lazr.restful that doesn't have bug 457146 (and has for months), I think it's safe to remove the client-side hack. I won't remove the corresponding code from lazr.restful, because old versions of lazr.restfulclient are still around and manifesting bug 457146.
