[#2020] set up the new data structures, we should start using them. A total repo refresh is in progress to backfill the data needed.
Fresh sandbox; got this error running Allura tests:
Traceback (most recent call last):
File "/home/tvansteenburgh/tvansteenburgh-6014/forge/Allura/allura/tests/model/test_repo.py", line 274, in test_set_last_commit_nodate
obj, isnew = M.RepoObject.upsert('foo1')
AttributeError: 'module' object has no attribute 'RepoObject'
forge:tv/3059 (I merged the original rc/3059 into this.)
My commits contain some cleanup based on my code review. I removed a lot of code that is no longer used, added a couple tests, and fixed some things that were broken. Just give it a quick code-review to make sure I didn't do anything dumb. Tests all pass. There is still an issue with the CommitRunBuilder which breaks the commit browser - that's being handled under [#3759].
For SVN commits, the view of that commit does not list which files changed. E.g. /p/test3059/code-0/4/ However, the specific diff on the file like /p/test3059/code-0/4/tree/trunk/x?diff=4f69ebd79c1040786f000037:3 does work.
Forking doesn't work right either. I made a git repo with just a few commits and pushed them up (separately, if that matters). Then I went and used the fork link to fork the repo. The forked repo's tree view had unknown and None fields. History and commit views seem to work fine. This also happens if you manually fork a repo and push it into a different tool (not using the builtin fork form).
Lastly, we need to figure how we want to handle this in production. Some pages don't work well until they are reindexed. But reindexing will take quite some time. Would be nice to push out the reindexing code first, but it isn't clearly separated from the view changes :(
+ 1ca6c12...9f665a5 tv/3759 -> tv/3759 (forced update)
Latest development is continuing on tv/3759. Rick committed fixes for SVN diffs and git/hg forking. I tested these, rebased the branch against dev, and pushed it back.
In the process of testing I found two new bugs:
Tested manually and these two problems are no longer occurring. I did get a test failure though:
Running tests in module ForgeHg
Traceback (most recent call last):
File "/var/local/env-allura/lib/python2.7/site-packages/nose-1.1.2-py2.7.egg/nose/case.py", line 197, in runTest
File "/home/tvansteenburgh/tvansteenburgh-6034/forge/ForgeHg/forgehg/tests/functional/test_controllers.py", line 141, in test_diff
resp = resp.click(description='diff')
File "/var/local/env-allura/lib/python2.7/site-packages/WebTest-1.3.1-py2.7.egg/webtest/app.py", line 224, in click
File "/var/local/env-allura/lib/python2.7/site-packages/WebTest-1.3.1-py2.7.egg/webtest/app.py", line 320, in _find_element
IndexError: No matching elements found (from 15 possible)
assigned_to: Tim Van Steenburgh --> Rick Copeland ☕
qa: ⸘Dave Brondsema‽ --> Tim Van Steenburgh
Gonna take a crack at this test failure since Rick can't get to it any time soon.
+ 24857f1...88f9384 tv/3759 -> tv/3759 (forced update)
Test was out-of-date after Cory's diff refactor, which was recently rebased into this branch.
QA: Tests all pass. Find more broken stuff?
Easy testing worked fine, so I went with a big test. I think the bug can probably be replicated on a smaller scale too. Here's what I did.
+ 2fb0961...54073ed tv/3759 -> tv/3759 (forced update)
QA: Same test - push the allura repo into a new Code tool, and confirm that the folder commits are correct now.
Log in to post a comment.