Proposed commit message:
www/py-aiohttp: update to 2.3.1 PR: 222525 Reviewed_by: koobs (maintainer) Approved_by: koobs (maintainer) Differential_Revision: https://reviews.freebsd.org/D12655
Details
Details
- Reviewers
philip
portlint: OK (looks fine.)
testport: OK (poudriere: 10amd64, 11amd64 and 12amd64)
maketest: OK (5 failed, 1965 passed, 36 skipped, 5 xfailed in 42.83 seconds)
The five failed tests are not actually failed -- all of them "fail" as follows:
E AssertionError: assert <CIMultiDict(...': 'header2')> == {'h1': 'h1', 'h2': 'header2'}
E Full diff:
E - <CIMultiDict('H1': 'h1', 'h2': 'header2')>
E + {'h1': 'h1', 'h2': 'header2'}
tests/test_client_session.py:153: AssertionErrorI'm not sure how the <CIMultiDict(> creeps into the output there. The actual output is as expected. I suspect outdated TEST_DEPENDS.
I did a run-test with the updated py3-threema-msgapi and everything seems quite happy. It wouldn't be nearly as happy if there was actual brokenness.
Diff Detail
Diff Detail
- Repository
- rP FreeBSD ports repository
- Lint
No Lint Coverage - Unit
No Test Coverage - Build Status
Buildable 12151 Build 12452: arc lint + arc unit
Event Timeline
Comment Actions
Okay: figure out the problem with the test results. multidict changed subtly between 3.0.0 and 3.2.1 (which we now have in the ports tree), causing the tests to fail. This is an upstream issue. It only affects the tests, it does not appear to affect functionality. The tests also fail with multidict 3.3.0 (latest upstream).
I will check with upstream if they've fixed the issue in master. If not I'll file a bug report. In any case, I don't think this should block updating the port on FreeBSD to 2.3.1.
Comment Actions
PR patch is otherwise approved (URL removed to this review).
If you'd like to leave this review linked to that bugzilla issue, update the differential here, add the review URL back to the PR