Conversation
|
Also note that there is a small test failure to do with a line change on E assert "Module 'info...ogus-data.bla" == "Module 'info...ogus-data.bla"
E - Module 'info' failed with status code 71:
E ? ^
E + Module 'info' failed with status code 72:
E ? |
Yes, please see this: GenericMappingTools/gmt#4347 (comment). Although the test failure is caused by upstream change, but I think it's a bad test to check the undocumented status codes. We can see if we can improve the test. |
E assert "Module 'info...ogus-data.bla" == "Module 'info...ogus-data.bla"
E - Module 'info' failed with status code 71:
E ? ^
E + Module 'info' failed with status code 72:
E ? Should we check that the status code is 71 for GMT<=6.1.1, and 72 for GMT>=6.2.0? It's really a bad idea to test against the undocumented status codes. |
|
I would do something like |
|
Let's wait for more discussions in GenericMappingTools/gmt#4347. |
Description of proposed changes
It's annoying to have failing CI tests. Every time I want to merge a PR, I have to change the branch protection rules, merge a PR, and then change the protection rules back.
This PR contains two changes:
pytest.mark.xfaildoesn't work as we expect.info()tests fail for the GMT master branch, unless the upstream issue Running info on vector memory inputs return UNIX timestamps instead of ISO datetimes gmt#4241 is fixed.Reminders
make formatandmake checkto make sure the code follows the style guide.doc/api/index.rst.