nits: sort imports, doc typos, minor cleanups
factored out of #612
@ryan-williams Thanks for the PR. We would be glad to get this PR and #218 + #323 going once we have merged #580 which is a significant change on master.
rebased this, and looks like #580 was merged, so maybe this is ready for another look.
"Python tests on ubuntu-latest" failed, GHA log says:
This step has been truncated due to its large size. Download the full logs from the menu once the workflow run has completed.
I downloaded the logs and see this toward the end of 1_tests Python tests on ubuntu-latest.txt (48M):
2022-01-06T15:30:01.2401382Z [pid 1759] ### The following tests failed: ###
2022-01-06T15:30:01.2402956Z [pid 1759] ### test 'CardTimeoutTest' graph 'simple-foreach' in context python3-all-local ###
2022-01-06T15:30:01.2404842Z [pid 1759] ### test 'CardTimeoutTest' graph 'single-and-branch' in context python3-all-local ###
2022-01-06T15:30:01.2406043Z [pid 1759] ### test 'CardTimeoutTest' graph 'nested-foreach' in context python3-all-local ###
2022-01-06T15:30:01.2407218Z [pid 1759] ### test 'CardTimeoutTest' graph 'nested-branches' in context python3-all-local ###
2022-01-06T15:30:01.2408392Z [pid 1759] ### test 'CardTimeoutTest' graph 'single-linear-step' in context python3-all-local ###
2022-01-06T15:30:01.2409566Z [pid 1759] ### test 'CardTimeoutTest' graph 'small-foreach' in context python3-all-local ###
2022-01-06T15:30:01.2410920Z [pid 1759] ### test 'CardTimeoutTest' graph 'small-parallel' in context python3-all-local ###
2022-01-06T15:30:01.3857245Z ERROR: InvocationError for command /home/runner/work/metaflow/metaflow/test_runner (exited with code 1)
2022-01-06T15:30:01.3860641Z ___________________________________ summary ____________________________________
2022-01-06T15:30:01.3861286Z ERROR: python: commands failed
2022-01-06T15:30:01.4089259Z ##[error]Process completed with exit code 1.
Not sure what that is about. investigating…
Update: this was due to the black issue fixed in #885
I think this is ready for review
going through some older PR's, this seems like a straightforward one to push through. Would you still be available for rebasing it against the latest master ?