Fixes spurious error messages in tests #1400
Conversation
The tests are still "passing" locally (ie- the test process doesn't fail and all 333 tests are marked as passing). It's just that there is a subprocess command that gets invoked and fails. The error is swallowed by the action (I think that's the behaviour we want). And most likely, we're not seeing the same messages on CI since all of the associated env vars are in place and the workflow file is where we expect it to be. |
Right...forgot about the second one. Let me get that in, too. |
Previously, `isAnalyzingDefaultBranch` was failing because there are some missing env vars: `GITHUB_SHA`, `GITHUB_REF`, and `GITHUB_EVENT_PATH`. Also, `checkout_path` is missing as an input. Rather than trying to set them to mock values, which would require setting the paths to existing paths in the file system, I chose to stub the entire function. I think this is fine since the point of the test is to check the ram and threads values, not testing the `isAnalyzingDefaultBranch` function.
aeisenberg
force-pushed
the
aeisenberg/fix-test-error
branch
from
e7f1de4
to
1384ce4
Compare
Previously,
isAnalyzingDefaultBranchwas failing because there are some missing env vars:GITHUB_SHA,GITHUB_REF, andGITHUB_EVENT_PATH. Also,checkout_pathis missing as an input.Rather than trying to set them to mock values, which would require setting the paths to existing paths in the file system, I chose to stub the entire function. I think this is fine since the point of the test is to check the ram and threads values, not testing the
isAnalyzingDefaultBranchfunction.Merge / deployment checklist