Test failures in code coverage GHA and how to investigate them further

[IndrajeetPatil]

I am using the example code-coverage GHA across multiple respos, and I am often seeing tests fail while computing coverage when they don't fail otherwise.

For example, even though none of the tests fail in R CMD Check, some tests seem to fail while computing code coverage.

But the error doesn't indicate which ones:

Error: Error: Failure in `/tmp/RtmpeQ0RJZ/R_LIBS324e25ce0131/bayestestR/bayestestR-tests/testthat.Rout.fail`
quasi_capture(...)
 4. │     ├─testthat .capture(...)
 5. │     │ └─base::withCallingHandlers(...)
 6. │     └─rlang::eval_bare(quo_get_expr(.quo), quo_get_env(.quo))
 7. └─testthat::expect_equal(...)

[ FAIL 2 | WARN 2 | SKIP 4 | PASS 365 ]
Error: Error: Test failures

I also downloaded logs to see if this information is available there, but to no avail. If I run covr::package_coverage() locally, it doesn't show any test failures either. So I am at a complete loss as to how to figure out which tests are failing.

Is it possible for this GHA to be modified in any way to provide details about which tests failed? Thanks.

[gaborcsardi]

Indeed, this is hard: #481. You can try to modify the covr call to do the check in a certain directory and then upload that directory as an artifact. I am not sure how difficult this is.

[gaborcsardi]

Closed by #643. No the testthat.Rout* files are printed, and the whole check directory is also uploaded as an artifact. Example: https://github.com/r-lib/actions/actions/runs/3297405227

[github-actions[bot]]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue and include a link to this issue