Test coverage
Bazel can collect test coverage with a dedicated bazel coverage command.
This re-builds source files with instrumentation that captures which lines are executed then spawns the test runner with flags to collect coverage data for supported languages/test runners. Finally Bazel combines coverage reports from all test runners, so you can report overall coverage in a multi-language project.
The quality of this support varies dramatically from ruleset to ruleset (i.e. across languages). Also note that the coverage command sets flags which cause the analysis cache to be discarded, which can cost minutes in CI. In practice, it was probably a mistake to make coverage a separate command. Consider using the equivalent bazel test --collect_coverage --instrumentation_filter=^//
Test Cases
Most test runners support the JUnit XML format for reporting results at the individual test case level. The implementation of the testing ruleset must have explicit support for this.
These are collected in the bazel-testlogs folder. Bazel can also report the individual test case results to its terminal output with the flag --test_summary=testcase.
Other test outputs
You might want to grab screenshots or other output files that a test generates. Bazel doesn't allow tests to produce outputs the way build steps can, since tests are not build "Actions" but rather just some program being run.
You can read the environment variable TEST_UNDECLARED_OUTPUTS_DIR from your test, and write files into that folder. After Bazel finishes, you can collect the results as folders from the bazel-testlogs folder.
