Fix double-counting of subtest failures in test summary #13986
Description
What's the problem this feature will solve?
When running tests using subtests in pytest, failures are being double-counted in the final summary. For example, if 1 test is being run with 3 subtests where 2 of the subtests pass and 1 fails, the summary incorrectly shows "2 failed, 2 subtests passed" instead of properly separating the counts.
This makes it difficult to understand the actual test results at a glance, as subtest failures are counted both as regular test failures and as subtest failures, leading to inflated failure counts.
Describe the solution you'd like
The summary should properly separate subtest failures from regular test failures. For the example above (1 test with 3 subtests, 1 subtest fails, 2 pass), the summary should show:
============= 1 failed, 1 subtests failed, 2 subtests passed in 0.02s =============
This provides clear visibility into:
- How many top-level tests failed (1 failed)
- How many subtests failed (1 subtests failed)
- How many subtests passed (2 subtests passed)
This makes it much easier to understand the test results and identify whether failures are in subtests or in the main test logic.
Alternative Solutions
Currently, users need to manually parse the detailed output to understand the actual failure counts, which is error-prone and time-consuming. There's no workaround that provides the same clarity as fixing the counting logic.
Additional context
This is a feature that improves the accuracy and clarity of pytest's test summary output when using subtests. The change ensures that subtest failures are counted in their own category ("subtests failed") rather than being double-counted as both regular failures and subtest failures.