separate subtests failures to give better indication on tests analysis

[RoniKishner]

Description

This PR fixes the double-counting of subtest failures in pytest's final test summary. Previously, when a test with subtests had failures, both the subtest failure and the top-level test failure were counted as regular "failed" tests, leading to inflated failure counts.

Changes

• Modified pytest_report_teststatus hook in src/_pytest/subtests.py to return "subtests failed" category instead of "failed" for subtest failures
• Updated _determine_main_color in src/_pytest/terminal.py to consider "subtests failed" when determining the main color (red)
• Enhanced parse_summary_nouns in src/_pytest/pytester.py:
  • Updated regex to handle multi-word outcomes like "subtests failed"
  • Added pluralization mapping for subtest outcomes
  • Added whitespace handling for parsed outcomes
• Updated test expectations in testing/test_subtests.py to match the new summary format

Example

Before:
===================== 2 failed, 2 subtests passed in 0.02s =====================
After:
===================== 1 failed, 2 subtests passed, 1 subtests failed 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)

Closes #13986

[nicoddemus]

Thanks @RoniKishner for the PR.

I'm OK with separating subtest failures from normal failures.

I left a suggestion regarding the changelog, could you also take a look at the failing tests?

[nicoddemus]

Suggested change

	Fixed double-counting of subtest failures in the final test summary. Subtest failures are now reported separately as "subtests failed" instead of being counted as regular "failed" tests, providing clearer statistics. For example, a test with 3 subtests where 1 fails and 2 pass now shows "1 failed, 1 subtests failed, 2 subtests passed" instead of "2 failed, 2 subtests passed".
	Show subtests failures separate from normal test failures in the final test summary.
	Subtest failures are now reported separately as "subtests failed" instead of being counted as regular "failed" tests, providing clearer statistics.
	For example, a test with 3 subtests where 1 fails and 2 pass now shows ``1 failed, 2 subtests passed, 1 subtests failed`` instead of ``2 failed, 2 subtests passed``.