[test](inverted index) remove regression case for inverted index cache

[shuke987]

Proposed changes

Backport master PR #58137 to branch-4.1:

• Remove unstable regression case fault_injection_p0/test_inverted_index_cache.groovy.
• Register the BE UT CacheCombinationMatrix, which covers the query/searcher cache hit/miss matrix.
• Keep branch-4.1's typed inverted-index reader test interface while resolving the cherry-pick conflict.

Root cause

The branch-4.1 x64 NonConcurrent case still runs an old debug-point regression test that master already removed in favor of BE UT coverage. The regression test can fail with an intentionally injected error when the runtime path reaches InvertedIndexReader.handle_searcher_cache_miss.

Failing example:

• Build: 201293
• Occurrence: 2000000108
• Test: fault_injection_p0.test_inverted_index_cache.test_inverted_index_cache
• Error: [INTERNAL_ERROR]handle searcher cache miss

Validation

• Verified commit 86b9973e2bf / PR [test](inverted index) remove regression case for inverted index cache, use ut instead #58137 was not in origin/branch-4.1.
• Checked no duplicate open branch-4.1 PR for this case.
• Resolved cherry-pick conflict by preserving branch-4.1's current typed query interface.
• Ran git diff --check origin/branch-4.1...HEAD.

Full BE UT was not run locally.

[hello-stephen]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR.

Please clearly describe your PR:

1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
3. What features were added. Why was this function added?
4. Which code was refactored and why was this part of the code refactored?
5. Which functions were optimized and what is the difference before and after the optimization?

[shuke987]

run buildall

[shuke987]

/review

[github-actions]

Comment-only review: I did not find a blocking correctness issue in this PR.

Checked the changed files and surrounding cache paths. The removed Groovy fault-injection suite is replaced by the newly exposed InvertedIndexReaderTest.CacheCombinationMatrix BE unit test, whose helper warms the query/searcher caches and asserts the expected hit/miss counters for the query-cache hit, query-cache disabled/searcher-cache hit, query-cache enabled/searcher-cache disabled, and both-disabled cases. StringTypeInvertedIndexReader::query() returns before handle_searcher_cache() on query-cache hits, matching the new assertion. be/test/CMakeLists.txt glob-includes this test file, and I found no stale references to the deleted regression suite outside the review bundle.

Critical checkpoints:

• Existing inline review context: live PR inline comments were empty, so there was no duplicate issue to suppress.
• User focus: no additional user-provided review focus was supplied.
• Regression/session coverage: existing SQL suites still exercise enable_inverted_index_query_cache and enable_inverted_index_searcher_cache propagation; this PR's BE test covers the cache branch matrix deterministically.
• Validation limits: static review only. The checkout is shallow and lacks the base commit locally, so GitHub PR file metadata was used as the authoritative diff. Local BE test execution was not possible because thirdparty/installed, thirdparty/installed/bin/protoc, and a built doris_be_test binary are missing. A path-limited git show --check produced no whitespace warnings.

Subagent conclusions: optimizer-rewrite found no optimizer/rewrite or semantic-equivalence candidate. tests-session-config found no regression, session/config, determinism, compatibility, or style candidate. Convergence round 1 ended with both live subagents replying NO_NEW_VALUABLE_FINDINGS for the final ledger/comment set.

[hello-stephen]

BE UT Coverage Report

Increment line coverage 🎉

Increment coverage report
Complete coverage report

Category	Coverage
Function Coverage	54.78% (20954/38252)
Line Coverage	38.21% (199997/523348)
Region Coverage	34.68% (156997/452693)
Branch Coverage	35.63% (68544/192357)

[shuke987]

run buildall

[hello-stephen]

BE UT Coverage Report

Increment line coverage 🎉

Increment coverage report
Complete coverage report

Category	Coverage
Function Coverage	54.78% (20954/38252)
Line Coverage	38.21% (199951/523348)
Region Coverage	34.67% (156956/452693)
Branch Coverage	35.62% (68522/192357)

[hello-stephen]

BE Regression && UT Coverage Report

Increment line coverage 100% (0/0) 🎉

Increment coverage report
Complete coverage report

Category	Coverage
Function Coverage	73.80% (27582/37375)
Line Coverage	57.40% (298954/520847)
Region Coverage	55.01% (250703/455781)
Branch Coverage	56.36% (108586/192649)