[test](regression) Remove parquet complex cross page case

[suxiaogang223]

What problem does this PR solve?

Issue Number: None

Related PR: None

Problem Summary: Remove the parquet complex cross page regression case and its expected output from the hive external table suite.

Release note

None

Check List (For Author)

• Test: No need to test (removes an existing regression case only)
• Behavior changed: No
• Does this need documentation: No

[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?

[suxiaogang223]

run buildall

[Gabriel39]

/review

[github-actions]

PR approved by at least one committer and no changes requested.

[github-actions]

PR approved by anyone and no changes requested.

[github-actions]

Review completed for PR #65251.

I did not find a substantiated issue requiring an inline comment. The PR is narrowly scoped to deleting test_parquet_complex_cross_page.groovy and its matching generated .out file. I verified the live GitHub PR file list/patch still contains only those two deletions at head 91b54636f893e543a60840bc778dea64844b720f, and there are no existing inline review comments to deduplicate against.

Critical checkpoint conclusions:

• Goal/test proof: the change removes one p2 EMR Hive parquet complex cross-page regression case and its expected output. The matching suite/output pair is deleted together.
• Scope: deletion-only and focused; no source code, config, persistence, transaction, FE/BE protocol, lifecycle, or concurrency paths are modified.
• Parallel paths and coverage: I checked the remaining regression coverage. Cross-page parquet nested/complex reads are still covered by external_table_p0/hive/test_parquet_nested_types.groovy across array, nested-array, array-struct, map, null, filter, and aggregate-expression queries, and by external_table_p0/tvf/test_tvf_p0.groovy through row_cross_pages and row_cross_pages_2 parquet TVF cases. I also found no remaining active references to the deleted suite/table-specific names.
• Expected results: deleting the .out file is consistent with deleting the suite file; no orphan expected-output mapping was found.
• Observability/performance/data correctness: not applicable to this test-only deletion beyond the coverage check above.

User focus: no additional user-provided review focus was present.

Subagent conclusions: optimizer-rewrite reported no candidates; tests-session-config reported no candidates. After the main ledger recorded the dismissed coverage/orphan-output checks and the empty proposed inline comment set, convergence round 1 ended with both live subagents returning NO_NEW_VALUABLE_FINDINGS for the same current ledger/comment set.

Validation limits: static review only. The checkout is shallow/grafted and local ancestry did not contain the advertised base tree for the deleted paths, so I treated GitHub PR patch/file metadata and base blob API results as authoritative. .worktree_initialized is missing, and I did not run regression tests for this deletion-only PR.