test: add Data Branch issue BVT coverage

[Ariznawlll]

What type of PR is this?

• API-change
• BUG
• Improvement
• Documentation
• Feature
• Test and CI
• Code Refactoring

Which issue(s) this PR fixes:

issue #26071, #26073, #26074, #26078, #26081, #26114, #26118, #26120, #26121, #26127, #26128, #26136, #26205

What this PR does / why we need it:

Add executable BVT .sql/.result regression cases for fixed Data Branch and CLONE issues. The cases cover quoted identifiers, snapshots, BOOL/BIT primary keys, bulk DML, cross-account quota, CSV output errors, vector values, and deep lineage.

[qodo-code-review]

Qodo reviews are paused for this user.

Troubleshooting steps vary by plan Learn more →

On a Teams plan?
Reviews resume once this user has a paid seat and their Git account is linked in Qodo.
Link Git account →

Using GitHub Enterprise Server, GitLab Self-Managed, or Bitbucket Data Center?
These require an Enterprise plan - Contact us
Contact us →

[XuPeng-SH]

Blocking finding on exact head 1489020:

[P1] The #26081 BVT records two assertion-query failures as the expected baseline instead of validating the PICK/MERGE result. In test/distributed/cases/git4data/branch/edge/issue_26081_hashmap_bulk.sql:40-45 and :53-55, SUM(boolean-expression) is unsupported; the paired result therefore contains "invalid argument aggregate function sum, bad value [BOOL]" at lines 43 and 55. Because there is no intentional-error directive and these are the only selected/unselected and per-change-type checks, the BVT can stay green while those behavioral assertions never execute.

Please rewrite the expressions to produce integers, for example SUM(CASE WHEN condition THEN 1 ELSE 0 END), regenerate the result, and verify the expected counts for selected updates/deletes/inserts, unselected controls, and merged change classes.

I reviewed all 13 SQL/result pairs and checked all other recorded errors against explicit @regex expectations. CI is green, but it currently confirms the erroneous baseline above, so it does not close this finding.

[XuPeng-SH]

Deep re-review complete on exact head 9c7528b89cc9a2eeeac1791e87fe7b539fa211ba; no blocking finding.

Mergify rebased the already-reviewed change onto current main (37694c49f2f450f7aa7bb5560279d7ab7007c05d). All 26 PR-owned SQL/result files are byte-identical to the reviewed tree; the only tree delta is eight unrelated files brought in from main, with no conflict overlap. The exact-head PR diff passes git diff --check.

The prior blocker remains fixed: both unsupported SUM(BOOL) assertions are integer CASE aggregates, and the regenerated result proves PICK selected/unselected controls plus MERGE update/delete/insert counts instead of recording query errors. The arithmetic closes against the 50k base and the 10k update/delete/insert transformations.