fix(cubesql): Resolve schema-qualified pg_catalog functions

[MazterQyou]

Check List

• Tests have been run in packages where changes have been made if available
• Linter has been run for changed code
• Tests for the changes have been added if not covered yet
• Docs have been added / updated if required

Description of Changes Made

This PR bumps cube-js/arrow-datafusion@c7e6855, fixing a regression introduced with sqlparser bump where parsing pg_catalog.array_agg leads to parsing it as a compound function name. Related test is included.

[claude]

Claude finished @MazterQyou's task in 27s —— View job

---

Code Review

Review complete — LGTM, no blocking issues.

• Read PR diff
• Reviewed dep bump, test, and snapshot
• Assessed risk/coverage

Summary

Targeted fix: bumps cube-js/arrow-datafusion to c7e6855, which restores parsing of compound (schema-qualified) function names like pg_catalog.array_agg that regressed with the recent sqlparser bump. Adds a snapshot-backed regression test.

Findings

• rust/cubesql/cubesql/Cargo.toml:13 — dep pin looks fine; matches Cargo.lock updates on both sides (rust/cubesql/Cargo.lock, packages/cubejs-backend-native/Cargo.lock).
• rust/cubesql/cubesql/src/compile/mod.rs:17279-17291 — the new test_pg_catalog_array_agg test is minimal but exactly targets the regression (schema-qualified aggregate in SELECT over VALUES). Snapshot at pg_catalog_array_agg.snap produces the expected {1,2,3} array literal.
• No changes to production Rust code beyond the dep bump — the actual fix lives upstream in the datafusion fork, so surface area here is limited.

Suggestions (non-blocking)

• Consider adding one more test covering a schema-qualified scalar (non-aggregate) function to lock down parity, e.g. pg_catalog.upper('x') or pg_catalog.pg_get_expr(...), since the underlying parser fix likely covers all compound identifiers, not just aggregates. Optional — the current test is sufficient to catch the specific regression.
• No docs/API changes needed.

Verdict: LGTM. Straightforward upstream fix pull-through with an appropriate regression test.

• Branch: `cubesql/compound-function-fix`

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.26%. Comparing base (f9398ab) to head (a75400a).
⚠️ Report is 2 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #11190      +/-   ##
==========================================
- Coverage   83.70%   79.26%   -4.45%     
==========================================
  Files         257      474     +217     
  Lines       79465    97024   +17559     
  Branches        0     3572    +3572     
==========================================
+ Hits        66513    76902   +10389     
- Misses      12952    19606    +6654     
- Partials        0      516     +516     

Flag	Coverage Δ
cube-backend	59.13% <ø> (?)
cubesql	83.70% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[claude]

Claude finished @MazterQyou's task in 27s —— View job

---

Code Review

LGTM — no blocking issues. Already approved by @ovr.

• Read PR diff
• Reviewed dep bump, test, and snapshot
• Assessed risk/coverage

Summary

Targeted fix: bumps cube-js/arrow-datafusion to c7e6855, restoring parsing of schema-qualified function names like pg_catalog.array_agg that regressed with the recent sqlparser bump. Adds a snapshot-backed regression test.

Findings

• rust/cubesql/cubesql/Cargo.toml:13 — dep pin looks fine; matches Cargo.lock updates on both sides (rust/cubesql/Cargo.lock, packages/cubejs-backend-native/Cargo.lock).
• rust/cubesql/cubesql/src/compile/mod.rs:17279-17291 — new test_pg_catalog_array_agg test is minimal and targets the regression precisely (schema-qualified aggregate in SELECT over VALUES). Snapshot at pg_catalog_array_agg.snap correctly asserts {1,2,3}.
• No production Rust changes beyond the dep bump — actual fix lives upstream in the datafusion fork, so local surface area is limited.

Quality / Security / Perf

• Code quality: clean, minimal, follows existing test conventions in mod.rs.
• Security: no auth/input-handling surface touched; upstream sqlparser fix only affects identifier parsing.
• Performance: no runtime paths altered.

Suggestions (non-blocking)

• Consider adding one more test covering a schema-qualified scalar (non-aggregate) function to lock down parity, e.g. pg_catalog.upper('x'), since the underlying parser fix likely covers all compound identifiers, not just aggregates. Optional.
• No docs/API changes needed.

Verdict: LGTM.

• Branch: `cubesql/compound-function-fix`