cl: conf.NonClass support globalFuncToMethod; tool: Config doc; cl: methodCheck.matchList

[xushiwei]

• Proposal: MethodCheck — support Logical-Type Method Detection in llcppg #964

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 88.10%. Comparing base (690938b) to head (a5dfb77).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #963      +/-   ##
==========================================
+ Coverage   88.09%   88.10%   +0.01%     
==========================================
  Files          23       23              
  Lines        2687     2691       +4     
==========================================
+ Hits         2367     2371       +4     
  Misses        320      320              

Flag	Coverage Δ
llgo-tests	88.10% <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.

[fennoai]

Review summary

Clean, well-scoped change. Extracting nonClass() is a solid DRY improvement, the refactored tryToMethod() preserves the prior 4-tuple contract and nil-safety (tn.Pkg() may be nil for predeclared types but is only compared, never dereferenced), and the per-field docs on tool.Config are a real readability win. The getCString -> clang_getCString Rename-key fix is correct and matches the regenerated fixtures (CXString/out.go, include/clang-c/CXString.h). No security or performance concerns — contains() is a tiny O(n) scan on an offline code-gen path.

A couple of minor documentation nits are inline. Two additional notes that don't map to changed lines:

• cl/ctx.go:132 — the internal field comment still reads // typedef names to be treated as non-classes. This PR deliberately gives NonClass a dual meaning and documents it in both Config structs, but conf.NonClass flows straight into ctx.nonClasses (cl/compile.go:274) and is now also consumed with the Go-type-name meaning at cl/func.go:333. Worth updating this comment so the two meanings stay in sync.
• (Optional) cl/func.go:308-323 — the *types.Alias branch keeps an extra tn.Obj().Pkg() == pkgTypes check on the unaliased type that the *types.Named branch doesn't need (this is pre-existing). A one-line comment on why the alias path needs that underlying-package check would help a future reader avoid "simplifying" it away.