Conversation
_window?.Refresh() parses as a conditional_access_expression, not a member_access_expression, so the invocation branch fell through to the raw-text fallback. That split "_window?.Refresh" on the dot and recorded the receiver as "_window?", which no receiver table types, so the call was dropped with no unresolved entry. Read both shapes through one helper that returns the name and receiver nodes, and use it for the call-site generic-argument walk as well, so recv?.Do<T>() keeps its references edges. The cast and partial-class receivers from Graphify-Labs#3797 are separate and not touched here.
|
Thanks for the pull request, @Cintu07. A maintainer will review it soon. Want to talk it through while it is in review? Come join us on our Discord server. For longer-form discussion there is also GitHub Discussions. A couple of things that speed up review: make sure the test suite passes on Python 3.10 and 3.13, and that the change keeps extraction deterministic. |
![graphify-labs[bot]](https://avatars.githubusercontent.com/in/4356270?s=60&v=4){kind=small}
There was a problem hiding this comment.
Graphify reviewed this change.
Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).
Graphify review — findings
Fixes C# null-conditional calls like _s?.Save() and this._s?.Save() being dropped. A new _csharp_member_call_parts reads the receiver and method name from both member_access_expression and conditional_access_expression, so ?. calls resolve through the receiver-type table like plain . calls. Previously the raw-text fallback typed the receiver as _s?, which matched nothing. The same helper lets r?.Do<T>() emit its generic_arg reference to T.
No blocking issues surfaced. 1 lower-confidence candidate did not survive cross-model review.
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 809 functions depend on the 329 functions this change touches.
Health — this change adds coupling hotspots:
- new:
_extract_generic()— 18 callers, 29 callees - new:
extract_js()— 87 callers, 4 callees - new:
extract_xaml()— 19 callers, 17 callees - new:
extract_objc()— 27 callers, 9 callees - new:
extract_julia()— 19 callers, 7 callees - new:
extract_cpp()— 31 callers, 3 callees - new:
extract_vue()— 10 callers, 7 callees - new:
walk()— 1 callers, 66 callees - …and 9 more — each is listed as a finding
Verification — 809 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 746 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
27 of 312 test file(s) selected (9%) via static blast radius.
tests/test_astro_extraction.py— impacttests/test_build.py— impacttests/test_cjs_module_extension.py— impacttests/test_cpp_nested_and_cli.py— impacttests/test_csharp_call_site_generic_args.py— impact, changed-testtests/test_csharp_member_calls.py— impact, changed-testtests/test_dotnet.py— impacttests/test_extract.py— impacttests/test_extract_php_closures.py— impacttests/test_import_extension_resolution.py— impacttests/test_indirect_call_block_scoped_shadow.py— impacttests/test_indirect_dispatch.py— impacttests/test_indirect_dispatch_assign_return.py— impacttests/test_indirect_dispatch_getattr.py— impacttests/test_js_exported_scalar_bindings.py— impacttests/test_languages.py— impacttests/test_multilang.py— impacttests/test_python_underscore_resolution.py— impacttests/test_rationale.py— impacttests/test_ruby_resolution.py— impacttests/test_scala_self_type.py— impacttests/test_swift_computed_properties.py— impacttests/test_swift_protocol_requirements.py— impacttests/test_trailing_newline_not_a_syntax_error.py— impacttests/test_ts_new_expression_calls.py— impacttests/test_typescript_module_extensions.py— impacttests/test_vue_extraction.py— impact
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
· 17 more finding(s) on lines outside this diff (see the check run).
2 participants
what does this pr do?
fixes the ?. case from #3797. _window?.Refresh() parses as a conditional_access_expression, not a member_access_expression, so the c# invocation branch fell through to the raw text fallback. that split "_window?.Refresh" on the dot and set the receiver to "_window?", which no receiver table can type, so the call was dropped and never parked as unresolved. _window.Refresh() on the same field resolved fine.
added _csharp_member_call_parts, which returns the name and receiver nodes for both shapes, and the existing receiver logic runs on it unchanged. the call-site generic arg walk from #2911 had the same member_access_expression check, so recv?.Do() lost its references edges too. it uses the helper now.
the cast receiver and the partial class call from #3797 are separate fixes and not in here.
type of change
verification & invariants
recv?.M() resolves exactly like recv.M(): same receiver typing, same god-node guard, INFERRED. an untypable receiver still gives no edge, and chained receivers like _h?.Inner?.Save() produce the same output as before. with Window and Panel both declaring Refresh(), _window?.Refresh() lands on Window.Refresh only.
how was this tested?
windows 11 arm64, python 3.14, venv from uv sync --all-extras.
graphify-specific checklist
uv run python -m tools.skillgen --bless) when changing their source fragments.