[xiongjianxu]

Summary

Closes #4668.

A cross-repo calls edge pointed source_file:source_location into the wrong file. source_file came from the caller node (an ObjC method node lives at its @interface in the .h) while source_location is the call-site line (in the .m), so the pair landed past the end of a short header.

• _park_unresolved_member_call now records the call-site file next to line.
• The extract-time source_file relativization pass relativizes the parked file too, so it never leaks the absolute checkout path into graph.json (consistent with the parking payload carrying names only, ObjC: #2591 self.<field> receiver typing emits zero edges via the CLI — objc_field_types keys are skipped by the #1529 id remap #3150).
• link_cross_repo_member_calls prefers the parked file, falling back to the caller node's source_file for graphs built before the field existed.

Testing

• New tests/test_parked_call_site_file.py: the parked file is the repo-relative call-site .m, never absolute.
• New cross-repo edge cases: the edge source_file is the call site, and a pre-merge-graphs / global add can never produce cross-repo member-call edges: the resolved receiver type is discarded at bail-out instead of persisted #3152 entry without file falls back to the caller's file.
• Updated the Go/PHP/TS parked-payload assertions to tolerate the new additive file key (its exact value depends on a root those unit cases extract without).
• pytest -k "objc or cross_repo or parked or receiver" green apart from 7 failures in test_ext_tagged_id_clash.py that are pre-existing on v8 (File node takes the label of a module-level variable when a same-stem sibling file exists (x.mjs + x.sql, const sql) #4449), unrelated to this change. ruff clean.

Based on v8.

[github-actions]

Thanks for the pull request, @xiongjianxu. 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]

Graphify reviewed this change.

No issues found by static checks — no coupling regressions or blocking issues in the code graph. That is not the same as safe to merge: see what was not checked.

Not checked: tests were not run; formal verification proved 0 of 3 changed function(s) (3 sampled, not proven).

Formal verification. PR-changed functions: 3/3 verified (0 proven, 3 may-equivalent, 0 distinguished) · 0 not verified.

---

Graphify review — findings

Fixes the source file on cross-repo member-call edges: they now use the file where the call actually happens (e.g. the .m) instead of the caller node's own file (often the .h declaring the method). Previously the call-site line number was paired with the wrong file, so it could point past the end of an unrelated file. _park_unresolved_member_call now records the call-site file on each parked entry, and extract relativizes it like source_file, so absolute checkout paths don't leak into graph.json or the merged edge. Graphs parked before this change have no file, so link_cross_repo_member_calls falls back to the caller's source_file, which is the old behavior. Adds tests for the new edge source_file, the pre-#3152 fallback, and relative-path parking on ObjC. Existing parked-call assertions in the Go, PHP and TS tests now ignore file, since they check shape rather than location.

No blocking issues surfaced. 1 lower-confidence candidate did not survive cross-model review.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 3244 functions depend on the 475 functions this change touches.

Health — this change adds coupling hotspots:

• worse: extract() — 989 callers, 59 callees
• worse: link_cross_repo_member_calls() — 34 callers, 10 callees

Verification — 3244 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: 2998 function(s) in the blast radius were not formally verified this run

Test selection

Test selection

206 of 437 test file(s) selected (47%) via static blast radius.

• tests/test_affected_overrides.py — impact
• tests/test_astro_extraction.py — impact
• tests/test_astro_import_ids.py — impact
• tests/test_blade_extractor.py — impact
• tests/test_build.py — impact
• tests/test_builtin_global_type_refs.py — impact
• tests/test_c_include_resolution.py — impact
• tests/test_cache.py — impact
• tests/test_case_sensitive_resolution.py — impact
• tests/test_cjs_module_extension.py — impact
• tests/test_cobol_extractor.py — impact
• tests/test_cpp_method_declarations.py — impact
• tests/test_cpp_nested_and_cli.py — impact
• tests/test_cpp_objc_cross_file_calls.py — impact
• tests/test_cpp_template_type_params.py — impact
• tests/test_cross_extension_reexport_self_cycle.py — impact
• tests/test_cross_language_call_resolution.py — impact
• tests/test_cross_repo_external_call_guards.py — impact
• tests/test_cross_repo_member_calls.py — impact, changed-test
• tests/test_csharp_call_site_generic_args.py — impact
• tests/test_csharp_delegate_declarations.py — impact
• tests/test_csharp_enum_members.py — impact
• tests/test_csharp_field_generic_args.py — impact
• tests/test_csharp_foreach_loop_variable.py — impact
• tests/test_csharp_generic_callsites.py — impact
• tests/test_csharp_interface_dispatch.py — impact
• tests/test_csharp_member_calls.py — impact
• tests/test_csharp_member_chain_calls.py — impact
• tests/test_csharp_member_nodes.py — impact
• tests/test_csharp_object_creation.py — impact
• tests/test_csharp_partial_classes.py — impact
• tests/test_csharp_tuple_type_refs.py — impact
• tests/test_csharp_type_resolution.py — impact
• tests/test_csharp_underscore_collisions.py — impact
• tests/test_csharp_var_call_return_type.py — impact
• tests/test_csharp_verbatim_interpolation.py — impact
• tests/test_definition_file_portability.py — impact
• tests/test_detect.py — impact
• tests/test_dotnet.py — impact
• tests/test_duplicate_annotation_edges.py — impact
• tests/test_ejs_extractor.py — impact
• tests/test_elixir_defdelegate.py — impact
• tests/test_elixir_import_resolution.py — impact
• tests/test_elixir_keyword_def_calls.py — impact
• tests/test_elixir_qualified_calls.py — impact
• tests/test_elixir_unqualified_call_scope.py — impact
• tests/test_erlang_extractor.py — impact
• tests/test_export_direction.py — impact
• tests/test_ext_tagged_id_clash.py — impact
• tests/test_extract.py — impact
• … and 156 more

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.

Formal verification

No difference found (not proven): No behavior difference found in link\_cross\_repo\_member\_calls (not a proof).

The verifier ran both versions of link\_cross\_repo\_member\_calls on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

No difference found (not proven): No behavior difference found in \_park\_unresolved\_member\_call (not a proof).

The verifier ran both versions of \_park\_unresolved\_member\_call on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

No difference found (not proven): No behavior difference found in extract (not a proof).

The verifier ran both versions of extract on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

· 2 more finding(s) on lines outside this diff (see the check run).