Fix #2738: Prevent all-punctuation identifiers from collapsing in make_id

[shard-c6]

Fixes #2738

Summary

This PR addresses the issue where private, all-punctuation identifiers (like Dart's const Foo._()) collapse during ID normalization and overwrite their parent file node's label, which subsequently breaks _is_file_node() detection and causes file nodes to be incorrectly classified as god nodes.

Changes

1. Primary Fix: Updated make_id() in graphify/ids.py. When an entity is given but normalizes to an empty string (i.e. it only contains punctuation like _), make_id() now explicitly falls back to emitting "underscore" instead of silently collapsing. This prevents the symbol's ID from colliding with the file's ID.
2. Secondary Fix: Updated node merging logic in build_from_json() inside graphify/build.py. When G.add_node() merges attributes for an existing node, it now preferentially keeps the label that matches the source filename instead of allowing arbitrary overlapping symbols to overwrite it.

[graphify-labs]

Graphify reviewed this change.

Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.

Formal verification. 1 change(s) tested, no difference found (not proven).

---

Graphify review — findings

This PR modifies node-handling logic in two files. In graphify/build.py, the node-insertion path in build_from_json is changed so that when a node ID already exists in the graph with a conflicting label, it conditionally drops the incoming label attribute based on whether the existing/new labels are considered file-node labels for the associated source file. In graphify/ids.py, the make_id helper is reworked to normalize each part individually (substituting "underscore" for parts that normalize to empty) before joining and re-normalizing. The surface area is the graph-building node merge behavior and ID-generation logic, which are shared by many of the listed build/id symbols.

Worth a look

• make_id changes normalization semantics: per-part normalize plus 'underscore' placeholder — graphify/ids.py:68 · Escalate · medium
  • agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review Execution auto-disposal is off for this run; enable it (with sandbox isolation) to have Graphify try to confirm or refute this automatically.
• make_id no longer strips leading/trailing '_.' from parts — graphify/ids.py:68 · Escalate · medium
  • agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review Execution auto-disposal is off for this run; enable it (with sandbox isolation) to have Graphify try to confirm or refute this automatically.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 2515 functions depend on the 87 functions this change touches.

Health — this change adds coupling hotspots:

• worse: build_from_json() — 151 callers, 19 callees

Verification — 2515 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: 1455 function(s) in the blast radius were not formally verified this run

Formal verification

Could not verify: Could not verify build\_from\_json.

The verifier did not have enough to check build\_from\_json, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: not verifiable: all 6 sampled inputs raised on both versions — the function never executed, so 'no divergence' would be vacuous (mostly NameError — names the real obstacle, not a sampling gap)

No difference found (not proven): No behavior difference found in make\_id (not a proof).

The verifier ran both versions of make\_id 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.

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