Repository navigation
#9911's fixture_closure_rustc_discrimination passes on ANY rustc refusal from its red fixture's module, so its text-boundary claim is weaker than it reads - #10084
Conversation
…no pass `fixture_closure_rustc_discrimination` accepted its red arm on ANY rustc refusal reported in the red fixture's own emitted module. A location says WHERE rustc refused and never WHAT, so a fixture edited into a syntax error, an emitter regression producing an unresolved path, or an arity fault would each have reddened the arm in the right file and been reported PASSED while the text-boundary class the arm names had stopped being its subject. The attributed diagnostic now carries its governing rustc header, and the pair additionally requires the error class the arm claims -- E0308, cited from the rustc error index and MEASURED on the pre-change tree with the arm's own enrolled invocation rather than inferred from the class description. Only an `error` may attribute: a warning naming the file refused nothing, and an error header elsewhere in the run does not govern it. The scan is factored out of `run_cargo` as a pure function so the parse the red arm rests on executes on every push through three unit tests, while the cargo route around it stays `#[ignore]`d candidate evidence. No rung changes. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Frq65YNWcMzSVcnAoHCQuB
… red The two `compiler_tests_rust` mirrors are regenerated by the emitter (`claim_executor --required-regen`, installed and taken to a fixed point: the mirror re-emits identical and only `compiler_tests.rs` drifted after the seed rebuild), so the committed bytes carry the strengthened assertion. `a_red_of_the_wrong_error_class_does_not_pass_the_pair` is the discriminating RED for the conjunct itself: same pair shape, only the rustc code varies, and `E0433` -- an unresolved path, which is what an emitter regression or an edited fixture produces -- FAILS where the arm previously reported PASSED. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Frq65YNWcMzSVcnAoHCQuB
Measured across the two runs taken on this lane: the same E0308 arrived once as `mismatched types` and once as `arguments to this function are incorrect`. Matching the wording would have reddened the arm on the second run with nothing about the emitter changed. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Frq65YNWcMzSVcnAoHCQuB
# Conflicts: # src/v1/stage0/src/compiler_tests.rs # src/v1/stage0/src/v1_compiler_compiler_tests_rust.rs
# Conflicts: # src/v1/stage0/src/cli_run.rs # src/v1/stage0/src/compiler_tests.rs # src/v1/stage0/src/v1_compiler_compiler_tests_rust.rs
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 2514ca4e7c
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| && fixture_closure_attributed_diagnostic(&pair.red) | ||
| .is_some_and(|diagnostic| diagnostic.contains(FIXTURE_RED_EXPECTED_RUSTC_CODE)) |
There was a problem hiding this comment.
fixture_discrimination_passed is also called by function_value_adapter_fixture_closure_discrimination, whose red fixture intentionally produces E0277 and explicitly verifies that code in compiler_tests.rs. Hard-coding E0308 here therefore makes that opt-in test fail deterministically at its subsequent predicate assertion even when both adapter fixtures behave exactly as expected; the expected code needs to be supplied per discrimination pair, or this class check must remain specific to the text-boundary pair.
Useful? React with 👍 / 👎.
1 participant
Auto-opened by session-dashboard for session
swift-swift-486.Pushing to
session/swift-swift-486advances this PR.Worker attestation
Before flipping this PR to ready for review, confirm each item:
npm test,cargo test) and the result.Closes #Ndirective.Summary
TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.
Test plan