journal: ct.3 codegen + mono cleanup

2026-05-11 09:40:26 +02:00
parent 666c78468f
commit 4530548bb3
1 changed files with 73 additions and 0 deletions
@@ -14307,3 +14307,76 @@ tests + the existing workspace test surface cover the
 invariants; the canonical E2E demonstration of the
 iter-23.3-Task-4 bug closure happens in ct.4 with the recreated
 `compare_primitives_smoke.ail.json` fixture).
+
+## 2026-05-11 — Iteration ct.3: codegen + mono cleanup
+
+Three tasks landed, plus one orchestrator-level tidy from
+quality-review feedback:
+
+- **ct.3.1** (`a8a58cc`, tidy `05d0cce`): deleted the
+  `lookup_ctor_by_type` imports-walk fallback in
+  `crates/ailang-codegen/src/lib.rs`. Two-branch direct lookup
+  remains: qualified via `import_map` →
+  `module_ctor_index[target_module]` (with `cref.type_name ==
+  suffix` check); bare via `module_ctor_index[self.module_name]`
+  (with `cref.type_name == type_name` check). Both hard-error on
+  miss. Symmetric to ct.2.3's `Term::Ctor` synth fix on the
+  typecheck side. The lookup function's docstring and the
+  surrounding lockstep-rationale comment were refreshed; the
+  small `-tidy` followup dropped a dated "ct.3 Task 1:" comment
+  prefix flagged by the quality reviewer.
+- **ct.3.2** (`9149801`): narrowed
+  `apply_per_module_ctor_index_overlay` to `env.types` only and
+  renamed it to `apply_per_module_types_overlay`. The
+  `env.ctor_index` half of the original 22c overlay became
+  decorative after ct.2.2 stopped consulting the index from
+  Pattern::Ctor; the `env.types` half is still load-bearing for
+  Term::Ctor bare-name lookup in mono's re-run synth. Both call
+  sites updated, the rationale comment rewritten to drop the
+  obsolete Pattern::Ctor framing, the now-unused `CtorRef` import
+  dropped, and the dangling rustdoc link in
+  `build_workspace_env`'s docstring re-pointed to the new
+  function name.
+- **ct.3.3** (`666c784`): retired the obsolete "parity with
+  `ambiguous-type` diagnostic" paragraph in the
+  `MigrateCanonicalTypes` doc-comment
+  (`crates/ail/src/main.rs`). The diagnostic it referenced was
+  removed in ct.2.3.
+
+All four obsolete mechanisms named in the ct.1 JOURNAL entry are
+now gone:
+- Pattern::Ctor imports-fallback (ct.2.2).
+- Term::Ctor synth imports-fallback (ct.2.3).
+- codegen `lookup_ctor_by_type` imports-fallback (ct.3.1).
+- mono `apply_per_module_ctor_index_overlay`'s `env.ctor_index`
+  half (ct.3.2 narrowing).
+
+Only ct.4 remains in the canonical-type-names milestone:
+DESIGN.md amendments to Decision 2, hash-pinned regression test
+re-baseline for the two migrated cross-module fixtures, prose
+round-trip extension covering the qualifier-trim-on-print
+behaviour, and the optional
+`examples/compare_primitives_smoke.ail.json` recreation that
+demonstrates end-to-end closure of the iter-23.3-Task-4 bug
+through the full pipeline (parse → typecheck → mono → codegen
+→ run).
+
+Out of scope this iter, deliberately:
+- Codegen `lookup_ctor_in_pattern` (lib.rs:1790) — uses
+  imports-walk but has no scrutinee type to type-anchor; would
+  require plumbing Type through pattern lowering, a structurally
+  bigger refactor than ct.3's scope.
+- The codegen `lower_workspace_inner` implicit-prelude
+  `import_map` entry (iter 23.2.4 fix `be882c4`) — different
+  problem (fn-name resolution for mono symbols, not Type::Con
+  names).
+- `check_in_workspace`'s analogous per-module overlay in
+  `crates/ailang-check/src/lib.rs:1234` — its `env.ctor_index`
+  half still serves the duplicate-detection diagnostic at
+  workspace-build time, not the runtime ctor lookup; narrowing
+  it would be a separate concern with its own correctness
+  rationale.
+
+Workspace at iter-ct.3 close: full `cargo build --workspace` and
+`cargo test --workspace` green (451 tests, 0 failed). Test count
+unchanged from ct.2 — this iter is pure code deletion + doc tidy.