fix(transform): #764 — module-init State() no longer breaks stateOnChange (v0.5.910)

.github/workflows/test.yml

@@ -350,6 +350,8 @@ jobs:
             test_issue_610_foreach \
             test_issue_610_smoke \
             test_issue_640_navstack_textfield \
+            test_issue_763_reactive_textfield \
+            test_issue_764_state_at_module_init \
             test_parity_assert \
             test_parity_async_hooks \
             test_parity_buffer \

CHANGELOG.md

@@ -2,6 +2,8 @@
 
 Detailed changelog for Perry. See CLAUDE.md for concise summaries.
 
+## v0.5.910 — fix(transform): #764 — `State()` at module-init level no longer breaks `stateOnChange` / `stateBindTextfield` / `stateBindToggle`. **Symptom (reported as #763).** `const cell = State(""); stateOnChange(cell, cb); cell.set("v")` at module top level (NOT inside `function main()`) updated `cell.value` correctly but never fired `cb`. The same shape inside a function worked. The HIR-level `try_desugar_reactive_text` pass (which lowers `Text(\`...${state.value}...\`)` into an IIFE that registers `stateOnChange`) inherits the same failure when its host module declares the state at module-init level, because the desugar emits exactly the kind of `stateOnChange(cell, ...)` call that gets stranded. **Root cause.** `crates/perry-transform/src/state_desugar.rs::collect_state_bindings` (issue #535, extended in #612 to also catch capital-`S` `State`) was rewriting *every* module-init-level `State(initial)` / `state(initial)` declaration into the synthetic keyed runtime: `let cell = undefined` + `__state_init("__state_N", initial)`, with downstream `cell.set(v)` → `__state_set("__state_N", v)` and `cell.value` → `__state_get("__state_N")`. The two-state-systems-in-one-namespace problem: the synthetic `__state_*` runtime is registry/key-keyed (one HashMap keyed by string id), the public handle-based API (`perry_ui_state_create` → integer 1-based handle → `perry_ui_state_set` → `perry_ui_state_on_change`) is *separate*. After the rewrite, `stateOnChange(cell, cb)` saw `cell = undefined` (the `Let { init: Undefined }` placeholder the desugar leaves behind to keep the LocalId addressable), `perry_ui_state_on_change(handle: i64, callback: f64)` extracted garbage as its handle, and the subsequent `__state_set("__state_N", v)` wrote to a completely different store — so the handle-based subscriber registry stayed empty and the callback never fired. The reactive-Text desugar's `textSetString(__h, fresh_concat)` thus only ran when something else triggered it (`stateBindTextfield` two-way binding wakes the textfield up at type-time, but nothing wakes up the standalone `Text`). **Fix.** Gate `state_desugar` so it only rewrites a binding when *no* handle-based state API consumes it. After `collect_state_bindings`, a new `collect_handle_based_state_uses(module)` walks all of `module.init`, every function body, and every class method/ctor body looking for `NativeMethodCall { module: "perry/ui", method: "stateOnChange" | "stateBindTextfield" | "stateBindToggle" | "stateBindSlider" | "stateBindVisibility" | "stateBindTextNumeric", args[0]: LocalGet(id) }`. Any `id` it finds is removed from the rewrite map — the binding stays as the original `State(...)` constructor call, the handle-based runtime owns it end-to-end, and `stateOnChange` / friends register against the real integer handle. The `set`/`value`/`get`/`text` / `NavStack(state, routes)` / `ForEach(state, render)` rewrites are unaffected (they already had complete keyed-side implementations) and lowercase `state(...)` still works exactly as #535 designed when the user goes purely keyed-API. **Why not just remove the uppercase-`State` branch from `collect_state_bindings`?** #612's NavStack(state, routes) flow specifically needs the synth_id machinery for the `__navstack_register_route` call, and `state` and `State` are both documented as the constructor for that flow. Pulling uppercase out would re-break #612. The gate restores the v0.5.112+ handle-based contract for `State` without disturbing #535 / #612. **Validation.** (1) Four new unit tests in `crates/perry-transform/src/state_desugar.rs::tests` lock in the gate: module-init uppercase + `stateOnChange` is skipped; module-init lowercase + `stateBindTextfield` is skipped; module-init State with *only* `.set` is still rewritten (NavStack / ForEach path unaffected); a handle use inside a function body still gates the module-init binding (matches the #763 shape where `State()` is at module-init and `stateOnChange` is inside `main()`). (2) New integration fixture `test-files/test_issue_763_reactive_textfield.ts` exercises the original repro under AppleScript automation on macOS — typing into the TextField and clicking the "set hello world" button both reactively update the standalone `Text(\`current state for text: ${text.value}\`)` widget (verified via System Events' static-text enumeration). (3) NavStack regression: `test-files/test_issue_640_navstack_textfield.ts` (the original #612 repro) still routes from A→B correctly. `cargo test --release -p perry-transform` green (23/0/0); workspace test suite (excluding cross-host UI crates per CLAUDE.md) green. **Issue #764 context.** Originally filed to track auto-binding template literals over state cells to `perry_ui_state_bind_text_template` — but that auto-bind has actually shipped since v0.5.112 (HIR `try_desugar_reactive_text`); it emits an IIFE + `stateOnChange` + `textSetString` rather than calling the template-binding FFI directly, which works correctly when state is in function scope. The user-reported repro in #763 was actually exposing this separate module-init regression; this commit fixes that path. The "use `perry_ui_state_bind_text_template` directly" optimization is left as a future architectural improvement (it would centralize re-render in the runtime instead of one closure-per-state on the JS side).
+
 ## v0.5.909 — feat: add `libDirs` to native library manifest (#762). New optional `targets.<target>.libDirs` field in `package.json`'s `perry.nativeLibrary` block — array of linker search paths emitted ahead of the `libs` list. **Surface.** Documented in `docs/src/native-libraries/manifest-v1.md` alongside the existing `libs` / `frameworks` / `pkgConfig` entries. Use it when a wrapper ships a vendored `.a`/`.dylib` outside the cargo crate (the standard `lib_name` lookup probes `target/<triple>/release/`); previously the only escape hatch was an absolute path glued into `libs` via a workaround. **Plumbing.** Three edits — (1) `TargetNativeConfig.lib_dirs: Vec<PathBuf>` field added in `crates/perry/src/commands/compile.rs`; (2) `parse_native_library_manifest` parses `libDirs` and anchors relative entries to `package_dir` via `package_dir.join(p)` — mirrors the existing `swift_sources` / `metal_sources` pattern so a `"./vendor/lib"` in the manifest resolves against the wrapper's directory, not the user's cwd (absolute entries pass through unchanged since `PathBuf::join` ignores the base on an absolute right-hand side); (3) `build_and_run_link` in `crates/perry/src/commands/compile/link.rs` emits each entry between the `frameworks` loop and the `libs` loop — `-L<dir>` on every non-Windows target, `/LIBPATH:<dir>` when `is_windows` (mirrors the `libs` loop's existing `{lib}.lib` vs `-l{lib}` MSVC/Unix split, so a `targets.windows.libDirs` entry actually resolves the `{lib}.lib` lookups instead of being a silent no-op against link.exe). **Validation.** Two new unit tests in `crates/perry/src/commands/compile/resolve.rs::manifest_parse_tests` — `lib_dirs_relative_paths_anchored_to_package_dir` round-trips a manifest with both `"vendor/lib"` and `"/abs/path"` and asserts the package_dir-anchored / pass-through behavior; `lib_dirs_defaults_to_empty_when_absent` confirms the field is optional. `cargo build --release` clean; `cargo test -p perry --lib` green. **Version-bump note.** Renumbered v0.5.908 → v0.5.909 because #760 (test: argon2 + ethers parity) landed on main mid-review and took the v0.5.908 slot. **Contributor.** Implementation by @Lebei2046; MSVC `/LIBPATH:` branch + `package_dir` anchoring + tests + manifest doc entry + this changelog folded in at merge time.
 
 ## v0.5.908 — test: #698 — behavioral parity tests for argon2 + ethers integrations (#760). Follow-up to #694, addressing #698. Adds two new parity fixtures and a `@covers` block on `test_parity_crypto.ts`, moving 18 FFI entries out of `test_ffi_surface_stdlib_integrations.ts` (156 → 138 unique names) into behavioral coverage. **`test_parity_argon2.ts`** — round-trips `argon2.hash` / `argon2.verify` against the `perry-ext-argon2` wrapper; non-deterministic salt is handled by shape-checking the `$argon2id$` prefix and asserting verify round-trips against a freshly produced hash. Only the async path is in `NativeModSig`, so this covers the two FFI entries (`js_argon2_hash`, `js_argon2_verify`) reachable end-to-end. **`test_parity_ethers.ts`** — pure deterministic helpers from `perry-ext-ethers`: `getAddress` (EIP-55 checksum on lower- and upper-case input), `parseEther`/`formatEther` round-trip on 1.5 ETH, `parseUnits`/`formatUnits` at 6- (USDC) and 9-decimal (gwei) positions. Covers the five `js_ethers_*` helpers. Both new fixtures use the `test-parity/expected/` mechanism — Perry routes the npm-style imports to the bundled `perry-ext-*` wrappers, but Node can't resolve the same names without `node_modules`, so they fall through to stored expected-output comparison. **`test_parity_crypto.ts`** — adds a `@covers` block declaring the crypto/webcrypto/crypto_e2e FFI surface it already exercises via `node:crypto` and `crypto.subtle` (digest, hash, pbkdf2/hkdf, subtle.sign/verify); no behavioral change. **Out of scope (each gets its own follow-up).** `jsonwebtoken.sign` — `NativeModSig` passes `payload`/`secret` as `NA_F64` (NaN-boxed) but `js_jwt_sign` reads them as `*const StringHeader`, producing garbage tokens (calling-convention mismatch, same shape as the #591 argon2 fix). `bcrypt.hash` — return value reports `typeof === "object"` in user code; `compare` round-trips correctly but `.startsWith(...)` on the hash fails. `cheerio` — `$.select(...)` returns a bare-number handle, so `.length()` / `.first()` etc. fail with `(number).method is not a function` at the prototype check. Service-backed integrations (MongoDB, MySQL, PG, ioredis, Nodemailer, Sharp) also remain in the inventory pending the local-container / opt-in test setup described in #698. **Audit.** `./test-coverage/audit.sh --markdown` keeps TS-side and combined FFI coverage at **1791/1791 (100.0%)**. `test_ffi_surface_stdlib_integrations.ts` re-regenerated via `regen_ts_surface_inventory.py` to fold in the new `@covers` blocks. Comment-only changes; no behavioral or runtime impact. Contributor: @TheHypnoo.

CLAUDE.md

@@ -8,7 +8,7 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
 
 Perry is a native TypeScript compiler written in Rust that compiles TypeScript source code directly to native executables. It uses SWC for TypeScript parsing and LLVM for code generation.
 
-**Current Version:** 0.5.909
+**Current Version:** 0.5.910
 
 
 ## TypeScript Parity Status