* Split volatile library identity code
- Emit build identity getters from a tiny standalone C translation unit
- Cache generated library program objects independently of volatile identity and archives
- Preserve toolchain, runtime, and translation-unit dependency invalidation coverage
* Preserve library identity during localization
* Keep library identity sources private
* Preserve library identity in public emission
* Preserve identity in kept library units
* Reuse IR across comment-only library edits
- Restore validated lowered IR when TypeScript edits change only non-semantic comments
- Rebase source locations and regenerate exact-source sidecar and build identities
- Add semantic invalidation, integrity, and real-cache coverage
* Fix semantic cache lexical validation
* Use canonical IR version in semantic cache
* Fix semantic cache lexical rebasing
* Fix semantic cache Unicode identifiers
* feat: support wasm32-wasi as a production target
- Add production LLVM WASI codegen with a 32-bit ABI, coroutines, and dynamic modules
- Integrate WASI builds and execution into the CLI with capability diagnostics
- Add differential coverage and document the fourth production target
Co-authored-by: Albert <57402359+bulatovv@users.noreply.github.com>
* fix: address wasm target review findings
* fix WASI guest paths and fetch alias detection
* fix WASI async bridge and fetch detection
---------
Co-authored-by: Albert <57402359+bulatovv@users.noreply.github.com>
- Add a strict versioned manifest and direct C ABI lowering for both backends.
- Link declared native inputs through the CLI and compiler APIs with precise diagnostics.
- Cover the ABI under plain and sanitized integration tests and document its contracts.
Chris Tate