omena-semantic
Rust crate for the style semantic boundary.
This crate is intentionally small: it consumes the parser boundary from
engine-style-parser and exposes semantic-facing summaries without moving the
parser implementation or changing existing parser consumers.
Current public products:
omena-semantic.style-semantic-graph— combined parser boundary, selector-identity, source-input evidence, promotion evidence, and lossless CST contract for semantic graph consumers.omena-semantic.selector-identity— canonical selector ids, BEM suffix identity, and rewrite-safety blockers.omena-semantic.promotion-evidence— explicit readiness/gap checklist for promotion beyond output parity.omena-semantic.source-input-evidence—EngineInputV2-backed reference site identity, binding origin, style module edge, value-domain explanation, and selector certainty reason evidence.omena-semantic.lossless-cst-contract— byte-span invariants used by precise rename, formatter, and recovery-oriented consumers.omena-semantic.theory-observation-harness— observation-only readiness summary for selector rewrite safety, source evidence explainability, semantic graph downstream readiness, and generic-vs-CME coupling boundaries.
Primary check:
cargo fmt --all --checkcargo testcargo clippy --all-targets --all-features -- -D warnings
Publish status:
publish = falseremains in place untilomena-engine-input-producersandomena-engine-style-parserare available as registry dependencies.
CLI smoke:
|
For EngineInputV2 source-side evidence:
|
For the combined style semantic graph product:
|
For the theory observation harness:
|
For the compact observation contract:
|
Downstream consumers can use either the free function or the
TheoryObservationHarnessInput trait. The trait is the dogfooding surface for
consumers that should depend on an observation contract instead of a concrete
builder entry point. Consumers that only need stable readiness state can call
summarize_theory_observation_contract or the trait method
summarize_theory_observation_contract.