NormalizerOptions

Defined in: packages/engine/src/normalizer/TokenNormalizer.ts:49

Configuration options for the normalization pass.

These control safety limits and diagnostic callbacks. The defaults are chosen to be generous enough for any realistic expression while preventing runaway token expansion from recursive rules.

Properties

ignoreRuleIndex?

optional ignoreRuleIndex?: boolean;

Defined in: packages/engine/src/normalizer/TokenNormalizer.ts:86

Try every registered rule at every position, ignoring both the shape index and the first-token buckets the first call uses.

Diagnostic only, and much slower. It exists so the filtered walks can be compared against the exhaustive one over a corpus: each filter is pure, so all of them must agree token for token, and a rule whose declared NormalizerRule.shape is too narrow shows up as a difference rather than as a feature that quietly stopped working. NormalizerIndexFidelity.spec and RuleHintsMatchGuards.spec are the consumers.

Default

false

---

maxPasses?

optional maxPasses?: number;

Defined in: packages/engine/src/normalizer/TokenNormalizer.ts:55

Maximum number of full passes over the token stream before bailing out. Prevents infinite loops from recursive rule chains.

Default

100

---

maxTokens?

optional maxTokens?: number;

Defined in: packages/engine/src/normalizer/TokenNormalizer.ts:63

Maximum number of tokens allowed after normalization. If exceeded, an Error is thrown rather than passing a bloated stream to the parser.

Default

10000

---

onFusion?

optional onFusion?: (fusion) => void;

Defined in: packages/engine/src/normalizer/TokenNormalizer.ts:70

Callback invoked for each fusion event during normalization. Used by diagnostic mode to populate NormalizerOutput.fusions. When undefined, fusions are still tracked internally but no callbacks fire.

Parameters

Parameter	Type
fusion	TokenFusion

Returns

void