ГЛОБАЛЬНЫЙ РЕФАКТОРИНГ АРХИТЕКТУРЫ - Рефакторинг этапов 3.1 - Добавлен строгий контракт семантического извлечения (валидность, качество, reason-codes, рекомендация применять canonical). Подключен semantic guard в predecompose/runtime: мягкое отклонение плохого canonical: отдельный safe-кейс для сырого полезного raw-фрагмента; fallback теперь выигрывает только когда это оправдано. Подключty semantic arbitration в tool-gate

This commit is contained in:
2026-04-11 16:45:43 +03:00
parent 66402439dc
commit a7ccc92a9c
18 changed files with 750 additions and 26 deletions
@@ -13,6 +13,8 @@ export type AddressPredecomposeAggregationProfile =
| "open_items"
| "unknown";
export type AddressSemanticExtractionQuality = "high" | "medium" | "low";
export interface AddressLlmPredecomposeContractV1 {
schema_version: "address_llm_predecompose_contract_v1";
source_message: string;
@@ -41,6 +43,102 @@ export interface AddressLlmPredecomposeContractV1 {
aggregation_profile: AddressPredecomposeAggregationProfile;
}
export interface AddressSemanticExtractionContractV1 {
schema_version: "address_semantic_extraction_contract_v1";
source_message: string;
canonical_message: string;
canonical_rewrite_applied: boolean;
extraction: {
mode: AddressQuestionMode;
mode_confidence: "high" | "medium" | "low";
query_shape: AddressQueryShape;
query_shape_confidence: "high" | "medium" | "low";
intent: AddressIntent;
intent_confidence: "high" | "medium" | "low";
aggregation_profile: AddressPredecomposeAggregationProfile;
};
entities: AddressLlmPredecomposeContractV1["entities"];
period: AddressLlmPredecomposeContractV1["period"];
guard_hints: {
source_data_signal_detected: boolean;
canonical_data_signal_detected: boolean;
data_scope_meta_query_detected: boolean;
deep_investigation_signal_detected: boolean;
required_anchor_missing: boolean;
unsupported_low_confidence: boolean;
semantic_drift_suspected: boolean;
};
quality: AddressSemanticExtractionQuality;
valid: boolean;
apply_canonical_recommended: boolean;
reason_codes: string[];
}
const ADDRESS_SEMANTIC_DATA_SIGNAL_PATTERN =
/(?:\u0434\u043e\u043a|\u0434\u043e\u0433\u043e\u0432\u043e\u0440|\u043a\u043e\u043d\u0442\u0440\u0430\u0433\u0435\u043d\u0442|\u043a\u043e\u043d\u0442\u0440\u0430\u043a\u0442|\u0441\u0447(?:\u0435|\u0451)\u0442|\u0441\u0430\u043b\u044c\u0434\u043e|\u043e\u0431\u043e\u0440\u043e\u0442|\u043f\u043b\u0430\u0442(?:\u0435|\u0451)\u0436|\u043e\u043f\u0435\u0440\u0430\u0446|\u043f\u0435\u0440\u0438\u043e\u0434|\u0433\u043e\u0434|counterparty|contract|document|account|balance|turnover|operations?|doki|doky|dokument|dogovor|kontragent|schet|saldo|platezh|oplata)/iu;
const ADDRESS_SEMANTIC_ENTITY_SIGNAL_PATTERN =
/(?:\u0437\u0430\u043a\u0430\u0437\u0447\u0438\u043a|\u043f\u043e\u0441\u0442\u0430\u0432\u0449\u0438\u043a|\u043a\u043e\u043d\u0442\u0440\u0430\u0433\u0435\u043d\u0442|\u043a\u043e\u043c\u043f\u0430\u043d|\u043e\u0440\u0433\u0430\u043d\u0438\u0437\u0430\u0446|\u043a\u043e\u043d\u0442\u043e\u0440|customer|supplier|counterparty|company|vendor|client)/iu;
const ADDRESS_SEMANTIC_SCOPE_META_PATTERN =
/(?:\u043a\u0430\u043a\u0430\u044f\s+\u0431\u0430\u0437\u0430|\u0431\u0430\u0437\u0430\s+\u043a\u0430\u043a\u043e\u0439\s+\u043a\u043e\u043d\u0442\u043e\u0440|\u043f\u043e\s+\u043a\u0430\u043a\u0438\u043c\s+\u043a\u043e\u043d\u0442\u043e\u0440|which\s+company\s+base|which\s+tenant|data\s+scope)/iu;
const ADDRESS_SEMANTIC_DEEP_INVESTIGATION_PATTERN =
/(?:\u043f\u0440\u043e\u0432\u0435\u0440(?:\u044c|\u0438\u0442\u044c)|\u0440\u0430\u0437\u0431\u0435\u0440(?:\u0438|\u0430\u0442\u044c)|\u043f\u043e\u0447\u0435\u043c\u0443|\u043c\u0435\u0445\u0430\u043d\u0438\u0437\u043c|\u0440\u0430\u0437\u0440\u044b\u0432|\u0445\u0432\u043e\u0441\u0442|root\s*cause|trace\s*chain|state\s+transition)/iu;
function normalizeCompact(value: unknown): string {
return String(value ?? "")
.toLowerCase()
.replace(/\s+/g, " ")
.trim();
}
function hasSemanticDataSignal(value: unknown): boolean {
return ADDRESS_SEMANTIC_DATA_SIGNAL_PATTERN.test(normalizeCompact(value));
}
function hasSemanticEntitySignal(value: unknown): boolean {
return ADDRESS_SEMANTIC_ENTITY_SIGNAL_PATTERN.test(normalizeCompact(value));
}
function hasSemanticDataScopeMetaSignal(value: unknown): boolean {
return ADDRESS_SEMANTIC_SCOPE_META_PATTERN.test(normalizeCompact(value));
}
function hasSemanticDeepInvestigationSignal(value: unknown): boolean {
return ADDRESS_SEMANTIC_DEEP_INVESTIGATION_PATTERN.test(normalizeCompact(value));
}
function requiredAnchorMissing(contract: AddressLlmPredecomposeContractV1): boolean {
const intent = contract.intent;
if (
intent === "list_documents_by_counterparty" ||
intent === "bank_operations_by_counterparty" ||
intent === "list_contracts_by_counterparty"
) {
return !toNonEmptyString(contract.entities.counterparty);
}
if (intent === "list_documents_by_contract" || intent === "bank_operations_by_contract") {
return !toNonEmptyString(contract.entities.contract);
}
return false;
}
function deriveSemanticQuality(input: {
valid: boolean;
unsupportedLowConfidence: boolean;
requiredAnchorMissing: boolean;
unknownIntentAndShape: boolean;
}): AddressSemanticExtractionQuality {
if (!input.valid) {
return "low";
}
if (input.unsupportedLowConfidence || input.requiredAnchorMissing || input.unknownIntentAndShape) {
return "medium";
}
return "high";
}
function toNonEmptyString(value: unknown): string | null {
if (value === null || value === undefined) {
return null;
@@ -167,3 +265,115 @@ export function buildAddressLlmPredecomposeContractV1(input: {
aggregation_profile: inferAggregationProfile(intent.intent, shape.shape)
};
}
export function buildAddressSemanticExtractionContractV1(input: {
sourceMessage: string;
canonicalMessage: string;
predecomposeContract?: AddressLlmPredecomposeContractV1 | null;
}): AddressSemanticExtractionContractV1 {
const sourceMessage = String(input.sourceMessage ?? "").trim();
const canonicalMessage = String(input.canonicalMessage ?? "").trim() || sourceMessage;
const predecomposeContract =
input.predecomposeContract ??
buildAddressLlmPredecomposeContractV1({
sourceMessage,
canonicalMessage
});
const sourceDataSignal = hasSemanticDataSignal(sourceMessage);
const canonicalDataSignal = hasSemanticDataSignal(canonicalMessage);
const canonicalEntitySignal = hasSemanticEntitySignal(canonicalMessage);
const dataScopeMetaSignal = hasSemanticDataScopeMetaSignal(sourceMessage);
const deepInvestigationSignal =
hasSemanticDeepInvestigationSignal(sourceMessage) || hasSemanticDeepInvestigationSignal(canonicalMessage);
const unsupportedLowConfidence =
predecomposeContract.mode === "unsupported" &&
(predecomposeContract.mode_confidence === "low" || predecomposeContract.mode_confidence === "medium");
const missingRequiredAnchor = requiredAnchorMissing(predecomposeContract);
const unknownIntentAndShape =
predecomposeContract.intent === "unknown" && predecomposeContract.query_shape === "UNKNOWN";
const rewriteApplied = normalizeCompact(sourceMessage) !== normalizeCompact(canonicalMessage);
const semanticDriftSuspected = rewriteApplied && sourceDataSignal && !canonicalDataSignal;
const reasonCodes: string[] = [];
if (dataScopeMetaSignal) {
reasonCodes.push("data_scope_meta_query_detected");
}
if (unsupportedLowConfidence) {
reasonCodes.push("unsupported_low_confidence_contract");
}
if (missingRequiredAnchor) {
reasonCodes.push("required_anchor_missing_for_intent");
}
if (semanticDriftSuspected) {
reasonCodes.push("semantic_drift_source_vs_canonical");
}
if (unknownIntentAndShape && rewriteApplied) {
reasonCodes.push("rewrite_without_structured_gain");
}
if (deepInvestigationSignal) {
reasonCodes.push("deep_investigation_signal_detected");
}
const valid =
!dataScopeMetaSignal &&
!semanticDriftSuspected &&
!(unsupportedLowConfidence && !canonicalDataSignal && !sourceDataSignal && !canonicalEntitySignal) &&
!(unknownIntentAndShape && rewriteApplied && !canonicalDataSignal && !canonicalEntitySignal);
const applyCanonicalRecommended =
valid &&
!(unsupportedLowConfidence && sourceDataSignal && rewriteApplied) &&
!(missingRequiredAnchor && rewriteApplied);
const quality = deriveSemanticQuality({
valid,
unsupportedLowConfidence,
requiredAnchorMissing: missingRequiredAnchor,
unknownIntentAndShape
});
return {
schema_version: "address_semantic_extraction_contract_v1",
source_message: sourceMessage,
canonical_message: canonicalMessage,
canonical_rewrite_applied: rewriteApplied,
extraction: {
mode: predecomposeContract.mode,
mode_confidence: predecomposeContract.mode_confidence,
query_shape: predecomposeContract.query_shape,
query_shape_confidence: predecomposeContract.query_shape_confidence,
intent: predecomposeContract.intent,
intent_confidence: predecomposeContract.intent_confidence,
aggregation_profile: predecomposeContract.aggregation_profile
},
entities: {
account: predecomposeContract.entities.account,
counterparty: predecomposeContract.entities.counterparty,
contract: predecomposeContract.entities.contract,
document_type: predecomposeContract.entities.document_type,
document_ref: predecomposeContract.entities.document_ref,
organization: predecomposeContract.entities.organization
},
period: {
scope: predecomposeContract.period.scope,
period_from: predecomposeContract.period.period_from,
period_to: predecomposeContract.period.period_to,
as_of_date: predecomposeContract.period.as_of_date,
has_explicit_period: predecomposeContract.period.has_explicit_period
},
guard_hints: {
source_data_signal_detected: sourceDataSignal,
canonical_data_signal_detected: canonicalDataSignal,
data_scope_meta_query_detected: dataScopeMetaSignal,
deep_investigation_signal_detected: deepInvestigationSignal,
required_anchor_missing: missingRequiredAnchor,
unsupported_low_confidence: unsupportedLowConfidence,
semantic_drift_suspected: semanticDriftSuspected
},
quality,
valid,
apply_canonical_recommended: applyCanonicalRecommended,
reason_codes: reasonCodes
};
}