Этап 4 / Волна 19.1 выравнивание живого контура для слоя адресного сбора доказательной базы

2026-03-29 08:33:54 +03:00
parent 60d8b96a14
commit d461cedf35
19 changed files with 2389 additions and 66 deletions
@@ -82,6 +82,83 @@ function extractExecutionState(normalized) {
        };
    });
 }
+function collectBusinessScopesFromNormalized(normalized) {
+    const scopes = [];
+    for (const item of extractFragments(normalized)) {
+        if (!item || typeof item !== "object") {
+            continue;
+        }
+        const scope = String(item.business_scope ?? "").trim();
+        if (scope) {
+            scopes.push(scope);
+        }
+    }
+    return Array.from(new Set(scopes));
+}
+function hasJuly2020SnapshotSignal(userMessage, companyAnchors) {
+    const lower = String(userMessage ?? "").toLowerCase();
+    if (/(?:\b2020[-/.]0?7\b|\bиюл[ьяе]?\b(?:\s+20\d{2})?|\bjuly\b(?:\s+20\d{2})?)/i.test(lower)) {
+        return true;
+    }
+    const periods = Array.isArray(companyAnchors?.periods) ? companyAnchors.periods : [];
+    const dates = Array.isArray(companyAnchors?.dates) ? companyAnchors.dates : [];
+    return [...periods, ...dates].some((item) => /2020[-/.]0?7|июл|july/i.test(String(item ?? "").toLowerCase()));
+}
+function hasP0DomainSignal(userMessage, companyAnchors) {
+    if (inferP0DomainFromMessage(userMessage)) {
+        return true;
+    }
+    const accounts = Array.isArray(companyAnchors?.accounts) ? companyAnchors.accounts : [];
+    if (accounts.some((item) => /^(?:01|02|08|19|20|21|23|25|26|28|29|44|51|60|62|68|76|97)(?:\.|$)/.test(String(item ?? "").trim()))) {
+        return true;
+    }
+    return /(?:ндс|vat|рбп|deferred|амортиз|supplier|customer|settlement|month\s*close|закрыти[ея]\s+месяц|поставщ|покупат)/i.test(String(userMessage ?? "").toLowerCase());
+}
+function resolveBusinessScopeAlignment(input) {
+    const rawScopes = collectBusinessScopesFromNormalized(input.normalized);
+    const needsCompanyGrounding = hasJuly2020SnapshotSignal(input.userMessage, input.companyAnchors) && hasP0DomainSignal(input.userMessage, input.companyAnchors);
+    const reasons = [];
+    if (needsCompanyGrounding) {
+        reasons.push("july_2020_snapshot_p0_signal");
+    }
+    if (!input.routeSummary || input.routeSummary.mode !== "deterministic_v2" || !needsCompanyGrounding) {
+        return {
+            business_scope_raw: rawScopes,
+            business_scope_resolved: rawScopes,
+            company_grounding_applied: false,
+            scope_resolution_reason: reasons,
+            route_summary_resolved: input.routeSummary
+        };
+    }
+    let changed = false;
+    const decisions = input.routeSummary.decisions.map((decision) => {
+        const scopeValue = String(decision.business_scope ?? "").trim();
+        if (scopeValue !== "generic_accounting" && scopeValue !== "unclear") {
+            return decision;
+        }
+        changed = true;
+        return {
+            ...decision,
+            business_scope: "company_specific_accounting"
+        };
+    });
+    const resolvedSummary = changed
+        ? {
+            ...input.routeSummary,
+            decisions
+        }
+        : input.routeSummary;
+    const resolvedScopes = changed
+        ? Array.from(new Set(decisions.map((decision) => String(decision.business_scope ?? "").trim()).filter(Boolean)))
+        : rawScopes;
+    return {
+        business_scope_raw: rawScopes,
+        business_scope_resolved: resolvedScopes,
+        company_grounding_applied: changed,
+        scope_resolution_reason: changed ? [...reasons, "generic_or_unclear_to_company_specific_override"] : reasons,
+        route_summary_resolved: resolvedSummary
+    };
+}
 function escapeRegex(value) {
    return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
 }
@@ -141,8 +218,9 @@ function extractDiscardedIntentSegments(normalized) {
 function collectDateSpans(text) {
    const spans = [];
    const datePatterns = [
-        /\b20\d{2}[-/.](?:0[1-9]|1[0-2])(?:[-/.](?:0[1-9]|[12]\d|3[01]))?\b/g,
-        /\b(?:0?[1-9]|[12]\d|3[01])[./-](?:0?[1-9]|1[0-2])[./-](?:\d{2}|\d{4})\b/g
+        /\b20\d{2}(?:[-/.](?:0?[1-9]|1[0-2]))(?:[-/.](?:0?[1-9]|[12]\d|3[01]))?\b/g,
+        /\b(?:0?[1-9]|[12]\d|3[01])[./-](?:0?[1-9]|1[0-2])[./-](?:\d{2}|\d{4})\b/g,
+        /\b(?:0?[1-9]|[12]\d|3[01])\s+(?:январ[ьяе]|феврал[ьяе]|март[ае]?|апрел[ьяе]|ма[йея]|июн[ьяе]?|июл[ьяе]?|август[ае]?|сентябр[ьяе]?|октябр[ьяе]?|ноябр[ьяе]?|декабр[ьяе]?|january|february|march|april|may|june|july|august|september|october|november|december)(?:\s+20\d{2})?\b/giu
    ];
    for (const datePattern of datePatterns) {
        let match = null;
@@ -155,6 +233,32 @@ function collectDateSpans(text) {
    }
    return spans;
 }
+function collectAmountSpans(text) {
+    const spans = [];
+    const amountPatterns = [/\b\d{1,3}(?:[ \u00A0]\d{3})+(?:[.,]\d{2})?\b/g, /\b\d+[.,]\d{2}\b/g];
+    for (const amountPattern of amountPatterns) {
+        let match = null;
+        while ((match = amountPattern.exec(text)) !== null) {
+            spans.push({
+                start: match.index,
+                end: match.index + match[0].length
+            });
+        }
+    }
+    return spans;
+}
+function collectPercentSpans(text) {
+    const spans = [];
+    const percentPattern = /\b\d{1,3}(?:[.,]\d+)?\s*%/g;
+    let match = null;
+    while ((match = percentPattern.exec(text)) !== null) {
+        spans.push({
+            start: match.index,
+            end: match.index + match[0].length
+        });
+    }
+    return spans;
+}
 function intersectsAnySpan(start, end, spans) {
    return spans.some((span) => start < span.end && end > span.start);
 }
@@ -230,7 +334,7 @@ function extractAccountTokens(text) {
    if (explicitAccounts.size > 0) {
        return Array.from(explicitAccounts);
    }
-    const spans = collectDateSpans(lower);
+    const spans = [...collectDateSpans(lower), ...collectAmountSpans(lower), ...collectPercentSpans(lower)];
    const hasAccountingLexeme = /(?:\bсчет(?:а|у|ом|ов)?\b|\bсч\.?\b|\baccount(?:s)?\b|\bschet(?:a|u|om|ov)?\b|оплат|расчет|аванс|долг|settlement|payment)/iu.test(lower);
    if (!hasAccountingLexeme) {
        return [];
@@ -846,7 +950,7 @@ function extractNormalizedPeriodLiteral(text) {
 }
 function extractFollowupAccountAnchorsLoose(text) {
    const lower = String(text ?? "").toLowerCase();
-    const spans = collectDateSpans(lower);
+    const spans = [...collectDateSpans(lower), ...collectAmountSpans(lower), ...collectPercentSpans(lower)];
    const anchors = [];
    const followupAccountPattern = /\b(?:01|02|08|19|20|21|23|25|26|28|29|44|51|60|62|68|76|97)(?:\.\d{2})?\b/g;
    let match = null;
@@ -1192,6 +1296,13 @@ export class AssistantService {
        };
        const normalized = await this.normalizerService.normalize(normalizePayload);
        const companyAnchors = (0, companyAnchorResolver_1.resolveCompanyAnchors)(userMessage);
+        const businessScopeResolution = resolveBusinessScopeAlignment({
+            userMessage,
+            companyAnchors,
+            normalized: normalized.normalized,
+            routeSummary: normalized.route_hint_summary
+        });
+        const resolvedRouteSummary = businessScopeResolution.route_summary_resolved;
        const inferredDomainByMessage = inferP0DomainFromMessage(userMessage);
        const focusDomainForGuards = inferredDomainByMessage === "settlements_60_62" ||
            inferredDomainByMessage === "vat_document_register_book" ||
@@ -1214,8 +1325,8 @@ export class AssistantService {
            focusDomainHint: focusDomainForGuards,
            primaryPeriod: temporalGuard.primary_period_window
        });
-        const requirementExtraction = extractRequirements(normalized.route_hint_summary, normalized.normalized, userMessage);
-        let executionPlan = toExecutionPlan(normalized.route_hint_summary, normalized.normalized, userMessage, requirementExtraction.byFragment);
+        const requirementExtraction = extractRequirements(resolvedRouteSummary, normalized.normalized, userMessage);
+        let executionPlan = toExecutionPlan(resolvedRouteSummary, normalized.normalized, userMessage, requirementExtraction.byFragment);
        executionPlan = (0, assistantRuntimeGuards_1.applyTemporalHintToExecutionPlan)(executionPlan, temporalGuard);
        executionPlan = (0, assistantRuntimeGuards_1.applyPolarityHintToExecutionPlan)(executionPlan, domainPolarityGuardInitial);
        const retrievalCalls = [];
@@ -1305,7 +1416,8 @@ export class AssistantService {
            polarity: polarityGuardResult.audit,
            evidence: evidenceGateResult.audit,
            claimAnchors: claimAnchorAudit,
-            targetedEvidenceHitRate: targetedEvidenceResult.audit.targeted_evidence_hit_rate
+            targetedEvidenceHitRate: targetedEvidenceResult.audit.targeted_evidence_hit_rate,
+            businessScopeResolved: businessScopeResolution.business_scope_resolved
        });
        const groundingCheck = (0, assistantRuntimeGuards_1.applyEligibilityToGroundingCheck)(groundingCheckBase, groundedAnswerEligibilityGuard);
        const focusDomainHint = followupBinding.usage?.applied
@@ -1317,7 +1429,7 @@ export class AssistantService {
        const normalizationPeriodExplicit = hasExplicitPeriodAnchorFromNormalized(normalized.normalized) || hasPeriodInCompanyAnchors;
        const composition = (0, answerComposer_1.composeAssistantAnswer)({
            userMessage,
-            routeSummary: normalized.route_hint_summary,
+            routeSummary: resolvedRouteSummary,
            retrievalResults,
            requirements: coverageEvaluation.requirements,
            coverageReport: coverageEvaluation.coverage,
@@ -1351,7 +1463,7 @@ export class AssistantService {
                timestamp: new Date().toISOString(),
                questionId: userItem.message_id,
                userMessage,
-                routeSummary: normalized.route_hint_summary,
+                routeSummary: resolvedRouteSummary,
                requirements: coverageEvaluation.requirements,
                coverageReport: coverageEvaluation.coverage,
                retrievalResults,
@@ -1367,11 +1479,11 @@ export class AssistantService {
            prompt_version: normalized.prompt_version,
            schema_version: normalized.schema_version,
            fallback_type: composition.fallback_type,
-            route_summary: normalized.route_hint_summary,
+            route_summary: resolvedRouteSummary,
            fragments: extractFragments(normalized.normalized),
            requirements_extracted: coverageEvaluation.requirements,
            coverage_report: coverageEvaluation.coverage,
-            routes: toDebugRoutes(normalized.route_hint_summary),
+            routes: toDebugRoutes(resolvedRouteSummary),
            retrieval_status: retrievalResults.map((item) => ({
                fragment_id: item.fragment_id,
                requirement_ids: item.requirement_ids,
@@ -1384,16 +1496,29 @@ export class AssistantService {
            dropped_intent_segments: extractDiscardedIntentSegments(normalized.normalized),
            question_type_class: questionTypeClass,
            company_anchors: companyAnchors,
+            business_scope_raw: businessScopeResolution.business_scope_raw,
+            business_scope_resolved: businessScopeResolution.business_scope_resolved,
+            company_grounding_applied: businessScopeResolution.company_grounding_applied,
+            scope_resolution_reason: businessScopeResolution.scope_resolution_reason,
            raw_time_anchor: temporalGuard.raw_time_anchor,
+            raw_time_scope: temporalGuard.raw_time_scope,
            resolved_time_anchor: temporalGuard.resolved_time_anchor,
+            resolved_primary_period: temporalGuard.resolved_primary_period,
+            temporal_alignment_status: temporalGuard.temporal_alignment_status,
            temporal_resolution_source: temporalGuard.temporal_resolution_source,
+            temporal_guard_basis: temporalGuard.temporal_guard_basis,
            temporal_guard_applied: temporalGuard.temporal_guard_applied,
            temporal_guard_outcome: temporalGuard.temporal_guard_outcome,
            temporal_guard: temporalGuard,
+            raw_numeric_tokens: polarityGuardResult.audit.raw_numeric_tokens,
+            classified_numeric_tokens: polarityGuardResult.audit.classified_numeric_tokens,
+            rejected_as_non_accounts: polarityGuardResult.audit.rejected_as_non_accounts,
+            resolved_account_anchors: polarityGuardResult.audit.resolved_account_anchors,
            domain_polarity_guard: polarityGuardResult.audit,
            claim_anchor_audit: claimAnchorAudit,
            targeted_evidence_acquisition: targetedEvidenceResult.audit,
            evidence_admissibility_gate: evidenceGateResult.audit,
+            eligibility_time_basis: groundedAnswerEligibilityGuard.eligibility_time_basis,
            grounded_answer_eligibility_guard: groundedAnswerEligibilityGuard,
            ...(followupBinding.usage ? { followup_state_usage: followupBinding.usage } : {}),
            problem_centric_answer_applied: composition.problem_centric_answer_applied ?? false,
@@ -1438,7 +1563,7 @@ export class AssistantService {
                normalizer_output: normalized.normalized,
                execution_plan: executionPlan,
                resolved_execution_state: extractExecutionState(normalized.normalized),
-                routes: toDebugRoutes(normalized.route_hint_summary),
+                routes: toDebugRoutes(resolvedRouteSummary),
                retrieval_calls: retrievalCalls,
                retrieval_results_raw: retrievalResultsRaw,
                retrieval_results_normalized: retrievalResults,
@@ -1460,16 +1585,29 @@ export class AssistantService {
                dropped_intent_segments: extractDiscardedIntentSegments(normalized.normalized),
                question_type_class: questionTypeClass,
                company_anchors: companyAnchors,
+                business_scope_raw: businessScopeResolution.business_scope_raw,
+                business_scope_resolved: businessScopeResolution.business_scope_resolved,
+                company_grounding_applied: businessScopeResolution.company_grounding_applied,
+                scope_resolution_reason: businessScopeResolution.scope_resolution_reason,
                raw_time_anchor: temporalGuard.raw_time_anchor,
+                raw_time_scope: temporalGuard.raw_time_scope,
                resolved_time_anchor: temporalGuard.resolved_time_anchor,
+                resolved_primary_period: temporalGuard.resolved_primary_period,
+                temporal_alignment_status: temporalGuard.temporal_alignment_status,
                temporal_resolution_source: temporalGuard.temporal_resolution_source,
+                temporal_guard_basis: temporalGuard.temporal_guard_basis,
                temporal_guard_applied: temporalGuard.temporal_guard_applied,
                temporal_guard_outcome: temporalGuard.temporal_guard_outcome,
                temporal_guard: temporalGuard,
+                raw_numeric_tokens: polarityGuardResult.audit.raw_numeric_tokens,
+                classified_numeric_tokens: polarityGuardResult.audit.classified_numeric_tokens,
+                rejected_as_non_accounts: polarityGuardResult.audit.rejected_as_non_accounts,
+                resolved_account_anchors: polarityGuardResult.audit.resolved_account_anchors,
                domain_polarity_guard: polarityGuardResult.audit,
                claim_anchor_audit: claimAnchorAudit,
                targeted_evidence_acquisition: targetedEvidenceResult.audit,
                evidence_admissibility_gate: evidenceGateResult.audit,
+                eligibility_time_basis: groundedAnswerEligibilityGuard.eligibility_time_basis,
                grounded_answer_eligibility_guard: groundedAnswerEligibilityGuard,
                ...(followupBinding.usage ? { followup_state_usage: followupBinding.usage } : {}),
                problem_centric_answer_applied: composition.problem_centric_answer_applied ?? false,