Initial import NDC_1C

This commit is contained in:
2026-03-26 10:38:25 +03:00
commit a162d77ef7
2943 changed files with 3615871 additions and 0 deletions
@@ -0,0 +1,199 @@
import type {
AssistantRequirement,
RequirementCoverageReport,
UnifiedRetrievalResult
} from "../types/assistant";
import type { RouteHintSummary } from "../types/normalizer";
import type {
InvestigationLastAnswerMode,
InvestigationNarrowingStatus,
InvestigationState
} from "../types/stage1Contracts";
import {
INVESTIGATION_MAX_EVIDENCE_REFS,
INVESTIGATION_MAX_PRIMARY_ACCOUNTS,
INVESTIGATION_MAX_REQUIREMENT_LINKS,
INVESTIGATION_MAX_UNCERTAINTIES,
INVESTIGATION_STATE_SCHEMA_VERSION
} from "../types/stage1Contracts";
interface UpdateInvestigationStateInput {
previous: InvestigationState;
timestamp: string;
questionId: string;
userMessage: string;
routeSummary: RouteHintSummary | null;
requirements: AssistantRequirement[];
coverageReport: RequirementCoverageReport;
retrievalResults: UnifiedRetrievalResult[];
replyType: InvestigationLastAnswerMode;
}
function uniqueStrings(values: string[]): string[] {
return Array.from(new Set(values.map((item) => item.trim()).filter(Boolean)));
}
function capStrings(values: string[], max: number): string[] {
return uniqueStrings(values).slice(0, max);
}
function detectAccounts(text: string): string[] {
return capStrings(text.match(/\b\d{2}(?:\.\d{2})?\b/g) ?? [], INVESTIGATION_MAX_PRIMARY_ACCOUNTS);
}
function detectPeriod(text: string): string | null {
const monthly = text.match(/\b(20\d{2})[-/.](0[1-9]|1[0-2])\b/);
if (monthly) return `${monthly[1]}-${monthly[2]}`;
const yearly = text.match(/\b(20\d{2})\b/);
if (yearly) return yearly[1];
return null;
}
function deriveDomain(routeSummary: RouteHintSummary | null): string | null {
if (!routeSummary) return null;
if (routeSummary.mode === "legacy_v1") {
return routeSummary.route_hint;
}
const routes = routeSummary.decisions.map((item) => item.route).filter((route) => route !== "no_route");
const uniqueRoutes = uniqueStrings(routes);
if (uniqueRoutes.length === 0) {
return "no_route";
}
return uniqueRoutes.join(",");
}
function deriveNarrowingStatus(
routeSummary: RouteHintSummary | null,
coverageReport: RequirementCoverageReport
): InvestigationNarrowingStatus {
if (!routeSummary) {
return "unknown";
}
if (routeSummary.mode === "legacy_v1") {
return "not_needed";
}
if (routeSummary.fallback.type === "clarification" || coverageReport.clarification_needed_for.length > 0) {
return "needs_clarification";
}
const hasNoRoute = routeSummary.decisions.some((item) => item.route === "no_route");
if (hasNoRoute) {
return "broad_guarded";
}
return routeSummary.decisions.length > 1 ? "applied" : "not_needed";
}
function deriveQueryModeHint(routeSummary: RouteHintSummary | null): InvestigationState["query_mode_hint"] {
if (!routeSummary) {
return "investigation_candidate";
}
if (routeSummary.mode === "legacy_v1") {
return "direct_answer";
}
return routeSummary.fallback.type === "none" ? "direct_answer" : "investigation_candidate";
}
function collectEvidenceRefs(retrievalResults: UnifiedRetrievalResult[]): string[] {
const refs = retrievalResults.flatMap((result) => result.evidence.map((item) => item.evidence_id));
return capStrings(refs, INVESTIGATION_MAX_EVIDENCE_REFS);
}
function collectOpenUncertainties(
coverageReport: RequirementCoverageReport,
retrievalResults: UnifiedRetrievalResult[]
): string[] {
const requirementNotes = [
...coverageReport.requirements_uncovered.map((item) => `uncovered:${item}`),
...coverageReport.requirements_partially_covered.map((item) => `partial:${item}`),
...coverageReport.clarification_needed_for.map((item) => `clarify:${item}`),
...coverageReport.out_of_scope_requirements.map((item) => `out_of_scope:${item}`)
];
const limitationNotes = retrievalResults.flatMap((result) => result.limitations).slice(0, 6);
return capStrings([...requirementNotes, ...limitationNotes], INVESTIGATION_MAX_UNCERTAINTIES);
}
export function cloneInvestigationState(state: InvestigationState | null): InvestigationState | null {
if (!state) return null;
return {
...state,
focus: {
...state.focus,
primary_accounts: [...state.focus.primary_accounts]
},
evidence_refs: [...state.evidence_refs],
open_uncertainties: [...state.open_uncertainties],
followup_context: state.followup_context
? {
...state.followup_context,
referenced_requirement_ids: [...state.followup_context.referenced_requirement_ids]
}
: null
};
}
export function createEmptyInvestigationState(sessionId: string, timestamp = new Date().toISOString()): InvestigationState {
return {
schema_version: INVESTIGATION_STATE_SCHEMA_VERSION,
session_id: sessionId,
status: "idle",
turn_index: 0,
updated_at: timestamp,
question_id: null,
focus: {
domain: null,
period: null,
primary_accounts: [],
active_query_subject: null
},
narrowing_status: "unknown",
evidence_refs: [],
open_uncertainties: [],
last_answer_mode: null,
followup_context: null,
query_mode_hint: "direct_answer"
};
}
export function updateInvestigationState(input: UpdateInvestigationStateInput): InvestigationState {
const previous = input.previous;
const focusFromMessage = capStrings(detectAccounts(input.userMessage), INVESTIGATION_MAX_PRIMARY_ACCOUNTS);
const requirementIds = capStrings(
input.requirements.map((item) => item.requirement_id),
INVESTIGATION_MAX_REQUIREMENT_LINKS
);
const mainRequirement = input.requirements[0]?.requirement_text ?? input.userMessage;
return {
schema_version: INVESTIGATION_STATE_SCHEMA_VERSION,
session_id: previous.session_id,
status: "active",
turn_index: previous.turn_index + 1,
updated_at: input.timestamp,
question_id: input.questionId,
focus: {
domain: deriveDomain(input.routeSummary) ?? previous.focus.domain,
period: detectPeriod(input.userMessage) ?? previous.focus.period,
primary_accounts: capStrings(
[...focusFromMessage, ...previous.focus.primary_accounts],
INVESTIGATION_MAX_PRIMARY_ACCOUNTS
),
active_query_subject: mainRequirement.slice(0, 180)
},
narrowing_status: deriveNarrowingStatus(input.routeSummary, input.coverageReport),
evidence_refs: capStrings(
[...collectEvidenceRefs(input.retrievalResults), ...previous.evidence_refs],
INVESTIGATION_MAX_EVIDENCE_REFS
),
open_uncertainties: collectOpenUncertainties(input.coverageReport, input.retrievalResults),
last_answer_mode: input.replyType,
followup_context: {
previous_question_id: previous.question_id,
last_user_message: input.userMessage.slice(0, 240),
referenced_requirement_ids: requirementIds
},
query_mode_hint: deriveQueryModeHint(input.routeSummary)
};
}