feat(docpage): D3 절 구조 탐색기 — 슬림 인사이트 레일 + 절 트리 (frontend only)

문서 상세 /documents/[id] 재구성 (BE 무변경):
- 우측 탭(정보/AI/관리) → 슬림 전역 인사이트 레일: 요약·심층·불일치를 탭
  게이트 없이 상시 노출(details open, 모바일은 접기 가능), 정보/관리는 접이.
  → 가공 자료가 탭/온디맨드에 묻히던 IA 문제(G1) 해소.
- SectionOutline 절 목차 레벨 기반 들여쓰기(평탄→트리 모양).
- 모바일: 본문 메인 + 절목차/인사이트/정보/관리 접이 + 절 탭 본문 이동(기존 구조 활용).
관련 문서(See Also)는 v1 제외(자리만 유지). 심화 목업 = comparisons/2026-06-13-ds-docpage-d3-deepened.html.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

app/api/documents.py

@@ -210,8 +210,14 @@ class DocumentDetailResponse(DocumentResponse):
 
 
 class AcceptSuggestionRequest(BaseModel):
-    """§1 accept-suggestion 요청 body — stale payload / doc 수정 검출."""
+    """§1 accept-suggestion 요청 body — stale payload / doc 수정 검출.
+
+    jurisdiction: 안전 자료실 A-2 — material_type 제안 승인 시 사용자가 지정하는 관할.
+    law 승인은 필수 (기본값 없음 — KR 자동 부여 시 외국 자료가 KR 법령으로 오염되는
+    경로를 차단, plan A-2 계약).
+    """
     expected_source_updated_at: datetime
+    jurisdiction: str | None = None
 
 
 class DocumentUpdate(BaseModel):
@@ -537,6 +543,8 @@ async def list_documents(
     category: str | None = Query(None, description="doc_category enum — 지정 시 기본 news/memo 제외 해제"),
     has_suggestion: bool | None = Query(None, description="true: ai_suggestion IS NOT NULL"),
     proposed_category: str | None = Query(None, description="ai_suggestion.proposed_category 필터"),
+    material_type: str | None = Query(None, description="안전 자료실 C-1: 자료유형. 지정 시 기본 exclude 해제"),
+    jurisdiction: str | None = Query(None, description="안전 자료실 C-1: 관할 (KR/US/...)"),
 ):
     """문서 목록 조회 (페이지네이션 + 필터).
 
@@ -550,6 +558,10 @@ async def list_documents(
     if category:
         # 명시적 카테고리 필터 — 기본 exclude 해제
         query = query.where(Document.category == category)
+    elif material_type:
+        # 안전 자료실 C-1: material_type 지정 = 기본 exclude(news·law_monitor·note) 해제.
+        # 안전 코퍼스 본체(KOSHA 사례·CSB·법령 등)가 전부 note/crawl 채널이라 exclude 면 빈 화면.
+        query = query.where(Document.material_type == material_type)
     else:
         # 기본 목록: 뉴스/메모/법령 제외 (문서함 용도)
         query = query.where(
@@ -558,6 +570,9 @@ async def list_documents(
             Document.file_type != "note",
         )
 
+    if jurisdiction:
+        query = query.where(Document.jurisdiction == jurisdiction)
+
     if has_suggestion is True:
         query = query.where(Document.ai_suggestion.isnot(None))
     elif has_suggestion is False:
@@ -1244,11 +1259,49 @@ async def accept_suggestion(
     # payload 적용
     proposed_category = doc.ai_suggestion.get("proposed_category")
     proposed_path = doc.ai_suggestion.get("proposed_path")
+    # 안전 자료실 A-2 — material_type 제안 (classify 의 document_type 결정적 매핑)
+    proposed_material = doc.ai_suggestion.get("proposed_material_type")
 
-    if not proposed_category:
-        raise HTTPException(status_code=422, detail="proposed_category 누락된 suggestion")
+    if not proposed_category and not proposed_material:
+        raise HTTPException(
+            status_code=422,
+            detail="proposed_category/proposed_material_type 둘 다 누락된 suggestion",
+        )
 
-    doc.category = proposed_category
+    if proposed_category:
+        doc.category = proposed_category
+
+    if proposed_material:
+        _MATERIAL_TYPES = {"law", "paper", "book", "incident", "manual", "standard", "guide"}
+        _JURISDICTIONS = {"KR", "US", "EU", "JP", "GB", "INT"}
+        if proposed_material not in _MATERIAL_TYPES:
+            raise HTTPException(
+                status_code=422, detail=f"허용 밖 material_type: {proposed_material}"
+            )
+        jur = body.jurisdiction or doc.ai_suggestion.get("proposed_jurisdiction")
+        if jur is not None and jur not in _JURISDICTIONS:
+            raise HTTPException(status_code=422, detail=f"허용 밖 jurisdiction: {jur}")
+        # law = 국가 필수 입력, 기본값 없음 (plan A-2 — KR 자동 부여 시 외국 법령 오염.
+        # DB CHECK(chk_documents_law_jurisdiction) 도 거부하지만 422 로 명시 안내).
+        if proposed_material == "law" and not jur:
+            raise HTTPException(
+                status_code=422,
+                detail="법령(law) 승인은 jurisdiction 필수 — body.jurisdiction 으로 국가를 지정하세요 (기본값 없음)",
+            )
+        doc.material_type = proposed_material
+        doc.jurisdiction = jur
+        # 미러 동기화 1문 — jurisdiction 부여/정정 시 청크 country 동반 UPDATE
+        # (leg 간 국가 불일치 방지, plan A-2 계약. 단일 지점 = 본 승인 경로).
+        if jur:
+            from sqlalchemy import update as sa_update
+
+            from models.chunk import DocumentChunk
+
+            await session.execute(
+                sa_update(DocumentChunk)
+                .where(DocumentChunk.doc_id == doc.id)
+                .values(country=jur)
+            )
 
     # user_tags append (중복 방지, normalize + dedup 통과)
     if proposed_path:

app/api/queue_overview.py

@@ -59,6 +59,20 @@ class SummarizeEta(BaseModel):
     eta_minutes: int | None
 
 
+class MachineDone(BaseModel):
+    """머신 1대의 summarize 완료 실적 (분담 표시용)."""
+    done_1h: int
+    done_today: int
+
+
+class SummarizeByMachine(BaseModel):
+    """summarize 풀의 머신별 완료 실적 분담 — 보드 레인의 '맥미니 vs 맥북'
+    오프로드 가시화용. rows_to_summarize_split 이 이미 계산하던 값의 노출
+    (ds-board-merged A-1, 신규 수집 SQL 0)."""
+    macmini: MachineDone
+    macbook: MachineDone
+
+
 class TrendBucket(BaseModel):
     """summarize 24h 추이 버킷 — hour 는 KST "HH:00" 라벨."""
     hour: str
@@ -93,6 +107,7 @@ class QueueOverviewResponse(BaseModel):
     machines: list[MachineCard]
     stages: list[StageRow]
     summarize_eta: SummarizeEta
+    summarize_by_machine: SummarizeByMachine
     trend_24h: list[TrendBucket]
     totals: Totals
 

app/api/search.py

@@ -12,6 +12,7 @@
 import asyncio
 import hmac
 import time
+from datetime import date
 from typing import Annotated, Literal
 
 from fastapi import APIRouter, BackgroundTasks, Depends, Header, Query
@@ -31,6 +32,8 @@ from services.search.fusion_service import DEFAULT_FUSION
 from services.search.grounding_check import check as grounding_check
 from services.search.refusal_gate import RefusalDecision, decide as refusal_decide
 from services.search import query_rewriter
+from services.search.retrieval_service import AxisFilter
+from services.search.result_decorate import compute_facets, decorate_version_status
 from services.search.search_pipeline import PipelineResult, run_search
 from services.search.synthesis_service import SynthesisResult, synthesize
 from services.search.verifier_service import VerifierResult, verify
@@ -70,6 +73,14 @@ class SearchResult(BaseModel):
     # PR-RAG-Time-1: freshness decay 디버그 메타. apply_freshness_decay 가 채움.
     # 비적용 row 도 채워짐(freshness_policy=None). base_score 는 항상 보존.
     freshness_debug: dict | None = None
+    # 안전 자료실 C-1: 분류 축 메타 (3 leg SELECT 에서 채움 — additive, ranking 무관).
+    # D-1 UI 결과 카드 유형별 렌더 + 해외 법령(B-5) 가동 시 국가 무표지 혼재 차단의 선행 조건.
+    material_type: str | None = None
+    jurisdiction: str | None = None
+    published_date: date | None = None
+    # 안전 자료실 C-1 후속: 법령 버전 상태(legal_meta.version_status) — wrapper 1회 decorate.
+    # law 결과만 채워짐(legal_meta 위성), 그 외/무매핑 law = None. D-1 버전 뱃지 선행.
+    version_status: str | None = None
 
 
 # ─── Phase 0.4: 디버그 응답 스키마 ─────────────────────────
@@ -101,6 +112,9 @@ class SearchResponse(BaseModel):
     query: str
     mode: str
     debug: SearchDebug | None = None
+    # 안전 자료실 C-1 후속: facets=true 일 때만 채워짐(미요청=None, byte 불변).
+    # top-K 결과 내 분류 축 분포 라벨 {axis: {label: count}}.
+    facets: dict[str, dict[str, int]] | None = None
 
 
 def _to_debug_candidates(rows: list[SearchResult], n: int = 20) -> list[DebugCandidate]:
@@ -205,9 +219,23 @@ async def search(
             "분리용. production 검색에는 사용 금지 (latency 큼)."
         ),
     ),
+    material_type: str | None = Query(
+        None, description="안전 자료실 C-1: 자료유형 필터 CSV (law,paper,incident,...). material_type = ANY"),
+    jurisdiction: str | None = Query(
+        None, description="안전 자료실 C-1: 관할 필터 (KR/US/EU/JP/GB/INT)"),
+    year_from: int | None = Query(None, ge=1900, le=2100, description="published_date 연도 하한 (NULL=created_at fallback)"),
+    year_to: int | None = Query(None, ge=1900, le=2100, description="published_date 연도 상한"),
+    facets: bool = Query(False, description="안전 자료실 C-1 후속: top-K 결과 분류 축 분포(material_type/jurisdiction/version_status)를 응답 facets 에 집계. 미지정=계산/노출 0"),
 ):
     """문서 검색 — FTS + ILIKE + 벡터 결합 (Phase 3.1 이후 run_search wrapper)"""
     try:
+        axis = AxisFilter(
+            material_types=[m.strip() for m in material_type.split(",") if m.strip()]
+            if material_type else None,
+            jurisdiction=jurisdiction,
+            year_from=year_from,
+            year_to=year_to,
+        )
         pr = await run_search(
             session,
             q,
@@ -223,6 +251,7 @@ async def search(
             rewrite_backend=rewrite_backend,
             corpus_variant=corpus_variant,
             exact_knn=exact_knn,
+            axis=axis,
         )
     except ValueError as e:
         # _resolve_backend / _resolve_reranker / _resolve_rewrite_backend / _resolve_corpus_variant unknown slug → HTTP 400
@@ -313,12 +342,17 @@ async def search(
 
     debug_obj = _build_search_debug(pr) if debug else None
 
+    # 안전 자료실 C-1 후속 — wrapper decoration (검색 코어 무접촉, ranking 무관)
+    await decorate_version_status(session, pr.results)   # 법령 결과에 version_status
+    facets_obj = compute_facets(pr.results) if facets else None
+
     return SearchResponse(
         results=pr.results,
         total=len(pr.results),
         query=q,
         mode=pr.mode,
         debug=debug_obj,
+        facets=facets_obj,
     )
 
 

app/main.py

@@ -53,8 +53,8 @@ async def lifespan(app: FastAPI):
     from workers.dedup_reconcile import run as dedup_reconcile_run
     from workers.digest_worker import run as global_digest_run
     from workers.file_watcher import watch_inbox
-    from workers.law_monitor import run as law_monitor_run
     from workers.mailplus_archive import run as mailplus_run
+    from workers.statute_collector import run as statute_run
     from workers.news_collector import run as news_collector_run
     from workers.fulltext_worker import reconcile_unresolved as fulltext_reconcile_run
     from workers.kosha_collector import run as kosha_collector_run
@@ -120,7 +120,9 @@ async def lifespan(app: FastAPI):
     # safety > law > manual 우선순위로 25건씩. 6720 레거시 → 야간당 ~150건 → 약 45일 소화.
     scheduler.add_job(tier_backfill_run, "interval", minutes=30, id="tier_backfill")
     # 일일 스케줄 (KST)
-    scheduler.add_job(law_monitor_run, CronTrigger(hour=7, timezone=KST), id="law_monitor")
+    # statute_collector = 구 law_monitor 대체 (safety-library-1 B-1 PR②) — poll→ingest→
+    # 생애주기 잡(버전 시리즈 승격·supersede·레거시 스윕·repeal) 통째 (R8-B1).
+    scheduler.add_job(statute_run, CronTrigger(hour=7, timezone=KST), id="statute_collector")
     scheduler.add_job(mailplus_run, CronTrigger(hour=7, timezone=KST), id="mailplus_morning")
     scheduler.add_job(mailplus_run, CronTrigger(hour=18, timezone=KST), id="mailplus_evening")
     scheduler.add_job(daily_digest_run, CronTrigger(hour=20, timezone=KST), id="daily_digest")

app/models/document.py

@@ -1,9 +1,9 @@
 """documents 테이블 ORM"""
 
-from datetime import datetime
+from datetime import date, datetime
 
 from pgvector.sqlalchemy import Vector
-from sqlalchemy import BigInteger, Boolean, DateTime, Enum, ForeignKey, Integer, String, Text
+from sqlalchemy import BigInteger, Boolean, Date, DateTime, Enum, ForeignKey, Integer, String, Text
 from sqlalchemy.dialects.postgresql import JSONB
 from sqlalchemy.orm import Mapped, mapped_column
 
@@ -146,6 +146,16 @@ class Document(Base):
     # /accept-suggestion 승인 시에만 category / user_tags 반영 (자동 전이 금지)
     ai_suggestion: Mapped[dict | None] = mapped_column(JSONB)
 
+    # === 안전 자료실 분류 축 (plan safety-library-1, migrations 340~345) ===
+    # 자료유형 — law/paper/book/incident/manual/standard/guide (TEXT+CHECK, enum 아님).
+    # 수집기 ingest 시점 deterministic 부여 (classify-skip 경로 다수 — classify_worker 의존 금지).
+    # AI 라우팅(subject_domain) 매칭 키 사용 금지 (axis separation — category 와 동일 불변식).
+    material_type: Mapped[str | None] = mapped_column(Text)
+    # 관할 — KR/US/EU/JP/GB/INT. law 는 CHECK 로 jurisdiction NOT NULL 구조 강제 (migration 344).
+    jurisdiction: Mapped[str | None] = mapped_column(Text)
+    # 유형별 대표 날짜 — 법령=COALESCE(시행일, 공포일) / 논문=발행일 / 재해=발생일
+    published_date: Mapped[date | None] = mapped_column(Date)
+
     # PR-B B-1: summary_triage (4B, 상시) / summary_deep (26B, 에스컬레이션) 분할 산출
     ai_tldr: Mapped[str | None] = mapped_column(Text)                          # ≤60자 TL;DR
     ai_bullets: Mapped[list | None] = mapped_column(JSONB)                     # 3~5개 핵심 bullets

app/models/legal_act.py

@@ -0,0 +1,73 @@
+"""legal_acts / legal_meta 테이블 ORM — 법령 레지스트리(워치리스트 겸) + 버전 위성
+
+plan: safety-library-1 (migrations 346~347).
+- legal_acts = 폴링 순회 대상 목록이 곧 테이블 (news_sources 패턴의 법령판).
+  KOSHA GUIDE(비법령)·KGS Code(watch-폴더 단독 트랙)는 비대상.
+- legal_meta = 법령 문서 1버전(또는 별표·해석례 1건)당 1행, documents 1:0..1 위성.
+  version_status 전이는 statute_collector 의 일일 잡이 유일한 코드 지점
+  (전 버전 pending 적재 → 잡이 승격·supersede·repeal 을 한 트랜잭션 처리).
+"""
+
+from datetime import date, datetime
+
+from sqlalchemy import BigInteger, Boolean, Date, DateTime, ForeignKey, Text, UniqueConstraint
+from sqlalchemy.orm import Mapped, mapped_column
+
+from core.database import Base
+
+
+class LegalAct(Base):
+    __tablename__ = "legal_acts"
+
+    # 'kr-law:{법령ID}' / 'us-cfr:29-1910' 형식. KGS 는 시드 비대상 (R3-M5).
+    family_id: Mapped[str] = mapped_column(Text, primary_key=True)
+    # 어댑터 상수 고정값 — 파싱 결과에서 추론 금지 (코어가 적재 직전 assert)
+    jurisdiction: Mapped[str] = mapped_column(Text, nullable=False)
+    # statute(법률) / decree(시행령) / rule(시행규칙·부령) / admin_rule(고시·예규) / code(법정 위임 상세기준)
+    law_level: Mapped[str] = mapped_column(Text, nullable=False)
+    title: Mapped[str] = mapped_column(Text, nullable=False)
+    title_ko: Mapped[str | None] = mapped_column(Text)
+    # 법률 → 시행령 → 시행규칙 계층
+    parent_family_id: Mapped[str | None] = mapped_column(ForeignKey("legal_acts.family_id"))
+    # 법령ID / CFR part / CELEX / e-Gov law_id 등 소스 고유 식별자
+    native_id: Mapped[str] = mapped_column(Text, nullable=False)
+    # 'law.go.kr' / 'ecfr' / 'cellar' / 'egov_v2' / 'leg_gov_uk'
+    source_api: Mapped[str] = mapped_column(Text, nullable=False)
+    # 시드 26개 전부 true — '우선순위'는 정렬일 뿐 watch 제외 아님 (R3-B1)
+    watch: Mapped[bool] = mapped_column(Boolean, nullable=False, default=True)
+    poll_cycle: Mapped[str] = mapped_column(Text, nullable=False, default="daily")
+    # 변경이력 폴링 워터마크 — 파싱 검증 통과 후에만 영속
+    watermark: Mapped[str | None] = mapped_column(Text)
+    # 어댑터는 폐지 감지 마킹만, repealed 전이는 일일 잡 (R3-M3)
+    repeal_detected_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))
+    created_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), default=datetime.now
+    )
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), default=datetime.now, onupdate=datetime.now
+    )
+
+
+class LegalMeta(Base):
+    __tablename__ = "legal_meta"
+    __table_args__ = (
+        # 버전 dedup 구조 강제 — annex 는 version_key='MST|별표N' 합성형 (R3-M4)
+        UniqueConstraint("family_id", "law_doc_kind", "version_key", name="uq_legal_meta_version"),
+    )
+
+    document_id: Mapped[int] = mapped_column(
+        BigInteger, ForeignKey("documents.id", ondelete="CASCADE"), primary_key=True
+    )
+    family_id: Mapped[str] = mapped_column(
+        ForeignKey("legal_acts.family_id"), nullable=False
+    )
+    # primary(본문) / annex(별표·서식) / interpretation(해석례)
+    law_doc_kind: Mapped[str] = mapped_column(Text, nullable=False, default="primary")
+    version_key: Mapped[str] = mapped_column(Text, nullable=False)
+    promulgation_date: Mapped[date | None] = mapped_column(Date)
+    effective_date: Mapped[date | None] = mapped_column(Date)
+    # pending → current → superseded / repealed. 전이는 일일 잡 단일 지점, KST 기준.
+    version_status: Mapped[str] = mapped_column(Text, nullable=False, default="pending")
+    created_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), default=datetime.now
+    )

app/models/news_source.py

@@ -53,3 +53,12 @@ class NewsSource(Base):
              name="source_channel"),
         default="news",
     )
+
+    # ── 안전 자료실 분류 축 (plan safety-library-1 A-2, migrations 352~355) ──
+    # 자료유형 기본값 — documents.material_type 으로 ingest 시점 전파 (NULL=비대상).
+    # jurisdiction 은 별도 컬럼 없이 country 전파, 단 paper 는 코드에서 NULL 강제.
+    material_type: Mapped[str | None] = mapped_column(Text)
+    # extract_meta.license 주입용 — kogl/ogl/public_domain/proprietary/unknown.
+    # 미확정 = 보수적(unknown + redistribute=false), 근거 확보 시 완화.
+    license_scheme: Mapped[str | None] = mapped_column(Text)
+    license_redistribute: Mapped[bool | None] = mapped_column(Boolean)

app/services/briefing/loader.py

@@ -15,11 +15,12 @@ from sqlalchemy import text
 
 from core.database import async_session
 from core.utils import setup_logger
+from services.search.license_filter import restricted_exclude_sql
 
 logger = setup_logger("briefing_loader")
 
 
-_NEWS_WINDOW_SQL = text("""
+_NEWS_WINDOW_SQL = text(f"""
     SELECT
         d.id,
         d.title,
@@ -41,6 +42,8 @@ _NEWS_WINDOW_SQL = text("""
       AND d.created_at < :window_end
       AND d.embedding IS NOT NULL
       AND d.ai_summary IS NOT NULL
+      -- 안전 자료실 B-4: licensed_restricted 발행 차단 (digest 와 동일 공유 술어, 경로 일관성)
+      AND {restricted_exclude_sql("d")}
 """)
 
 
@@ -49,7 +52,7 @@ _SOURCE_COUNTRY_SQL = text("""
 """)
 
 
-_HISTORICAL_CANDIDATES_SQL = text("""
+_HISTORICAL_CANDIDATES_SQL = text(f"""
     SELECT
         d.id,
         d.title,
@@ -63,6 +66,8 @@ _HISTORICAL_CANDIDATES_SQL = text("""
       AND d.created_at < :hist_end
       AND d.embedding IS NOT NULL
       AND d.ai_summary IS NOT NULL
+      -- 안전 자료실 B-4: licensed_restricted 발행 차단 (공유 술어)
+      AND {restricted_exclude_sql("d")}
 """)
 
 

app/services/digest/loader.py

@@ -15,11 +15,12 @@ from sqlalchemy import text
 
 from core.database import async_session
 from core.utils import setup_logger
+from services.search.license_filter import restricted_exclude_sql
 
 logger = setup_logger("digest_loader")
 
 
-_NEWS_WINDOW_SQL = text("""
+_NEWS_WINDOW_SQL = text(f"""
     SELECT
         d.id,
         d.title,
@@ -41,6 +42,9 @@ _NEWS_WINDOW_SQL = text("""
       AND d.created_at < :window_end
       AND d.embedding IS NOT NULL
       AND d.ai_summary IS NOT NULL
+      -- 안전 자료실 B-4: licensed_restricted 발행 차단 (모든 경로 공유 술어 = license_filter).
+      -- news 채널엔 현재 restricted 부재 = 방어적 게이트(미래 유료 news 소스 대비, 경로 누락 방지).
+      AND {restricted_exclude_sql("d")}
 """)
 
 

app/services/queue_overview.py

@@ -213,6 +213,16 @@ def build_summarize_eta(stage_stats: dict[str, dict]) -> dict:
     }
 
 
+def build_summarize_by_machine(summarize_split: dict[str, dict]) -> dict:
+    """summarize 머신별 완료 실적 분담 (macmini vs macbook) — 보드 레인의
+    오프로드 가시화용. rows_to_summarize_split 이 이미 만든 값을 응답 형태로
+    투영(done_1h/done_today 만, done_15m 은 내부 state 판정 전용이라 제외)."""
+    def m(key: str) -> dict:
+        s = summarize_split.get(key, {})
+        return {"done_1h": int(s.get("done_1h", 0)), "done_today": int(s.get("done_today", 0))}
+    return {"macmini": m("macmini"), "macbook": m("macbook")}
+
+
 def build_trend(
     inflow_buckets: dict[str, int],
     done_buckets: dict[str, int],
@@ -292,6 +302,7 @@ def compose_overview(
         ),
         "stages": build_stages(stage_stats),
         "summarize_eta": build_summarize_eta(stage_stats),
+        "summarize_by_machine": build_summarize_by_machine(summarize_split),
         "trend_24h": build_trend(inflow_buckets, done_buckets, now_kst),
         "totals": build_totals(stage_stats),
     }

app/services/search/fusion_service.py

@@ -72,6 +72,10 @@ class LegacyWeightedSum(FusionStrategy):
                     score=existing.score + r.score * 0.5,
                     snippet=existing.snippet,
                     match_reason=f"{existing.match_reason}+vector",
+                    # C-1: 분류 축 메타 전파 (재구성 시 누락 = D-1 유형 표시 None)
+                    material_type=existing.material_type,
+                    jurisdiction=existing.jurisdiction,
+                    published_date=existing.published_date,
                 )
             elif r.score > 0.3:
                 merged[r.id] = r
@@ -128,6 +132,10 @@ class RRFOnly(FusionStrategy):
                     score=rrf_score,
                     snippet=base.snippet,
                     match_reason="+".join(reasons),
+                    # C-1: 분류 축 메타 전파 (재구성 시 누락 = D-1 유형 표시 None)
+                    material_type=base.material_type,
+                    jurisdiction=base.jurisdiction,
+                    published_date=base.published_date,
                 )
             )
         return merged[:limit]

app/services/search/license_filter.py

@@ -0,0 +1,28 @@
+"""안전 자료실 B-4 — licensed_restricted 단일 술어 (a안 U-2①, 모든 경로 공유 정의).
+
+색인은 허용하되 restricted=true(구매 전자책·유료자료)의 verbatim span 이 RAG 증거·발행물
+(검색/ask·digest·morning_briefing·study 풀이)에 들어가는 모든 경로를 구조적으로 차단.
+경로마다 술어를 복붙하지 않고 이 한 정의를 공유 — 가드 누락/드리프트 방지
+([[feedback_structural_integrity_over_path_discipline]]).
+개인 파일 열람(GET /documents/{id}?download)은 a안상 허용 = 미적용.
+
+두 표현(raw SQL / ORM)은 의미 동일: restricted 부재·false·extract_meta NULL = COALESCE 로
+미제외(redistribute=false 여도 restricted 부재면 미제외 — redistribute≠restricted 가 핵심).
+"""
+
+
+def restricted_exclude_sql(alias: str = "") -> str:
+    """raw text() 쿼리용 bare 술어('AND' 미포함). alias='' = 컬럼 직접 참조."""
+    p = (alias + ".") if alias else ""
+    return f"COALESCE({p}extract_meta -> 'license' ->> 'restricted', 'false') <> 'true'"
+
+
+def restricted_exclude_orm():
+    """SQLAlchemy ORM .where() 절 — restricted_exclude_sql 과 동일 의미(JSONB extract_meta)."""
+    from sqlalchemy import func
+
+    from models.document import Document
+
+    return func.coalesce(
+        Document.extract_meta["license"]["restricted"].astext, "false"
+    ) != "true"

app/services/search/result_decorate.py

@@ -0,0 +1,55 @@
+"""안전 자료실 C-1 후속 — 검색 결과 wrapper decoration (version_status + facets).
+
+엔드포인트 wrapper 에서 run_search() 결과에 1회 적용 — 검색 코어(run_search) 무접촉(r3).
+- version_status: 법령 결과(material_type='law')에 legal_meta.version_status
+  (current/superseded/pending/repealed) 부착. legal_meta.document_id 1:0..1 위성 →
+  매핑 없는 law(레거시 등)는 None 유지. law 결과 없으면 query skip.
+- facets: top-K 결과 내 분류 축(material_type/jurisdiction/version_status) 분포 라벨(r2-M4).
+  facets=true 일 때만 계산(미요청 시 None = byte 불변·ranking 무관).
+"""
+
+from __future__ import annotations
+
+from collections import Counter
+from typing import TYPE_CHECKING
+
+from sqlalchemy import text
+from sqlalchemy.ext.asyncio import AsyncSession
+
+if TYPE_CHECKING:
+    from api.search import SearchResult
+
+
+async def decorate_version_status(
+    session: AsyncSession, results: list["SearchResult"]
+) -> None:
+    """법령 결과에 legal_meta.version_status 부착 (in-place). law 결과 없으면 query skip."""
+    law_ids = [r.id for r in results if r.material_type == "law" and r.id is not None]
+    if not law_ids:
+        return
+    rows = await session.execute(
+        text(
+            "SELECT document_id, version_status FROM legal_meta "
+            "WHERE document_id = ANY(:ids)"
+        ),
+        {"ids": law_ids},
+    )
+    status_by_id = {row.document_id: row.version_status for row in rows}
+    for r in results:
+        if r.id in status_by_id:
+            r.version_status = status_by_id[r.id]
+
+
+def compute_facets(results: list["SearchResult"]) -> dict[str, dict[str, int]]:
+    """top-K 결과의 분류 축 분포 라벨. None 값은 제외(present 라벨만, 빈 축은 미포함)."""
+    axes = {
+        "material_type": [r.material_type for r in results],
+        "jurisdiction": [r.jurisdiction for r in results],
+        "version_status": [getattr(r, "version_status", None) for r in results],
+    }
+    facets: dict[str, dict[str, int]] = {}
+    for axis, vals in axes.items():
+        counter = Counter(v for v in vals if v is not None)
+        if counter:
+            facets[axis] = dict(counter.most_common())
+    return facets

app/services/search/retrieval_service.py

@@ -24,6 +24,7 @@ import asyncio
 import hashlib
 import re
 import time
+from dataclasses import dataclass
 from typing import TYPE_CHECKING, Any
 
 from sqlalchemy import text
@@ -98,6 +99,63 @@ QWEN3_QUERY_INSTRUCT = (
     "\nQuery: "
 )
 
+# ─── 안전 자료실 C-1: 분류 축 명시 필터 (3 leg 동등, byte 불변) ───────────────
+# 미지정(active=False) 시 모든 SQL 절이 빈 문자열 → 기존 SQL byte 불변(run_eval 회귀 0).
+# year 는 published_date NULL fallback created_at (freshness 와 동일 COALESCE 사상).
+@dataclass
+class AxisFilter:
+    material_types: list[str] | None = None   # CSV → list, material_type = ANY
+    jurisdiction: str | None = None
+    year_from: int | None = None
+    year_to: int | None = None
+
+    def active(self) -> bool:
+        return bool(self.material_types or self.jurisdiction
+                    or self.year_from is not None or self.year_to is not None)
+
+
+def _axis_sql(alias: str, af: "AxisFilter | None", params: dict) -> str:
+    """alias 기준 axis 필터 SQL — 미지정 시 '' (byte 불변). 반환 형태 ' AND ...'.
+
+    alias='' 이면 컬럼 직접 참조(단일 테이블 FROM documents 경로). 파라미터는 af_ prefix
+    로 호출측 기존 bind 와 충돌 방지.
+    """
+    if af is None or not af.active():
+        return ""
+    p = (alias + ".") if alias else ""
+    cl: list[str] = []
+    if af.material_types:
+        cl.append(f"{p}material_type = ANY(:af_mt)")
+        params["af_mt"] = af.material_types
+    if af.jurisdiction:
+        cl.append(f"{p}jurisdiction = :af_jur")
+        params["af_jur"] = af.jurisdiction
+    if af.year_from is not None:
+        cl.append(f"COALESCE({p}published_date, {p}created_at::date) >= make_date(:af_yf, 1, 1)")
+        params["af_yf"] = af.year_from
+    if af.year_to is not None:
+        cl.append(f"COALESCE({p}published_date, {p}created_at::date) <= make_date(:af_yt, 12, 31)")
+        params["af_yt"] = af.year_to
+    return " AND " + " AND ".join(cl)
+
+
+# ─── 안전 자료실 B-4: licensed_restricted 단일 술어 (a안 U-2① — 항상 적용) ──────
+def _license_sql(alias: str) -> str:
+    """licensed_restricted(extract_meta.license.restricted=true) 문서를 retrieval 에서 제외.
+
+    a안: 색인은 허용하되, 구매 전자책/유료자료의 verbatim span 이 RAG 증거·digest 발행에
+    들어가는 경로를 구조적으로 차단. 이 단일 술어를 모든 retrieval leg + digest loader 가
+    공유 — 경로별 가드 누락 방지([[feedback_structural_integrity_over_path_discipline]]).
+    개인 파일 열람(GET /documents/{id}?download)은 a안상 허용이라 미적용.
+
+    axis 필터(조건부)와 달리 항상 적용. restricted 부재/false = COALESCE 로 미제외 →
+    기존 코퍼스(restricted=true 0건)에서 결과 불변. 반환 ' AND ...' (alias='' = 컬럼 직접).
+    술어 정의 = license_filter.restricted_exclude_sql 공유(digest/briefing/study 풀이와 단일 source).
+    """
+    from services.search.license_filter import restricted_exclude_sql
+    return " AND " + restricted_exclude_sql(alias)
+
+
 # 2단계 gate (R2-B1) — SQL string interpolation 직전 final allowlist.
 _VALID_DOCS_TABLE = re.compile(r"^(documents|documents_cand_[a-z0-9_]+)$")
 # corpus_chunks = document_chunks WHERE in_corpus=true 뷰 (Hier-Decomp-1 c2 choke point).
@@ -235,7 +293,7 @@ def query_embed_cache_stats() -> dict[str, int]:
 
 
 async def search_text(
-    session: AsyncSession, query: str, limit: int
+    session: AsyncSession, query: str, limit: int, *, axis: "AxisFilter | None" = None
 ) -> list["SearchResult"]:
     """FTS + trigram 필드별 가중치 검색 (Phase 1.2-B UNION 분해).
 
@@ -266,8 +324,12 @@ async def search_text(
     # SQLAlchemy async session 내 두 execute는 같은 connection 사용
     await session.execute(text("SELECT set_limit(0.15)"))
 
+    _params: dict[str, Any] = {"q": query, "limit": limit}
+    # license(항상) + axis(조건부). license 가 항상 ' AND ...' 이라 WHERE 는 늘 존재.
+    _where = _license_sql("d") + _axis_sql("d", axis, _params)
+
     result = await session.execute(
-        text("""
+        text(f"""
             WITH candidates AS (
                 -- title trigram (idx_documents_title_trgm)
                 SELECT id FROM documents
@@ -320,13 +382,15 @@ async def search_text(
                        WHEN similarity(coalesce(d.ai_summary, ''), :q) >= 0.3 THEN 'summary'
                        WHEN similarity(coalesce(d.extracted_text, ''), :q) >= 0.3 THEN 'content'
                        ELSE 'fts'
-                   END AS match_reason
+                   END AS match_reason,
+                   d.material_type, d.jurisdiction, d.published_date
             FROM documents d
             JOIN candidates c ON d.id = c.id
+            WHERE{_where[4:]}
             ORDER BY score DESC
             LIMIT :limit
         """),
-        {"q": query, "limit": limit},
+        _params,
     )
     return [SearchResult(**row._mapping) for row in result]
 
@@ -341,6 +405,7 @@ async def search_vector(
     snapshot_chunk_id_max: int | None = None,
     corpus_variant: str | None = None,
     exact_knn: bool = False,
+    axis: "AxisFilter | None" = None,
 ) -> list["SearchResult"]:
     """Hybrid 벡터 검색 — doc + chunks 동시 retrieval (Phase 1.2-G).
 
@@ -415,6 +480,7 @@ async def search_vector(
                 docs_table=docs_table,
                 snapshot_doc_id_max=snapshot_doc_id_max,
                 exact_knn=exact_knn,
+                axis=axis,
             )
 
     async def _chunks_call() -> list["SearchResult"]:
@@ -424,6 +490,7 @@ async def search_vector(
                 chunks_table=chunks_table,
                 snapshot_chunk_id_max=snapshot_chunk_id_max,
                 exact_knn=exact_knn,
+                axis=axis,
             )
 
     doc_results, chunk_results = await asyncio.gather(_docs_call(), _chunks_call())
@@ -439,6 +506,7 @@ async def _search_vector_docs(
     docs_table: str = "documents",
     snapshot_doc_id_max: int | None = None,
     exact_knn: bool = False,
+    axis: "AxisFilter | None" = None,
 ) -> list["SearchResult"]:
     """documents (또는 documents_cand_<slug>).embedding 직접 검색.
 
@@ -463,28 +531,34 @@ async def _search_vector_docs(
         if snapshot_doc_id_max is not None:
             snapshot_clause = " AND id <= :snapshot_doc_id_max"
             params["snapshot_doc_id_max"] = snapshot_doc_id_max
+        axis_clause = _axis_sql("", axis, params)   # alias 없음 (단일 FROM documents)
+        license_clause = _license_sql("")            # B-4: restricted 항상 제외
         sql = f"""
             SELECT id, title, ai_domain, ai_summary, file_format,
                    (1 - (embedding <=> cast(:embedding AS vector))) AS score,
                    left(extracted_text, 1200) AS snippet,
                    'vector_doc' AS match_reason,
-                   NULL::bigint AS chunk_id, NULL::integer AS chunk_index, NULL::text AS section_title
+                   NULL::bigint AS chunk_id, NULL::integer AS chunk_index, NULL::text AS section_title,
+                   material_type, jurisdiction, published_date
             FROM documents
-            WHERE embedding IS NOT NULL AND deleted_at IS NULL{snapshot_clause}
+            WHERE embedding IS NOT NULL AND deleted_at IS NULL{snapshot_clause}{axis_clause}{license_clause}
             ORDER BY embedding <=> cast(:embedding AS vector)
             LIMIT :limit
         """
     else:
         # candidate: docs_table 은 (doc_id, embed_input, embed_input_hash, embedding) 만 보유 → JOIN documents
+        axis_clause = _axis_sql("d", axis, params)
+        license_clause = _license_sql("d")           # B-4: restricted 항상 제외
         sql = f"""
             SELECT d.id, d.title, d.ai_domain, d.ai_summary, d.file_format,
                    (1 - (c.embedding <=> cast(:embedding AS vector))) AS score,
                    left(d.extracted_text, 1200) AS snippet,
                    'vector_doc' AS match_reason,
-                   NULL::bigint AS chunk_id, NULL::integer AS chunk_index, NULL::text AS section_title
+                   NULL::bigint AS chunk_id, NULL::integer AS chunk_index, NULL::text AS section_title,
+                   d.material_type, d.jurisdiction, d.published_date
             FROM {docs_table} c
             JOIN documents d ON d.id = c.doc_id
-            WHERE d.deleted_at IS NULL
+            WHERE d.deleted_at IS NULL{axis_clause}{license_clause}
             ORDER BY c.embedding <=> cast(:embedding AS vector)
             LIMIT :limit
         """
@@ -500,6 +574,7 @@ async def _search_vector_chunks(
     chunks_table: str = "document_chunks",
     snapshot_chunk_id_max: int | None = None,
     exact_knn: bool = False,
+    axis: "AxisFilter | None" = None,
 ) -> list["SearchResult"]:
     """document_chunks (또는 document_chunks_cand_<slug>).embedding window partition.
 
@@ -525,12 +600,25 @@ async def _search_vector_chunks(
         snapshot_clause = " AND c.id <= :snapshot_chunk_id_max"
         params["snapshot_chunk_id_max"] = snapshot_chunk_id_max
 
+    # C-1: axis 필터는 inner topk 에 JOIN (R6 결정 — outer post-filter 면 ANN top-:inner_k
+    # 후보를 뽑은 뒤 거르므로 좁은 필터(GB 법령 등)에서 후보 붕괴). 미지정 시 JOIN 없음 = byte 불변.
+    if axis and axis.active():
+        chunk_join = " JOIN documents df ON df.id = c.doc_id"
+        chunk_axis = _axis_sql("df", axis, params)
+    else:
+        chunk_join = ""
+        chunk_axis = ""
+
+    # B-4: restricted 제외 — outer 가 documents d 를 항상 JOIN 하므로 post-rank 위치.
+    # restricted 는 소수(구매자료)라 inner topk 후 제외해도 candidate collapse 없음(axis 와 상이).
+    license_clause = _license_sql("d")
+
     sql = f"""
         WITH topk AS (
             SELECT c.id AS chunk_id, c.doc_id, c.chunk_index, c.section_title, c.text,
                    c.embedding <=> cast(:embedding AS vector) AS dist
-            FROM {chunks_table} c
-            WHERE c.embedding IS NOT NULL{snapshot_clause}
+            FROM {chunks_table} c{chunk_join}
+            WHERE c.embedding IS NOT NULL{snapshot_clause}{chunk_axis}
             ORDER BY c.embedding <=> cast(:embedding AS vector)
             LIMIT :inner_k
         ),
@@ -543,10 +631,12 @@ async def _search_vector_chunks(
                d.ai_summary AS ai_summary, d.file_format AS file_format,
                (1 - r.dist) AS score, left(r.text, 1200) AS snippet,
                'vector_chunk' AS match_reason,
-               r.chunk_id AS chunk_id, r.chunk_index AS chunk_index, r.section_title AS section_title
+               r.chunk_id AS chunk_id, r.chunk_index AS chunk_index, r.section_title AS section_title,
+               d.material_type AS material_type, d.jurisdiction AS jurisdiction,
+               d.published_date AS published_date
         FROM ranked r
         JOIN documents d ON d.id = r.doc_id
-        WHERE r.rn <= 2 AND d.deleted_at IS NULL
+        WHERE r.rn <= 2 AND d.deleted_at IS NULL{license_clause}
         ORDER BY r.dist
         LIMIT :limit
     """

app/services/search/search_pipeline.py

@@ -47,6 +47,7 @@ from .rerank_service import (
     rerank_chunks,
 )
 from .retrieval_service import (
+    AxisFilter,
     compress_chunks_to_docs,
     search_text,
     search_vector,
@@ -148,6 +149,7 @@ async def run_search(
     rewrite_backend: str | None = None,
     corpus_variant: str | None = None,
     exact_knn: bool = False,
+    axis: AxisFilter | None = None,
 ) -> PipelineResult:
     """검색 파이프라인 실행.
 
@@ -275,6 +277,7 @@ async def run_search(
                 snapshot_chunk_id_max=snapshot_chunk_id_max,
                 corpus_variant=corpus_variant,
                 exact_knn=exact_knn,
+                axis=axis,
             )
         timing["vector_ms"] = (time.perf_counter() - t0) * 1000
         if not raw_chunks:
@@ -284,7 +287,7 @@ async def run_search(
         results = vector_results
     else:
         t0 = time.perf_counter()
-        text_results = await search_text(session, q, limit)
+        text_results = await search_text(session, q, limit, axis=axis)
         timing["text_ms"] = (time.perf_counter() - t0) * 1000
 
         if mode == "hybrid":
@@ -306,6 +309,7 @@ async def run_search(
                     snapshot_chunk_id_max=snapshot_chunk_id_max,
                     corpus_variant=corpus_variant,
                     exact_knn=exact_knn,
+                    axis=axis,
                 )
             timing["vector_ms"] = (time.perf_counter() - t1) * 1000
 
@@ -458,6 +462,10 @@ def _rrf_fuse_variants(
             score=rrf_score,
             snippet=doc.snippet,
             match_reason=f"{doc.match_reason}+multi_query_rrf",
+            # C-1: 분류 축 메타 전파 (SearchResult 재구성 지점 — fusion 2곳과 동기)
+            material_type=doc.material_type,
+            jurisdiction=doc.jurisdiction,
+            published_date=doc.published_date,
         ))
     return fused[:limit]
 

app/services/study/explanation_rag.py

@@ -24,6 +24,7 @@ from models.chunk import DocumentChunk
 from models.document import Document
 from models.study_question import StudyQuestion
 from models.study_topic import StudyTopicDocument
+from services.search.license_filter import restricted_exclude_orm
 
 logger = logging.getLogger(__name__)
 
@@ -124,11 +125,14 @@ async def _gather_document_evidence(
         return []
 
     # 매핑된 documents 메타 (제목·요약 표기)
+    # B-4: licensed_restricted 제외 → valid_doc_ids 에서 빠지므로 아래 청크 쿼리(doc_id IN)도
+    # 자동 차단. study 풀이 RAG 도 retrieval/digest 와 동일 단일 술어 공유(a안 U-2①).
     doc_meta_rows = (
         await session.execute(
             select(Document.id, Document.title, Document.ai_summary).where(
                 Document.id.in_(doc_ids),
                 Document.deleted_at.is_(None),
+                restricted_exclude_orm(),
             )
         )
     ).all()

app/workers/api_standards_collector.py

@@ -175,10 +175,16 @@ async def _ingest_detail(session, source: NewsSource, url: str) -> str:
         ai_domain="Engineering",
         ai_sub_group=_SOURCE_NAME,
         ai_tags=["Engineering/API 표준 공지"],
+        # 안전 자료실 A-2 — 표준 '공지' = standard (코드 본문 아님 — ASME/API 본문은 paywall)
+        material_type="standard",
+        jurisdiction="US",
+        published_date=pub_dt.date() if pub_dt else None,
         extract_meta={
             "source_id": source.id,
             "source_name": _SOURCE_NAME,
             "published_at": pub_dt.isoformat() if pub_dt else None,
+            "license": {"scheme": "proprietary", "redistribute": False,
+                        "attribution": "American Petroleum Institute"},
             "fulltext": {
                 "status": "api_announcement",
                 "engine": engine,

app/workers/chunk_worker.py

@@ -311,6 +311,10 @@ async def process(document_id: int, session: AsyncSession) -> None:
     country, source, src_lang = await _lookup_news_source(session, doc)
     if src_lang:
         language = src_lang
+    # 안전 자료실 A-2 — 뉴스 lookup 미해당(crawl/law/업로드) 문서는 jurisdiction 을
+    # chunk.country 미러로 (leg 간 국가 일치. EU/INT 도 이 경로로 첫 유입 — String(10) 수용).
+    if country is None and doc.jurisdiction:
+        country = doc.jurisdiction
     domain_category = "news" if doc.source_channel == "news" else "document"
 
     # 기존 chunks 삭제 (재처리)

app/workers/classify_worker.py

@@ -62,6 +62,15 @@ FACET_DOCTYPES = {"발주서", "세금계산서", "명세표", "도면", "증명
 # 자료실 자동 분류 제안 대상 (거래 하위)
 LIBRARY_SUGGESTION_DOCTYPES = {"발주서", "세금계산서", "명세표"}
 
+# 안전 자료실 A-2 — document_type → material_type 결정적 매핑 (제안 전용, 자동 전이 금지).
+# 모호한 doctype(Reference/Report 등)은 매핑하지 않음 — 무리한 전수 분류 시도 금지 (plan 0-1).
+_DOCTYPE_TO_MATERIAL = {
+    "Law_Document": "law",
+    "Academic_Paper": "paper",
+    "Manual": "manual",
+    "Standard": "standard",
+}
+
 # PR-B prompt_version task 이름
 SUMMARY_TRIAGE_TASK = "p3a_short_summary"
 
@@ -492,6 +501,24 @@ async def process(
         if not doc.document_type:
             doc.document_type = doc_type if doc_type in DOCUMENT_TYPES else "Note"
 
+        # ─── 안전 자료실 A-2: material_type 제안 (업로드 경로 — LLM 직접 부여 금지) ───
+        # document_type → material_type 결정적 매핑만 제안으로 적재 (프롬프트 변경 0).
+        # 승인(accept-suggestion) 시에만 전이 — law 는 국가 필수 입력 (KR 기본값 오염 차단,
+        # 자동 전이 금지 사상은 category 와 동일). 수집기 deterministic 경로는 이미 채워져
+        # 있어(material_type IS NOT NULL) 본 제안 비대상. 거래문서 제안(ai_suggestion 점유)과
+        # 충돌 시 기존 제안 우선 (두 제안이 겹치는 문서는 실무상 없음 — 거래 vs 안전자료).
+        _mt_prop = _DOCTYPE_TO_MATERIAL.get(doc.document_type or "")
+        if _mt_prop and doc.material_type is None and doc.ai_suggestion is None:
+            doc.ai_suggestion = {
+                "proposed_material_type": _mt_prop,
+                "proposed_jurisdiction": None,
+                "confidence": doc.ai_confidence,
+                "source_updated_at": (
+                    doc.updated_at.isoformat() if doc.updated_at else None
+                ),
+                "reason": "document_type→material_type 결정적 매핑",
+            }
+
         # confidence
         confidence = parsed.get("confidence", 0.5)
         doc.ai_confidence = max(0.0, min(1.0, float(confidence)))

app/workers/csb_collector.py

@@ -202,7 +202,12 @@ async def _ingest_pdf(session, page_slug: str, pdf_url: str) -> bool:
         import_source="csb_sitemap",
         edit_url=pdf_url,
         ai_tags=["Safety/CSB/보고서"],
-        extract_meta={"csb": {"page_slug": page_slug, "kind": "report_pdf"}},
+        # 안전 자료실 A-2 — ingest 시점 deterministic. CSB = 미 연방기관 = public domain.
+        material_type="incident",
+        jurisdiction="US",
+        extract_meta={"csb": {"page_slug": page_slug, "kind": "report_pdf"},
+                      "license": {"scheme": "public_domain", "redistribute": True,
+                                  "attribution": "U.S. Chemical Safety Board"}},
     )
     session.add(doc)
     await session.flush()
@@ -290,10 +295,16 @@ async def _ingest_url(session, source: NewsSource, url: str, lastmod: datetime)
         ai_domain="Safety",
         ai_sub_group=_SOURCE_NAME,
         ai_tags=["Safety/CSB"],
+        # 안전 자료실 A-2 — ingest 시점 deterministic (classify-skip 경로)
+        material_type="incident",
+        jurisdiction="US",
+        published_date=lastmod.date() if lastmod else None,
         extract_meta={
             "source_id": source.id,
             "source_name": _SOURCE_NAME,
             "published_at": lastmod.isoformat(),
+            "license": {"scheme": "public_domain", "redistribute": True,
+                        "attribution": "U.S. Chemical Safety Board"},
             "fulltext": {
                 "status": "csb_sitemap",
                 "engine": engine,

app/workers/file_watcher.py

@@ -58,6 +58,23 @@ SCAN_TARGETS: list[tuple[str, str | None]] = [
     ("Videos", "video"),
 ]
 
+# 안전 자료실 A-2/B-4 — watch 타깃별 (material_type, jurisdiction, license) deterministic 축.
+# 키 = 타깃 경로의 마지막 성분. license = extract_meta.license 에 그대로 주입(None=미주입).
+#   restricted=true → retrieval_service._license_sql 가 RAG 증거·digest 에서 제외(a안 U-2① —
+#   구매자료 verbatim span 차단, 색인 자체는 허용. 개인 파일 열람은 미차단).
+# 사용자 결정(2026-06-13): Books/Papers=proprietary+restricted / Manuals=proprietary·restricted=false
+#   (검색·RAG 활용) / KGS=법정 위임 상세기준 law/KR·KOGL 공공·restricted 아님.
+_TARGET_AXIS: dict[str, tuple[str, str | None, dict | None]] = {
+    "KGS_Code": ("law", "KR", {"scheme": "kogl", "redistribute": True,
+                               "restricted": False, "attribution": "한국가스안전공사(KGS)"}),
+    "Books": ("book", None, {"scheme": "proprietary", "redistribute": False,
+                             "restricted": True, "attribution": "구매 도서"}),
+    "Papers_Purchased": ("paper", None, {"scheme": "proprietary", "redistribute": False,
+                                         "restricted": True, "attribution": "구매 논문"}),
+    "Manuals": ("manual", None, {"scheme": "proprietary", "redistribute": False,
+                                 "restricted": False, "attribution": "기술 매뉴얼"}),
+}
+
 
 def should_skip(path: Path) -> bool:
     if path.name in SKIP_NAMES or path.name.startswith("._"):
@@ -242,6 +259,11 @@ async def watch_inbox():
             if not scan_root.exists():
                 continue
 
+            # 안전 자료실 A-2/B-4 — 타깃 폴더 기반 (material, jurisdiction, license)
+            target_mt, target_jur, target_license = _TARGET_AXIS.get(
+                Path(sub).name, (None, None, None)
+            )
+
             for file_path in scan_root.rglob("*"):
                 if not file_path.is_file() or should_skip(file_path):
                     continue
@@ -275,7 +297,14 @@ async def watch_inbox():
                         source_channel="drive_sync",
                         category=category,
                         needs_conversion=needs_conversion,
+                        # 안전 자료실 A-2/B-4 — watch 타깃 매핑 (KGS=law/KR 등, 비대상=NULL)
+                        material_type=target_mt,
+                        jurisdiction=target_jur,
                     )
+                    # B-4 — 타깃 폴더 license 주입(restricted 포함, 비대상=미주입). classify 는
+                    # material_type IS NULL 일 때만 제안 + extract_meta 미기록이라 주입 보존.
+                    if target_license:
+                        doc.extract_meta = {"license": dict(target_license)}
                     session.add(doc)
                     await session.flush()
 
@@ -291,6 +320,15 @@ async def watch_inbox():
                         existing.category = category
                     if needs_conversion and not getattr(existing, "needs_conversion", False):
                         existing.needs_conversion = True
+                    # B-4 — 축/license 보정(B-4 이전 적재분이 재변경 시): material 미설정 시 주입,
+                    # license 부재 시에만 merge 주입(clobber 회피 — 기존 extract_meta 키 보존).
+                    if existing.material_type is None and target_mt is not None:
+                        existing.material_type = target_mt
+                        existing.jurisdiction = target_jur
+                    if target_license and not (existing.extract_meta or {}).get("license"):
+                        meta = dict(existing.extract_meta or {})
+                        meta["license"] = dict(target_license)
+                        existing.extract_meta = meta
 
                     if next_stage:
                         await enqueue_stage(session, existing.id, next_stage)

app/workers/kosha_collector.py

@@ -1,14 +1,17 @@
 """C-2 KOSHA Open API 수집 워커 (plan crawl-24x7-1).
 
-3 API (2026-06-10 실키 live 검증 + fixture 박제 — tests/fixtures/kosha_*_response.json):
+4 API (2026-06-10/06-13 실키 live 검증 + fixture 박제 — tests/fixtures/kosha_*_response.json):
   재해사례 게시판: GET /B552468/disaster_api02/getdisaster_api02            callApiId=1060
   재해사례 첨부:   GET /B552468/disaster_attach_api02/Disaster_attach_api02 callApiId=1070
   KOSHA GUIDE:    GET /B552468/koshaguide/getKoshaGuide                    callApiId=1050
+  사망사고 속보:   GET /B552468/news_api02/getNews_api02                     callApiId=1040
 
 daily 스케줄 1회 (main.py):
   재해사례 = 최근 페이지만 diff (boardno dedup) — 사례 본문 Document(텍스트 네이티브)
             + 첨부 PDF/HWP 다운로드 → /documents/crawl_raw/kosha/{boardno}/ 저장
             → 파일 Document + extract enqueue (kordoc HWP/PDF 기존 파이프라인 재사용).
+  사망사고 = 최근 페이지만 diff (arno dedup) — 속보 본문 Document(HTML → _clean_html).
+            첨부 API 없음·business 필드 없음. 등록일 = arno 접두 8자리(YYYYMMDD).
   GUIDE   = 전체 레지스트리 메타 diff (1039건, 100/page = 11 call) → 신규/개정만,
             일일 ingest cap(기본 25) = backlog 자동 점진 백필(~6주) + 부하 평탄화.
             cap 으로 미처리 잔량은 매회 로그 (silent cap 금지).
@@ -23,7 +26,7 @@ import hashlib
 import os
 import random
 import re
-from datetime import datetime, timezone
+from datetime import date, datetime, timezone
 from pathlib import Path
 
 import httpx
@@ -38,6 +41,7 @@ from models.news_source import NewsSource
 from models.queue import enqueue_stage
 from workers.news_collector import (
     FeedError,
+    _clean_html,
     _get_or_create_health,
     _record_failure,
     _record_success,
@@ -49,17 +53,36 @@ _BASE = "https://apis.data.go.kr/B552468"
 _BOARD_EP = f"{_BASE}/disaster_api02/getdisaster_api02"
 _ATTACH_EP = f"{_BASE}/disaster_attach_api02/Disaster_attach_api02"
 _GUIDE_EP = f"{_BASE}/koshaguide/getKoshaGuide"
+_FATAL_EP = f"{_BASE}/news_api02/getNews_api02"
 
 _CASE_SOURCE = "KOSHA 재해사례"
 _GUIDE_SOURCE = "KOSHA GUIDE"
+_FATAL_SOURCE = "KOSHA 사망사고"
 
 _CASE_PAGES = 2          # daily diff 범위 (30×2 = 최근 60건 — 등록일 역순 API)
 _CASE_ROWS = 30
+_FATAL_PAGES = 2         # 사망사고 속보 daily diff (30×2 = 최근 60건 — 등록일 역순)
+_FATAL_ROWS = 30
 _GUIDE_ROWS = 100
 _GUIDE_DAILY_CAP = int(os.getenv("KOSHA_GUIDE_DAILY_CAP", "25"))
 _MAX_FILE_BYTES = 50 * 1024 * 1024
 _DOWNLOAD_DELAY = (2.0, 5.0)  # portal.kosha.or.kr 파일서버 — 연속 다운로드 간격
 
+# 안전 자료실 A-2 — KOSHA 산출물 라이선스 (KOGL 유형 미확정 → 보수적 redistribute=False,
+# 근거 확보 시 완화. 0-3 license 메타 deterministic 주입).
+_KOSHA_LICENSE = {"scheme": "kogl", "redistribute": False, "attribution": "한국산업안전보건공단(KOSHA)"}
+
+
+def _ymd_to_date(ymd: str | None) -> date | None:
+    """'YYYYMMDD'/'YYYY-MM-DD' → date. 형식 불일치는 None (fail-quiet — 날짜는 보조 축)."""
+    digits = re.sub(r"\D", "", ymd or "")
+    if len(digits) != 8:
+        return None
+    try:
+        return date(int(digits[:4]), int(digits[4:6]), int(digits[6:8]))
+    except ValueError:
+        return None
+
 
 def _api_key() -> str:
     key = os.getenv("KOSHA_API_KEY", "")
@@ -93,6 +116,29 @@ def _items(payload: dict) -> list[dict]:
     return [item] if isinstance(item, dict) else list(item)
 
 
+def _fatal_fields(item: dict) -> dict | None:
+    """사망사고 item(arno/keyword/contents 3필드 고정) → Document 필드 매핑.
+
+    순수 함수(httpx/DB 불요 — fixture 단위 테스트 대상). 필수 = arno+keyword,
+    부재 시 None(skip). 날짜 전용 필드가 없어 등록 식별자 arno 접두에서 유도:
+    arno = 'YYYYMMDDHHMMSS' + 임의 6자 (2019~ 라이브 전수 동형 검증). 접두 8자리=KST
+    등록일 → published_date, 14자리=등록시각 → reg_dt(원문 그대로, tz 해석 미주장).
+    """
+    arno = str(item.get("arno") or "").strip()
+    title = (item.get("keyword") or "").strip()
+    if not arno or not title:
+        return None
+    text = _clean_html(item.get("contents") or "", max_len=None)
+    reg_dt = arno[:14] if re.fullmatch(r"\d{14}", arno[:14]) else None
+    return {
+        "arno": arno,
+        "title": title,
+        "text": text,
+        "published_date": _ymd_to_date(arno[:8]),
+        "reg_dt": reg_dt,
+    }
+
+
 def _safe_filename(name: str) -> str:
     """NAS 파일명 정화 — 경로분리자/제어문자/공백연쇄 제거 (쉘 함정 회피)."""
     name = re.sub(r"[/\\\x00-\x1f]", "_", name).strip()
@@ -155,7 +201,11 @@ async def _ingest_attachment(session, boardno: str, filenm: str, filepath: str)
         import_source="kosha_api",
         edit_url=filepath,
         ai_tags=["Safety/KOSHA재해사례/첨부"],
-        extract_meta={"kosha": {"boardno": boardno, "kind": "case_attachment"}},
+        # 안전 자료실 A-2 — ingest 시점 deterministic (classify 경유해도 LLM 비의존)
+        material_type="incident",
+        jurisdiction="KR",
+        extract_meta={"kosha": {"boardno": boardno, "kind": "case_attachment"},
+                      "license": dict(_KOSHA_LICENSE)},
     )
     session.add(doc)
     await session.flush()
@@ -213,12 +263,16 @@ async def collect_disaster_cases(session) -> int:
                 ai_domain="Safety",
                 ai_sub_group=_CASE_SOURCE,
                 ai_tags=[f"Safety/KOSHA재해사례/{business or '기타'}"],
+                # 안전 자료실 A-2 — ingest 시점 deterministic (classify-skip 경로)
+                material_type="incident",
+                jurisdiction="KR",
                 extract_meta={
                     "source_id": source.id,
                     "source_name": _CASE_SOURCE,
                     "published_at": None,
                     "kosha": {"boardno": boardno, "business": business,
                               "atcflcnt": item.get("atcflcnt")},
+                    "license": dict(_KOSHA_LICENSE),
                 },
             )
             session.add(doc)
@@ -250,6 +304,83 @@ async def collect_disaster_cases(session) -> int:
     return new_count
 
 
+async def collect_fatal_accidents(session) -> int:
+    """사망사고 속보 daily diff — 최근 _FATAL_PAGES 페이지, arno dedup.
+
+    재해사례(1060)와 별 채널(1040): business 필드·첨부 API 없음, contents=HTML.
+    본문 = 텍스트 네이티브(_clean_html) → md 변환 비대상, summarize/embed/chunk 큐.
+    """
+    key = _api_key()
+    source = await _get_or_create_source(session, _FATAL_SOURCE, _FATAL_EP)
+    new_count = 0
+
+    for page in range(1, _FATAL_PAGES + 1):
+        payload = await _api_get(
+            f"{_FATAL_EP}?serviceKey={key}&callApiId=1040&pageNo={page}&numOfRows={_FATAL_ROWS}"
+        )
+        items = _items(payload)
+        if not items:
+            break
+        page_all_dup = True
+        for item in items:
+            fields = _fatal_fields(item)
+            if fields is None:
+                continue
+            arno = fields["arno"]
+            fhash = hashlib.sha256(f"kosha-fatal|{arno}".encode()).hexdigest()[:32]
+            existing = await session.execute(
+                select(Document).where(Document.file_hash == fhash).limit(1)
+            )
+            if existing.scalars().first():
+                continue
+            page_all_dup = False
+
+            text = fields["text"]
+            now = datetime.now(timezone.utc)
+            doc = Document(
+                file_path=f"crawl/{_FATAL_SOURCE}/{arno}",
+                file_hash=fhash,
+                file_format="article",
+                file_size=len(text.encode()),
+                file_type="note",
+                title=fields["title"],
+                extracted_text=f"{fields['title']}\n\n{text}",
+                extracted_at=now,
+                extractor_version="kosha_api",
+                md_status="skipped",
+                md_extraction_error="kosha fatal: 텍스트 네이티브, markdown 변환 비대상",
+                source_channel="crawl",
+                data_origin="external",
+                review_status="approved",
+                ai_domain="Safety",
+                ai_sub_group=_FATAL_SOURCE,
+                ai_tags=["Safety/KOSHA사망사고"],
+                # 안전 자료실 A-2 — ingest 시점 deterministic (classify-skip 경로)
+                material_type="incident",
+                jurisdiction="KR",
+                published_date=fields["published_date"],
+                extract_meta={
+                    "source_id": source.id,
+                    "source_name": _FATAL_SOURCE,
+                    "published_at": None,
+                    "kosha": {"arno": arno, "kind": "fatal_accident",
+                              "reg_dt": fields["reg_dt"]},
+                    "license": dict(_KOSHA_LICENSE),
+                },
+            )
+            session.add(doc)
+            await session.flush()
+            await enqueue_stage(session, doc.id, "summarize")
+            await enqueue_stage(session, doc.id, "embed")
+            await enqueue_stage(session, doc.id, "chunk")
+            new_count += 1
+        if page_all_dup:
+            break  # 등록일 역순 — 페이지 전체가 기존이면 이후 페이지도 기존
+
+    logger.info(f"[kosha] 사망사고 신규 {new_count}건")
+    return new_count
+
+
 async def collect_kosha_guide(session, cap: int = _GUIDE_DAILY_CAP) -> int:
     """GUIDE 레지스트리 전체 메타 diff → 신규/개정만 다운로드 (일일 cap 점진 백필)."""
     key = _api_key()
@@ -307,8 +438,13 @@ async def collect_kosha_guide(session, cap: int = _GUIDE_DAILY_CAP) -> int:
             import_source="kosha_api",
             edit_url=spec["url"],
             ai_tags=["Safety/KOSHA GUIDE"],
+            # 안전 자료실 A-2 — GUIDE = 구속력 없는 권고 기술지침 (law 아님, plan 0-1)
+            material_type="guide",
+            jurisdiction="KR",
+            published_date=_ymd_to_date(spec["ymd"]),
             extract_meta={"kosha": {"kind": "guide", "techGdlnNo": spec["no"],
-                                    "ofancYmd": spec["ymd"]}},
+                                    "ofancYmd": spec["ymd"]},
+                          "license": dict(_KOSHA_LICENSE)},
         )
         session.add(doc)
         await session.flush()
@@ -325,6 +461,7 @@ async def run() -> None:
     """daily 1회 — 소스별 실패 격리 (재해사례 실패가 GUIDE 를 막지 않게)."""
     now = datetime.now(timezone.utc)
     for name, collector in ((_CASE_SOURCE, collect_disaster_cases),
+                            (_FATAL_SOURCE, collect_fatal_accidents),
                             (_GUIDE_SOURCE, collect_kosha_guide)):
         async with async_session() as session:
             result = await session.execute(select(NewsSource).where(NewsSource.name == name))

app/workers/law_monitor.py

@@ -6,7 +6,7 @@
 
 import os
 import re
-from datetime import datetime, timezone
+from datetime import date, datetime, timezone
 from pathlib import Path
 from xml.etree import ElementTree as ET
 
@@ -262,6 +262,16 @@ async def _save_law_split(
                 f"개정구분: {revision_type}"
             )
 
+        # 안전 자료실 A-2 — 공포일 파싱 (law published_date = COALESCE(시행일, 공포일) 계약,
+        # 본 레거시 워커는 공포일만 보유 — 시행일 기반 버전 체인은 B-1 statute_collector 소관)
+        _digits = re.sub(r"\D", "", str(proclamation_date or ""))
+        pub_date = None
+        if len(_digits) == 8:
+            try:
+                pub_date = date(int(_digits[:4]), int(_digits[4:6]), int(_digits[6:8]))
+            except ValueError:
+                pub_date = None
+
         doc = Document(
             file_path=rel_path,
             file_hash=file_hash(file_path),
@@ -272,6 +282,13 @@ async def _save_law_split(
             source_channel="law_monitor",
             data_origin="work",
             category="law",
+            # 안전 자료실 A-2 — ingest 시점 deterministic. 법령 텍스트 = 저작권법 제7조
+            # 비보호 저작물 (public domain). 본 워커는 휴면(LAW_OC 미설정)이나 코드 경로 유지.
+            material_type="law",
+            jurisdiction="KR",
+            published_date=pub_date,
+            extract_meta={"license": {"scheme": "public_domain", "redistribute": True,
+                                      "attribution": "국가법령정보센터"}},
             user_note=note or None,
         )
         session.add(doc)

app/workers/news_collector.py

@@ -341,11 +341,35 @@ def _entry_body(source: NewsSource, entry, summary: str) -> tuple[str, str]:
 
 def _build_extract_meta(source: NewsSource, pub_dt: datetime) -> dict:
     """fulltext_worker / 패널이 쓰는 출처 메타 (documents 에 source FK 가 없어 여기 기록)."""
-    return {
+    meta = {
         "source_id": source.id,
         "source_name": source.name,
         "published_at": pub_dt.isoformat(),
     }
+    # 안전 자료실 A-2: 소스 레지스트리의 라이선스를 deterministic 주입 (0-3 license 메타).
+    # P3 다이제스트/발행류가 redistribute=false 소스를 구조적으로 제외하는 게이트 입력.
+    if source.license_scheme:
+        meta["license"] = {
+            "scheme": source.license_scheme,
+            "redistribute": bool(source.license_redistribute),
+            "attribution": source.name,
+        }
+    return meta
+
+
+def _material_axis(source: NewsSource) -> tuple[str | None, str | None]:
+    """안전 자료실 분류 축 (material_type, jurisdiction) — 레지스트리 deterministic.
+
+    - material_type = news_sources.material_type (NULL = 비대상, 뉴스/철학 등)
+    - jurisdiction = source.country 전파. 단 paper 는 NULL 강제
+      (국제 학술지에 관할 개념 부적합 — plan 0-1 계약. 레지스트리 country=US 여도 미전파).
+    """
+    mt = source.material_type
+    if not mt:
+        return None, None
+    if mt == "paper":
+        return mt, None
+    return mt, source.country
 
 
 def _doc_identity(source: NewsSource, source_short: str, category: str) -> dict:
@@ -354,17 +378,22 @@ def _doc_identity(source: NewsSource, source_short: str, category: str) -> dict:
     file_path 접두사가 곧 채널 디렉토리. ai_domain 은 다이제스트/검색 필터의 분기 축이라
     crawl 채널이 'News' 를 오염시키지 않게 분리 (0-5 채널 레벨 분리 사상).
     """
+    material_type, jurisdiction = _material_axis(source)
     if source.source_channel == "crawl":
         domain = category if category and category != "Other" else "Domain"
         return {
             "path_prefix": "crawl",
             "ai_domain": domain,
             "ai_tags": [f"{domain}/{source_short}"],
+            "material_type": material_type,
+            "jurisdiction": jurisdiction,
         }
     return {
         "path_prefix": "news",
         "ai_domain": "News",
         "ai_tags": [f"News/{source_short}/{category}"],
+        "material_type": material_type,
+        "jurisdiction": jurisdiction,
     }
 
 
@@ -528,6 +557,10 @@ async def _fetch_rss(session, source: NewsSource) -> tuple[int, str]:
             ai_domain=ident["ai_domain"],
             ai_sub_group=source_short,
             ai_tags=ident["ai_tags"],
+            # 안전 자료실 A-2 — 레지스트리 deterministic (classify-skip 경로라 ingest 시점 필수)
+            material_type=ident["material_type"],
+            jurisdiction=ident["jurisdiction"],
+            published_date=pub_dt.date() if pub_dt else None,
             extract_meta=_build_extract_meta(source, pub_dt),
         )
         session.add(doc)
@@ -661,6 +694,10 @@ async def _fetch_api_guardian(session, source: NewsSource) -> tuple[int, str]:
             ai_domain=ident["ai_domain"],
             ai_sub_group=source_short,
             ai_tags=ident["ai_tags"],
+            # 안전 자료실 A-2 — 레지스트리 deterministic (classify-skip 경로라 ingest 시점 필수)
+            material_type=ident["material_type"],
+            jurisdiction=ident["jurisdiction"],
+            published_date=pub_dt.date() if pub_dt else None,
             extract_meta=_build_extract_meta(source, pub_dt),
         )
         session.add(doc)
@@ -757,6 +794,10 @@ async def _fetch_api_nyt(session, source: NewsSource) -> tuple[int, str]:
             ai_domain=ident["ai_domain"],
             ai_sub_group=source_short,
             ai_tags=ident["ai_tags"],
+            # 안전 자료실 A-2 — 레지스트리 deterministic (classify-skip 경로라 ingest 시점 필수)
+            material_type=ident["material_type"],
+            jurisdiction=ident["jurisdiction"],
+            published_date=pub_dt.date() if pub_dt else None,
             extract_meta=_build_extract_meta(source, pub_dt),
         )
         session.add(doc)

app/workers/statute_adapters/__init__.py

@@ -0,0 +1,43 @@
+"""statute_collector 나라별 어댑터 패키지 (plan safety-library-1 B-1).
+
+어댑터 계약 (2함수 + 상수):
+  JURISDICTION: str  — 어댑터 상수 고정. 코어가 적재 직전 assert (파싱 결과 추론 금지).
+  poll_changes(client, watch_rows) -> list[ChangeEvent]  — 개정 감지만 (경량 호출).
+  fetch_version(client, act, change) -> list[VersionPayload]  — PR②.
+    payload 리스트: primary + annex 각각 자기 version_key (R4-M4).
+
+ChangeEvent.kind: amend / repeal / bootstrap(합성 — PR② 부트스트랩이 amend 와
+동일 ingest 경로 재사용, R6-m2).
+"""
+
+from dataclasses import dataclass, field
+
+
+@dataclass
+class ChangeEvent:
+    """개정 감지 이벤트 — poll_changes 산출물."""
+    family_id: str
+    kind: str                    # amend / repeal / bootstrap
+    new_version_key: str         # KR = MST (법령일련번호)
+    title: str
+    promulgation_date: str | None = None   # YYYYMMDD
+    effective_date: str | None = None      # YYYYMMDD (목록 시행일자 — 조문별 차등 시행 주의)
+    revision_type: str | None = None       # 제개정구분명
+
+
+@dataclass
+class VersionPayload:
+    """fetch_version 산출물 1건 — primary 또는 annex 각자 자기 version_key (R4-M4).
+
+    전문 1콜 스냅샷 의미론(R7-M3 fixture 판정): 한 응답에서 primary + annex 전부 생성.
+    annex version_key = 'MST|{별표번호}-{별표가지번호}' (zero-padded 구조화 필드 그대로 —
+    suffix 문자열 파싱 아닌 필드 기반, R7-B1 a 업그레이드).
+    """
+    law_doc_kind: str            # primary / annex
+    version_key: str
+    title: str
+    content: str                 # 조문/별표 markdown 텍스트
+    promulgation_date: str | None = None   # YYYYMMDD (본문 기본정보)
+    effective_date: str | None = None      # YYYYMMDD (본문 기본정보 — 목록값과 다를 수 있음)
+    annex_label: str | None = None         # '별표1' / '별표5의2' (표시용)
+    meta: dict = field(default_factory=dict)

app/workers/statute_adapters/kr.py

@@ -0,0 +1,213 @@
+"""KR 법령 어댑터 — 국가법령정보센터 (law.go.kr DRF) (plan safety-library-1 B-1 PR①).
+
+poll_changes = lawSearch 목록 diff: 워치리스트 행별 정식 법령명 exact 조회 →
+MST(법령일련번호) != watermark 이면 ChangeEvent. law_monitor 의 검증된 호출 형태 재사용.
+
+fixture (2026-06-13 라이브 박제, tests/fixtures/statute_kr/):
+  - lawsearch_*.xml — 목록 필드: 법령ID(불변)·법령일련번호(MST)·공포일자·시행일자·제개정구분명
+  - lawservice_*.xml.gz — 전문 1콜 XML: 조문단위 853(산안기준규칙) + 별표단위 23 전부 포함
+    = 스냅샷 의미론 확정(R7-M3 ①: annex 부분 fetch 실패 개념 없음 — 같은 응답에 없는
+    별표 = 삭제 간주 가능). 별표번호+별표가지번호 = 구조화 필드(R7-M3 ② — suffix 문자열
+    파싱 불요, version_key 합성은 이 필드 기반. PR② fetch_version 소관).
+  - 조문 취득 방식 판정(R2-m1): 전문 1콜 + 로컬 파싱 확정 — lawjosub 조 단위 호출이면
+    산안기준규칙(853조)은 개정당 호출 폭증. lawjosub fixture 는 보조 박제.
+
+주의: 응답의 '법령상세링크' 필드에 OC 키가 포함됨 — fixture/로그에 raw 응답을 남길 때
+새니타이즈 필수 (repo fixture 는 __OC_REDACTED__ 처리됨).
+"""
+
+import asyncio
+import os
+import xml.etree.ElementTree as ET
+
+import httpx
+
+from core.crawl_politeness import CRAWL_UA
+from core.utils import setup_logger
+from workers.statute_adapters import ChangeEvent, VersionPayload
+
+logger = setup_logger("statute_kr")
+
+JURISDICTION = "KR"
+SOURCE_API = "law.go.kr"
+
+LAW_SEARCH_URL = "https://www.law.go.kr/DRF/lawSearch.do"
+LAW_SERVICE_URL = "https://www.law.go.kr/DRF/lawService.do"
+
+# 같은 도메인 연속 호출 간격 (일 1회 x 26콜 — 보수적)
+_POLL_DELAY_S = 1.5
+
+
+def _oc() -> str:
+    oc = os.getenv("LAW_OC", "")
+    if not oc:
+        raise RuntimeError("LAW_OC 미설정 — statute KR 어댑터 사용 불가")
+    return oc
+
+
+def parse_search_hit(xml_text: str, official_title: str) -> dict | None:
+    """lawSearch XML 에서 정식 법령명 exact match 1건 추출 (순수 함수 — fixture 테스트 대상).
+
+    정식명 기준 exact match — 워치리스트 title 이 정식명(가운뎃점 포함)이므로 안전.
+    (law_monitor 의 하드코딩 '유해위험작업...'(점 없음)이 영구 미매칭이던 함정의 교훈:
+    조회 키는 반드시 레지스트리의 정식명을 쓴다.)
+    """
+    root = ET.fromstring(xml_text)
+    for law in root.findall(".//law"):
+        if (law.findtext("법령명한글") or "").strip() != official_title:
+            continue
+        mst = (law.findtext("법령일련번호") or "").strip()
+        if not mst:
+            continue
+        return {
+            "mst": mst,
+            "law_id": (law.findtext("법령ID") or "").strip(),
+            "promulgation_date": (law.findtext("공포일자") or "").strip() or None,
+            "effective_date": (law.findtext("시행일자") or "").strip() or None,
+            "revision_type": (law.findtext("제개정구분명") or "").strip() or None,
+            "status_code": (law.findtext("현행연혁코드") or "").strip() or None,
+        }
+    return None
+
+
+def detect_change(hit: dict | None, act_family_id: str, act_title: str,
+                  watermark: str | None) -> ChangeEvent | None:
+    """목록 hit + 워터마크 → ChangeEvent (순수 함수 — fixture 테스트 대상).
+
+    - hit 없음 = 감지 불가 (None — 호출측이 fail-loud 로그. 폐지 단정 금지:
+      검색 누락/표기 변경 가능성과 구분 불가하므로 repeal 은 제개정구분명 기준만)
+    - MST == watermark = 변경 없음
+    - 제개정구분명에 '폐지' = repeal, 그 외 = amend
+    """
+    if hit is None:
+        return None
+    if watermark and hit["mst"] == watermark:
+        return None
+    kind = "repeal" if (hit.get("revision_type") or "").find("폐지") >= 0 else "amend"
+    return ChangeEvent(
+        family_id=act_family_id,
+        kind=kind,
+        new_version_key=hit["mst"],
+        title=act_title,
+        promulgation_date=hit.get("promulgation_date"),
+        effective_date=hit.get("effective_date"),
+        revision_type=hit.get("revision_type"),
+    )
+
+
+def _article_markdown(art: ET.Element) -> str:
+    """조문단위 1건 → 텍스트. 조문내용(이미 '제N조(제목) ...' 형태) + 항/호/목 전체.
+
+    메타 필드(조문번호/조문여부/조문시행일자 등)는 제외 — 조문내용과 항 서브트리만.
+    """
+    parts = []
+    body = (art.findtext("조문내용") or "").strip()
+    if body:
+        parts.append(body)
+    for hang in art.findall("항"):
+        text = "\n".join(t.strip() for t in hang.itertext() if t.strip())
+        if text:
+            parts.append(text)
+    return "\n".join(parts)
+
+
+def parse_service_payloads(xml_text: str, official_title: str, mst: str) -> list[VersionPayload]:
+    """lawService 전문 XML → VersionPayload 리스트 (순수 함수 — fixture 테스트 대상).
+
+    스냅샷 의미론: 응답에 있는 별표가 그 버전의 별표 전체 (R7-M3 fixture 판정).
+    - primary 1건: 전 조문 markdown (조문여부 != '조문' 행 = 장/절 헤더 → '## ' 처리)
+    - annex N건: 별표단위별 — version_key = 'MST|{별표번호}-{가지번호}' (zero-padded 그대로)
+    """
+    root = ET.fromstring(xml_text)
+    base = root.find(".//기본정보")
+    prom = (base.findtext("공포일자") or "").strip() or None if base is not None else None
+    eff = (base.findtext("시행일자") or "").strip() or None if base is not None else None
+
+    lines: list[str] = [f"# {official_title}", ""]
+    for art in root.findall(".//조문단위"):
+        is_article = (art.findtext("조문여부") or "").strip() == "조문"
+        text = _article_markdown(art)
+        if not text:
+            continue
+        if is_article:
+            lines.append(f"### {text}" if not text.startswith("제") else text)
+        else:
+            lines.append(f"## {text}")
+        lines.append("")
+    primary_content = "\n".join(lines).strip()
+
+    payloads = [VersionPayload(
+        law_doc_kind="primary",
+        version_key=mst,
+        title=official_title,
+        content=primary_content,
+        promulgation_date=prom,
+        effective_date=eff,
+    )]
+
+    for annex in root.findall(".//별표단위"):
+        no = (annex.findtext("별표번호") or "").strip()
+        sub = (annex.findtext("별표가지번호") or "").strip() or "00"
+        kind = (annex.findtext("별표구분") or "별표").strip()   # 별표 / 서식 — 별도 차원!
+        a_title = (annex.findtext("별표제목") or "").strip()
+        a_body = (annex.findtext("별표내용") or "").strip()
+        if not no:
+            continue
+        # 삭제 tombstone — KR 은 별표/서식 삭제가 absence 가 아니라 '삭제 <날짜>' 명시 행
+        # (fixture 실측: 산안기준규칙 서식1·2). 내용 없는 tombstone 은 적재 skip.
+        # 시리즈의 구버전 current 잔존 처리 = PR③ 관찰 후보 (absence 추론은 불요 확정).
+        if a_title.startswith("삭제") and len(a_body) < 50:
+            continue
+        label = f"{kind}{int(no)}" + (f"의{int(sub)}" if sub not in ("", "0", "00") else "")
+        payloads.append(VersionPayload(
+            law_doc_kind="annex",
+            # 구분 차원 포함 — (번호,가지)만으로는 별표1 vs 서식1 충돌 (fixture 실측)
+            version_key=f"{mst}|{kind}{no}-{sub}",
+            title=f"{official_title} {label} {a_title}".strip(),
+            content=f"# {official_title} {label}\n## {a_title}\n\n{a_body}".strip(),
+            promulgation_date=prom,
+            effective_date=eff,
+            annex_label=label,
+        ))
+    return payloads
+
+
+async def fetch_version(client: httpx.AsyncClient, act, change: ChangeEvent) -> list[VersionPayload]:
+    """전문 1콜 → payload 리스트 (R2-m1 판정: lawjosub 조 단위 호출 안 함 — 853조 폭증 회피)."""
+    resp = await client.get(
+        LAW_SERVICE_URL,
+        params={"OC": _oc(), "target": "law", "MST": change.new_version_key, "type": "XML"},
+        headers={"User-Agent": CRAWL_UA},
+    )
+    resp.raise_for_status()
+    payloads = parse_service_payloads(resp.text, act.title, change.new_version_key)
+    if not payloads or len(payloads[0].content) < 200:
+        # 파싱 검증 floor — 미달 시 예외 = 워터마크 미영속 (재시도 가능 상태 유지)
+        raise ValueError(f"전문 파싱 결과 빈약 ({act.family_id}): payloads={len(payloads)}")
+    return payloads
+
+
+async def poll_changes(client: httpx.AsyncClient, watch_rows: list) -> list[ChangeEvent]:
+    """워치리스트 행별 lawSearch diff. 행 단위 실패 격리 (한 법령 실패가 나머지를 막지 않음)."""
+    oc = _oc()
+    events: list[ChangeEvent] = []
+    for act in watch_rows:
+        try:
+            resp = await client.get(
+                LAW_SEARCH_URL,
+                params={"OC": oc, "target": "law", "type": "XML", "query": act.title},
+                headers={"User-Agent": CRAWL_UA},
+            )
+            resp.raise_for_status()
+            hit = parse_search_hit(resp.text, act.title)
+            if hit is None:
+                # fail-loud: 정식명 미매칭 = 표기 변경/검색 누락 의심 — 침묵 skip 금지
+                logger.warning(f"[statute-kr] 목록 미매칭: {act.family_id} {act.title!r}")
+            else:
+                ev = detect_change(hit, act.family_id, act.title, act.watermark)
+                if ev:
+                    events.append(ev)
+        except Exception as e:
+            logger.error(f"[statute-kr] poll 실패 ({act.family_id}): {type(e).__name__}: {e!r}")
+        await asyncio.sleep(_POLL_DELAY_S)
+    return events

app/workers/statute_collector.py

@@ -0,0 +1,381 @@
+"""statute_collector — 법령 수집 코어 (plan safety-library-1 B-1, PR②).
+
+구성 (잡 코드 통째 — R8-B1: 승격과 스윕의 PR 분리 = 배포 갭 이중 노출 윈도):
+  poll_changes(어댑터) → fetch_version(전문 1콜, payload 리스트) → ingest(전 버전
+  pending 적재 + 4축 주입) → 생애주기 잡(버전 시리즈 단위 승격·supersede + 상태 기반
+  레거시 스윕 + repeal — 단일 트랜잭션, KST 기준).
+
+핵심 계약 (카드 = 스펙):
+  - 워터마크 영속 = ingest 파싱 검증 통과 후에만 (실패 시 다음 폴링이 재감지)
+  - 승격·supersede 단위 = 버전 시리즈 = (family_id, law_doc_kind, annex 식별자)
+    — R7-B1: family 단위 구현 금지 (annex 승격이 primary 를 소거하는 본문 소실 경로)
+  - 레거시 스윕 = 상태 기반: 매 잡 실행, primary 시리즈 current 보유 + repeal 미감지
+    family 의 법령명 매핑 레거시(law_monitor 스냅샷) 청크 in_corpus=false (멱등)
+  - 매핑 = 정확 일치 가정 금지: title 의 '법령명 (YYYYMMDD)' 패턴에서 법령명 추출 후
+    정규화(공백·가운뎃점 변형 흡수) **동등** 비교 — prefix 비교 금지 ('산업안전보건법'이
+    '산업안전보건법 시행령' 레거시를 오폭하는 경로 차단)
+  - ingest 4축 (R8-M1): material_type='law' / jurisdiction=어댑터 상수 /
+    published_date=COALESCE(시행일, 공포일) / license=public_domain(저작권법 제7조)
+  - 부트스트랩(--bootstrap) = kind='bootstrap' 합성 이벤트, amend 와 동일 경로 +
+    extract_meta.backfill=true (E-1 게이트 집계 제외 마커)
+  - 가시성: source_health 성공/실패 기록 (HC.io 는 2026-05-30 알림 레이어 폐기로 부재 —
+    silent-skip 가드 정신은 crawl-health 보드 + health 행으로 대체)
+
+실행:
+  스케줄 = daily 07:00 KST (main.py — 구 law_monitor 슬롯 승계)
+  수동   = docker compose exec -T fastapi python -m workers.statute_collector [--bootstrap]
+"""
+
+import argparse
+import asyncio
+import hashlib
+import re
+import unicodedata
+from datetime import date, datetime, timezone
+from zoneinfo import ZoneInfo
+
+import httpx
+from sqlalchemy import select, update
+
+from core.database import async_session
+from core.utils import setup_logger
+from models.chunk import DocumentChunk
+from models.document import Document
+from models.legal_act import LegalAct, LegalMeta
+from models.news_source import NewsSource
+from models.queue import enqueue_stage
+from workers.news_collector import _get_or_create_health, _record_failure, _record_success
+from workers.statute_adapters import ChangeEvent, VersionPayload
+from workers.statute_adapters import kr
+
+logger = setup_logger("statute_collector")
+
+_KST = ZoneInfo("Asia/Seoul")
+_SOURCE_NAME = "KR 법령 (law.go.kr)"
+_LICENSE = {"scheme": "public_domain", "redistribute": True, "attribution": "국가법령정보센터"}
+_FETCH_DELAY_S = 2.5   # lawService 전문(최대 ~1.3MB) 연속 호출 간격
+
+# jurisdiction → 어댑터 모듈 (Phase 1 = KR 단독, 해외는 B-5 게이트 뒤)
+_ADAPTERS = {"KR": kr}
+
+
+# ─── 법령명 매핑 (R8-m1: 정확 일치 가정 금지 — 변형 흡수 정규화 + 동등 비교) ───
+
+_LEGACY_TITLE_RE = re.compile(r"^(.*?)\s*\((\d{8})\)")
+
+
+def normalize_law_name(name: str) -> str:
+    """공백·가운뎃점 변형 흡수 — NFC 정규화 후 공백/ㆍ·・ 제거."""
+    s = unicodedata.normalize("NFC", name or "")
+    return re.sub(r"[\sㆍ·・]", "", s)
+
+
+def legacy_law_name(title: str) -> str | None:
+    """레거시 law_monitor title('법령명 (YYYYMMDD) 섹션')에서 법령명 추출."""
+    m = _LEGACY_TITLE_RE.match(title or "")
+    return m.group(1).strip() if m else None
+
+
+def series_suffix(version_key: str) -> str | None:
+    """버전 시리즈의 annex 식별자 — version_key 'MST|NNNN-SS' 의 '|' 뒤 (primary=None)."""
+    return version_key.split("|", 1)[1] if "|" in version_key else None
+
+
+def _to_date(ymd: str | None) -> date | None:
+    digits = re.sub(r"\D", "", ymd or "")
+    if len(digits) != 8:
+        return None
+    try:
+        return date(int(digits[:4]), int(digits[4:6]), int(digits[6:8]))
+    except ValueError:
+        return None
+
+
+# ─── ingest (전 버전 pending 적재 — R2-B2/R3 계약) ──────────────────────────────
+
+async def _ingest_payload(session, act: LegalAct, ev: ChangeEvent,
+                          payload: VersionPayload, backfill: bool) -> bool:
+    """payload 1건 → Document + legal_meta(pending). 반환 = 신규 여부 (dedup 멱등)."""
+    fhash = hashlib.sha256(
+        f"statute|{act.jurisdiction}|{act.native_id}|{payload.version_key}".encode()
+    ).hexdigest()[:32]
+    existing = await session.execute(
+        select(Document.id).where(Document.file_hash == fhash).limit(1)
+    )
+    if existing.scalars().first():
+        return False
+
+    prom = _to_date(payload.promulgation_date or ev.promulgation_date)
+    eff = _to_date(payload.effective_date or ev.effective_date)
+    now = datetime.now(timezone.utc)
+    extra = {"backfill": True} if backfill else {}
+    doc = Document(
+        file_path=f"crawl/statute/{act.family_id}/{payload.version_key.replace('|', '_')}",
+        file_hash=fhash,
+        file_format="article",
+        file_size=len(payload.content.encode()),
+        file_type="note",
+        title=f"{payload.title} ({payload.promulgation_date or ev.promulgation_date or ''})".strip(),
+        extracted_text=payload.content,
+        extracted_at=now,
+        extractor_version="statute_kr@law.go.kr",
+        md_status="skipped",
+        md_extraction_error="statute: 텍스트 네이티브, markdown 변환 비대상",
+        source_channel="crawl",
+        data_origin="external",
+        review_status="approved",
+        ai_domain="법령",
+        ai_sub_group=act.title,
+        ai_tags=[f"법령/KR/{act.title}"],
+        # 안전 자료실 ingest 4축 (R8-M1 — classify-skip 경로라 ingest 시점 필수)
+        material_type="law",
+        jurisdiction=kr.JURISDICTION,
+        published_date=eff or prom,
+        extract_meta={
+            "statute": {"family_id": act.family_id, "law_id": act.native_id,
+                        "kind": payload.law_doc_kind, "version_key": payload.version_key,
+                        "annex_label": payload.annex_label,
+                        "event_kind": ev.kind, "revision_type": ev.revision_type},
+            "license": dict(_LICENSE),
+            **extra,
+        },
+    )
+    session.add(doc)
+    await session.flush()
+
+    session.add(LegalMeta(
+        document_id=doc.id,
+        family_id=act.family_id,
+        law_doc_kind=payload.law_doc_kind,
+        version_key=payload.version_key,
+        promulgation_date=prom,
+        effective_date=eff,
+        version_status="pending",   # 전 버전 pending 적재 — 승격은 생애주기 잡만
+    ))
+    # summarize 안 함 (조문 자체가 정본 — 맥미니 부하 0), embed+chunk 만
+    await enqueue_stage(session, doc.id, "embed")
+    await enqueue_stage(session, doc.id, "chunk")
+    return True
+
+
+# ─── 생애주기 잡 (전이·supersede·스윕·repeal 의 유일한 코드 지점) ────────────────
+
+async def _flip_chunks(session, doc_ids: list[int]) -> int:
+    if not doc_ids:
+        return 0
+    result = await session.execute(
+        update(DocumentChunk)
+        .where(DocumentChunk.doc_id.in_(doc_ids), DocumentChunk.in_corpus.is_(True))
+        .values(in_corpus=False)
+    )
+    return result.rowcount or 0
+
+
+async def _legacy_doc_ids(session, act: LegalAct) -> list[int]:
+    """법령명 매핑 레거시(law_monitor) 문서 id — 정규화 동등 비교 (prefix 금지)."""
+    result = await session.execute(
+        select(Document.id, Document.title).where(
+            Document.source_channel == "law_monitor",
+            Document.deleted_at.is_(None),
+        )
+    )
+    want = normalize_law_name(act.title)
+    ids = []
+    for doc_id, title in result.all():
+        name = legacy_law_name(title or "")
+        if name and normalize_law_name(name) == want:
+            ids.append(doc_id)
+    return ids
+
+
+async def run_lifecycle(session) -> dict:
+    """일 1회 생애주기 잡 — 호출측이 단일 트랜잭션 commit. KST 기준, 멱등."""
+    today = datetime.now(_KST).date()
+    stats = {"promoted": 0, "superseded": 0, "repealed": 0,
+             "legacy_flipped_docs": 0, "legacy_flipped_chunks": 0}
+
+    acts_result = await session.execute(select(LegalAct).where(LegalAct.watch.is_(True)))
+    acts = {a.family_id: a for a in acts_result.scalars().all()}
+
+    lm_result = await session.execute(
+        select(LegalMeta).where(LegalMeta.family_id.in_(list(acts.keys())))
+    )
+    metas = lm_result.scalars().all()
+
+    # 1) repeal — 마킹된 family: current+pending 전부 repealed + 청크 flip + 레거시 flip (R7-M2)
+    repeal_families = {fid for fid, a in acts.items() if a.repeal_detected_at is not None}
+    for fid in repeal_families:
+        rows = [m for m in metas if m.family_id == fid and m.version_status in ("pending", "current")]
+        for m in rows:
+            m.version_status = "repealed"
+            stats["repealed"] += 1
+        await _flip_chunks(session, [m.document_id for m in rows])
+        legacy_ids = await _legacy_doc_ids(session, acts[fid])
+        stats["legacy_flipped_chunks"] += await _flip_chunks(session, legacy_ids)
+
+    # 2) 승격 + supersede — 버전 시리즈 단위 (R7-B1 a: family 단위 금지)
+    series: dict[tuple, list[LegalMeta]] = {}
+    for m in metas:
+        if m.family_id in repeal_families:
+            continue
+        series.setdefault(
+            (m.family_id, m.law_doc_kind, series_suffix(m.version_key)), []
+        ).append(m)
+
+    for key, rows in series.items():
+        due = sorted(
+            (m for m in rows if m.version_status == "pending"
+             and (m.effective_date or m.promulgation_date)
+             and (m.effective_date or m.promulgation_date) <= today),
+            key=lambda m: (m.effective_date or m.promulgation_date),
+        )
+        for m in due:
+            prev = [c for c in rows if c.version_status == "current" and c is not m]
+            for c in prev:
+                c.version_status = "superseded"
+                stats["superseded"] += 1
+            await _flip_chunks(session, [c.document_id for c in prev])
+            m.version_status = "current"
+            stats["promoted"] += 1
+
+    # 3) 레거시 스윕 — 상태 기반 (R6-B1 a / R7-B1 b: primary 시리즈 current 보유 한정)
+    for fid, act in acts.items():
+        if fid in repeal_families:
+            continue
+        has_primary_current = any(
+            m.family_id == fid and m.law_doc_kind == "primary" and m.version_status == "current"
+            for m in metas
+        )
+        if not has_primary_current:
+            continue   # R3-B1 ② 내장 — fetch 실패 family 의 레거시 보존
+        legacy_ids = await _legacy_doc_ids(session, act)
+        flipped = await _flip_chunks(session, legacy_ids)
+        if flipped:
+            stats["legacy_flipped_docs"] += len(legacy_ids)
+            stats["legacy_flipped_chunks"] += flipped
+
+    return stats
+
+
+# ─── 메인 런 ─────────────────────────────────────────────────────────────────────
+
+async def run(bootstrap: bool = False) -> None:
+    """poll → fetch → ingest(가족 단위 커밋) → 생애주기 잡. 가족 단위 실패 격리."""
+    async with async_session() as session:
+        result = await session.execute(
+            select(LegalAct).where(LegalAct.watch.is_(True)).order_by(LegalAct.family_id)
+        )
+        rows = result.scalars().all()
+        if not rows:
+            logger.warning("[statute] 워치리스트 비어 있음 — 시드(migration 356) 미적용?")
+            return
+        source = await _get_source(session)
+        await session.commit()
+        source_id = source.id
+
+    ingested = 0
+    failed = 0
+    by_jur: dict[str, list] = {}
+    for row in rows:
+        by_jur.setdefault(row.jurisdiction, []).append(row)
+
+    async with httpx.AsyncClient(timeout=60) as client:
+        for jur, acts in by_jur.items():
+            adapter = _ADAPTERS.get(jur)
+            if adapter is None:
+                logger.warning(f"[statute] 어댑터 없는 jurisdiction skip: {jur}")
+                continue
+            assert adapter.JURISDICTION == jur, \
+                f"어댑터/행 jurisdiction 불일치: {adapter.JURISDICTION} != {jur}"
+
+            events = await adapter.poll_changes(client, acts)
+            acts_by_id = {a.family_id: a for a in acts}
+            for ev in events:
+                if bootstrap:
+                    ev.kind = "bootstrap"   # 합성 이벤트 — amend 와 동일 경로 (R6-m2)
+                act_ref = acts_by_id[ev.family_id]
+                try:
+                    payloads = await adapter.fetch_version(client, act_ref, ev)
+                    async with async_session() as session:
+                        act = await session.get(LegalAct, ev.family_id)
+                        new_docs = 0
+                        for p in payloads:
+                            if await _ingest_payload(session, act, ev, p, backfill=bootstrap):
+                                new_docs += 1
+                        # 워터마크 영속 = 파싱 검증(payload floor) 통과 후에만
+                        act.watermark = ev.new_version_key
+                        if ev.kind == "repeal":
+                            act.repeal_detected_at = datetime.now(timezone.utc)
+                        await session.commit()
+                    ingested += new_docs
+                    logger.info(f"[statute] ingest {ev.family_id} ({ev.kind}): "
+                                f"payload {len(payloads)}건 중 신규 {new_docs}건")
+                except Exception as e:
+                    failed += 1
+                    logger.error(f"[statute] ingest 실패 ({ev.family_id}): "
+                                 f"{type(e).__name__}: {e!r} — 워터마크 미영속, 다음 폴링 재감지")
+                await asyncio.sleep(_FETCH_DELAY_S)
+
+    # 생애주기 잡 — 수집 사이클 직후, 단일 트랜잭션 (0-2 ②)
+    async with async_session() as session:
+        stats = await run_lifecycle(session)
+        await session.commit()
+    logger.info(f"[statute] lifecycle: {stats}")
+
+    # health — fail-loud 가시성 (HC.io 폐기로 보드/health 행이 1차 관측면)
+    async with async_session() as session:
+        h = await _get_or_create_health(session, source_id)
+        now = datetime.now(timezone.utc)
+        if failed:
+            _record_failure(h, f"ingest 실패 {failed}건", now)
+        else:
+            _record_success(h, ingested, False, now)
+        await session.commit()
+
+    logger.info(f"[statute] run 완료 — 신규 문서 {ingested}건, 실패 {failed}건"
+                + (" (bootstrap)" if bootstrap else ""))
+
+
+async def _get_source(session) -> NewsSource:
+    result = await session.execute(select(NewsSource).where(NewsSource.name == _SOURCE_NAME))
+    source = result.scalars().first()
+    if source is None:
+        source = NewsSource(
+            name=_SOURCE_NAME, feed_url=kr.LAW_SEARCH_URL, feed_type="rss",
+            fetch_method="api", fulltext_policy="none", source_channel="crawl",
+            category="Safety", language="ko", country="KR",
+            enabled=False,   # 6h 뉴스 사이클 비대상 — 본 워커가 daily 폴링
+        )
+        session.add(source)
+        await session.flush()
+    return source
+
+
+async def poll_once() -> int:
+    """관찰 전용 폴링 (PR① 잔존 CLI — 상태 변경 0)."""
+    async with async_session() as session:
+        result = await session.execute(
+            select(LegalAct).where(LegalAct.watch.is_(True)).order_by(LegalAct.family_id)
+        )
+        rows = result.scalars().all()
+    total = 0
+    async with httpx.AsyncClient(timeout=30) as client:
+        events = await kr.poll_changes(client, [r for r in rows if r.jurisdiction == "KR"])
+        for ev in events:
+            logger.info(f"[statute] 변경 감지 ({ev.kind}): {ev.family_id} {ev.title} "
+                        f"MST={ev.new_version_key}")
+        total = len(events)
+    logger.info(f"[statute] poll 완료 — 변경 {total}건 (관찰 전용)")
+    return total
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--bootstrap", action="store_true",
+                        help="26 family 현행판 1회 부트스트랩 (backfill 마커, R4-M1)")
+    parser.add_argument("--poll-only", action="store_true", help="관찰 전용 폴링")
+    args = parser.parse_args()
+    if args.poll_only:
+        asyncio.run(poll_once())
+    else:
+        asyncio.run(run(bootstrap=args.bootstrap))

frontend/src/lib/components/ProcessingFlowBoard.svelte

@@ -1,16 +1,18 @@
 <script lang="ts">
-  // 처리 머신 보드 v2 — 파이프라인 흐름 뷰 (plan ds-board-engines-1, R2 통합안).
-  // 메인 = 좌→우 흐름 노드(병목 amber·실패 뱃지), 노드 클릭 = 상세 패널(안1 변형),
-  // 실패 뱃지 클릭 = 실패 처리 드로어 (재시도/건너뛰기 — 영구 실패의 유일한 조치 경로).
-  // 데이터 = GET /api/queue/overview (60s 폴링 store) + GET /api/queue/failed (드로어 열 때).
+  // 처리 머신 보드 v3 — 통합안 (plan ds-board-merged: C2 머신레인 + C3 번다운/정직ETA).
+  //   · 머신 3레인(GPU/맥미니/맥북) = "누가 일하나" + 요약 오프로드(맥북 합류) 가시화
+  //   · 지배 백로그 번다운 패널 = "언제 끝나나" + 유입 차감한 정직 ETA(summarize_eta)
+  //   · 신선도 '갱신 N초 전' + stale 경고 / 실패 드로어·상세 패널은 v2 자산 재사용.
+  // 데이터 = GET /api/queue/overview (60s 폴링 store) + GET /api/queue/failed (드로어).
   import { api } from '$lib/api';
-  import { refreshQueueOverview } from '$lib/stores/queueOverview';
+  import { refreshQueueOverview, queueUpdatedAt } from '$lib/stores/queueOverview';
   import { addToast } from '$lib/stores/toast';
   import {
     AUX_NODES,
     FLOW_NODES,
     MACHINE_META,
     type FlowNodeDef,
+    type FlowMachine,
     etaShort,
     flowStageLabel,
     formatAgeSec,
@@ -20,6 +22,7 @@
     FailedItem,
     FailedListResponse,
     MachineCurrentItem,
+    MachineOverview,
     QueueOverview,
     QueueStageRow,
     RetryResponse,
@@ -82,14 +85,6 @@
   );
   const totalFailed = $derived(overview.totals.failed);
 
-  // 머신 스트립 — overview.machines 의 state/처리율 + 정적 모델 메타
-  const machineStrip = $derived(
-    overview.machines.map((m) => ({
-      ...m,
-      meta: MACHINE_META[m.key],
-    })),
-  );
-
   // ─── 선택 상태 (노드 상세 / 실패 드로어 — 동시에 하나만) ───
   let selected = $state<string | null>(null);
   let failOpen = $state(false);
@@ -194,22 +189,105 @@
     await Promise.all([loadFailures(), refreshQueueOverview()]);
   }
 
-  // ─── trend_24h 스파크라인 (summarize 유입 vs 소화 — API 가 주는데 미렌더이던 슬롯) ───
-  const spark = $derived.by(() => {
+  // ─── 머신 레인 (C2) — mainNodes 를 머신별로 그룹 + 머신 카드(state/처리율) 결합 ───
+  const machineByKey = $derived(
+    new Map<FlowMachine, MachineOverview>(overview.machines.map((m) => [m.key as FlowMachine, m])),
+  );
+  const LANE_ORDER: FlowMachine[] = ['gpu', 'macmini', 'macbook'];
+  const lanes = $derived(
+    LANE_ORDER.map((key) => ({
+      key,
+      meta: MACHINE_META[key],
+      card: machineByKey.get(key) ?? null,
+      nodes: mainNodes.filter((n) => n.def.machine === key),
+    })),
+  );
+
+  // 요약 오프로드 분담 — 맥미니 vs 맥북 (A-1 summarize_by_machine)
+  const split = $derived(overview.summarize_by_machine);
+  const splitTotal1h = $derived(Math.max(1, split.macmini.done_1h + split.macbook.done_1h));
+  const macbookSharePct = $derived(Math.round((split.macbook.done_1h / splitTotal1h) * 100));
+  // 맥북이 요약을 실제로 가져가는 중인가 (합류 표식 게이트)
+  const offloadActive = $derived(split.macbook.done_1h > 0);
+
+  // ─── 지배 백로그 = 요약. 정직 ETA(유입 차감) — summarize_eta ───
+  const eta = $derived(overview.summarize_eta);
+  // 정직 ETA 라벨: eta_minutes null = 유입이 소화를 앞섬(소진 불가)
+  const honestEtaLabel = $derived(
+    eta.pending === 0
+      ? '비어 있음'
+      : eta.eta_minutes != null
+        ? etaShort(eta.eta_minutes)
+        : '소진 불가',
+  );
+  const honestEtaWarn = $derived(eta.pending > 0 && eta.eta_minutes == null);
+
+  /** 단계별 정직 ETA(순소화율) — 노드용. 유입>소화면 null(소진 불가) */
+  function netEtaLabel(n: NodeStats): string | null {
+    if (n.pending === 0) return '한가';
+    const net = n.done1h - n.created1h;
+    if (net > 0) return etaShort(Math.round((n.pending / net) * 60));
+    if (n.created1h > n.done1h) return '유입 우세';
+    return null;
+  }
+
+  // ─── 신선도 (B-4) — '갱신 N초 전' + stale 경고 (폴링 60s) ───
+  let now = $state(Date.now());
+  $effect(() => {
+    const id = setInterval(() => (now = Date.now()), 1000);
+    return () => clearInterval(id);
+  });
+  const ageSec = $derived(
+    $queueUpdatedAt != null ? Math.max(0, Math.round((now - $queueUpdatedAt) / 1000)) : null,
+  );
+  const stale = $derived(ageSec != null && ageSec > 90);
+  const freshLabel = $derived(
+    ageSec == null
+      ? '갱신 대기'
+      : ageSec < 60
+        ? `갱신 ${ageSec}초 전`
+        : `갱신 ${Math.round(ageSec / 60)}분 전`,
+  );
+
+  // ─── 24h 번다운 (C3) — 요약 유입 vs 소화 + 맥북 합류 변곡점 마커 ───
+  const burn = $derived.by(() => {
     const t = overview.trend_24h;
     if (!t || t.length === 0) return null;
     const max = Math.max(1, ...t.map((b) => Math.max(b.inflow, b.done)));
-    const w = 120;
-    const h = 24;
+    const w = 300;
+    const h = 64;
     const step = w / Math.max(1, t.length - 1);
-    const pts = (sel: (b: (typeof t)[number]) => number) =>
-      t.map((b, i) => `${(i * step).toFixed(1)},${(h - (sel(b) / max) * (h - 3) + 1).toFixed(1)}`).join(' ');
-    return { inflow: pts((b) => b.inflow), done: pts((b) => b.done) };
+    const y = (v: number) => (h - (v / max) * (h - 8) + 4).toFixed(1);
+    const line = (sel: (b: (typeof t)[number]) => number) =>
+      t.map((b, i) => `${(i * step).toFixed(1)},${y(sel(b))}`).join(' ');
+    const doneLine = line((b) => b.done);
+    const area = `0,${h} ${doneLine} ${w.toFixed(1)},${h}`;
+    // 합류 변곡점 = done 최대 버킷 (맥북 야간 drain 합류 추정)
+    let mi = 0;
+    t.forEach((b, i) => {
+      if (b.done > t[mi].done) mi = i;
+    });
+    return {
+      w,
+      h,
+      area,
+      doneLine,
+      inflowLine: line((b) => b.inflow),
+      markX: (mi * step).toFixed(1),
+      markHour: t[mi].hour,
+      markDone: t[mi].done,
+      peak: max,
+    };
   });
+
+  // 머신 상태 dot 색 클래스
+  function dotClass(state: string): string {
+    return state === 'active' ? 'bg-success' : state === 'deferred' ? 'bg-warning' : 'bg-faint';
+  }
 </script>
 
 <div class="mt-5">
-  <!-- 헤더: 타이틀 + 요약 24h 스파크라인 + 실패 합계 -->
+  <!-- 헤더: 타이틀 + 신선도 + 실패 합계 -->
   <div class="flex items-center justify-between gap-3 mb-3">
     <div class="text-[11px] font-bold text-dim uppercase tracking-wider">처리 머신</div>
     <div class="flex items-center gap-3">
@@ -219,80 +297,107 @@
           onclick={openFailures}
         >실패 {totalFailed}건 처리</button>
       {/if}
-      {#if spark}
-        <div class="flex items-center gap-2 text-[10px] text-faint tabular-nums" title="요약(summarize) 단계 24시간 — 유입(회색) vs 소화(녹색)">
-          <svg width="120" height="24" viewBox="0 0 120 24" class="block">
-            <polyline points={spark.inflow} fill="none" stroke="currentColor" stroke-width="1.5" class="text-faint" />
-            <polyline points={spark.done} fill="none" stroke="currentColor" stroke-width="1.5" class="text-success" />
-          </svg>
-          <span>요약 24h 유입/소화</span>
-        </div>
-      {/if}
+      <span class="flex items-center gap-1.5 text-[10px] tabular-nums {stale ? 'text-warning' : 'text-faint'}" title="60초 폴링">
+        <span class="w-1.5 h-1.5 rounded-full {stale ? 'bg-warning' : 'bg-success'}"></span>
+        {freshLabel}{#if stale} · 갱신 지연{/if}
+      </span>
     </div>
   </div>
 
-  <!-- 머신 스트립 -->
-  <div class="flex flex-wrap gap-2 mb-3">
-    {#each machineStrip as m (m.key)}
-      <div class="flex items-center gap-2 bg-surface border border-default rounded-full px-3.5 py-1.5 text-xs">
-        <span class="w-2 h-2 rounded-full shrink-0 {m.state === 'active' ? 'bg-success' : m.state === 'deferred' ? 'bg-warning' : 'bg-faint'}"></span>
-        <span class="font-bold text-text">{m.meta?.label ?? m.label}</span>
-        <span class="text-[10px] text-faint font-mono">{m.meta?.model}</span>
-        <span class="text-[11px] text-dim tabular-nums">{formatRate(m.done_1h)}/h</span>
-        {#if m.key === 'macbook' && m.deferred_pending > 0}
-          <span class="text-[10px] font-semibold text-warning tabular-nums">보류 {m.deferred_pending}</span>
-        {/if}
+  <!-- 지배 백로그 스트립 (요약) + 정직 ETA -->
+  <div class="flex items-center flex-wrap gap-x-3 gap-y-1 bg-surface border border-warning/50 rounded-card px-3.5 py-2 mb-3">
+    <span class="text-[9px] font-bold text-warning border border-warning/60 rounded-full px-2 py-px">지배 백로그</span>
+    <span class="text-xs font-bold text-text">요약</span>
+    <span class="text-[11px] text-dim tabular-nums">대기 <b class="text-text">{eta.pending.toLocaleString()}</b> · 순소화 <b class="text-text">{formatRate(eta.done_rate_1h)}</b>/h · 유입 {formatRate(eta.inflow_rate_1h)}/h</span>
+    <span class="ml-auto flex items-center gap-1.5 border rounded-full px-2.5 py-0.5 {honestEtaWarn ? 'border-warning text-warning' : 'border-accent text-accent'}">
+      <span class="text-[10px] font-semibold">정직 ETA</span>
+      <span class="text-xs font-bold tabular-nums">{honestEtaLabel}</span>
+    </span>
+  </div>
+
+  <!-- 머신 레인 (누가 일하나 + 요약 오프로드) -->
+  <div class="grid gap-2 mb-3">
+    {#each lanes as lane (lane.key)}
+      <div class="bg-surface border border-default rounded-card px-3.5 py-2.5">
+        <div class="flex items-center gap-2 flex-wrap mb-2">
+          <span class="w-2 h-2 rounded-full shrink-0 {dotClass(lane.card?.state ?? 'idle')}"></span>
+          <span class="text-[9px] font-bold rounded px-1.5 py-px mtag-{lane.key}">{lane.meta.label}</span>
+          <span class="text-[10px] text-faint font-mono">{lane.meta.model}</span>
+          <span class="text-[11px] text-dim tabular-nums ml-1">{formatRate(lane.card?.done_1h ?? 0)}/h</span>
+          {#if lane.key === 'macbook' && (lane.card?.deferred_pending ?? 0) > 0}
+            <span class="text-[10px] font-semibold text-warning tabular-nums">보류 {lane.card?.deferred_pending}</span>
+          {/if}
+          {#if lane.card?.state === 'deferred'}
+            <span class="text-[9px] text-warning">잠듦 — 요약은 맥미니로 복귀</span>
+          {/if}
+        </div>
+        <div class="flex items-stretch gap-1.5 flex-wrap">
+          {#each lane.nodes as n (n.def.key)}
+            {@const idle = n.pending + n.processing + n.doneToday + n.failed === 0}
+            <button
+              class="relative text-left rounded-lg border px-2.5 py-1.5 transition-colors cursor-pointer hover:bg-surface-hover min-w-[96px]
+                {idle ? 'border-dashed border-default opacity-55' : n.inflowDominant ? 'border-warning' : 'border-default'}
+                {selected === n.def.key ? 'node-sel' : ''}"
+              onclick={() => toggleNode(n.def.key)}
+              title="{n.def.label} — 클릭하면 상세"
+            >
+              {#if n.failed > 0}
+                <span class="absolute -top-1.5 -right-1 text-[9px] font-extrabold bg-error text-white rounded-full px-1.5">{n.failed}</span>
+              {/if}
+              <div class="flex items-center gap-1 text-[11px] font-semibold text-text whitespace-nowrap">
+                {n.def.label}
+                {#if n.processing > 0}<span class="inline-block w-1.5 h-1.5 rounded-full bg-accent animate-pulse"></span>{/if}
+              </div>
+              <div class="text-sm font-extrabold tabular-nums leading-tight text-text">{n.pending.toLocaleString()}<span class="text-[9px] text-faint font-normal ml-0.5">대기</span></div>
+              <div class="text-[9px] text-dim tabular-nums whitespace-nowrap">{formatRate(n.done1h)}/h · 오늘 {n.doneToday.toLocaleString()}</div>
+              {#if n.def.key === 'summarize'}
+                <div class="mt-1 h-1 w-full rounded-full overflow-hidden flex" title="맥미니 {split.macmini.done_1h}/h · 맥북 {split.macbook.done_1h}/h">
+                  <span class="block h-full mtag-macmini-bar" style="width:{100 - macbookSharePct}%"></span>
+                  <span class="block h-full mtag-macbook-bar" style="width:{macbookSharePct}%"></span>
+                </div>
+                <div class="text-[9px] text-faint tabular-nums whitespace-nowrap mt-0.5">맥미니 {split.macmini.done_1h} · 맥북 {split.macbook.done_1h}/h</div>
+              {/if}
+            </button>
+          {/each}
+          {#if lane.key === 'macbook' && offloadActive}
+            <button
+              class="text-left rounded-lg border border-dashed border-warning/50 px-2.5 py-1.5 cursor-pointer hover:bg-surface-hover min-w-[96px]"
+              onclick={() => toggleNode('summarize')}
+              title="맥북이 요약을 맥미니에서 가져와 처리 중"
+            >
+              <div class="flex items-center gap-1 text-[11px] font-semibold text-text whitespace-nowrap">요약 합류 <span class="text-[8px] font-bold text-warning">OFFLOAD</span></div>
+              <div class="text-sm font-extrabold tabular-nums leading-tight text-text">{split.macbook.done_1h}<span class="text-[9px] text-faint font-normal ml-0.5">/h</span></div>
+              <div class="text-[9px] text-dim tabular-nums whitespace-nowrap">요약의 {macbookSharePct}% 담당</div>
+            </button>
+          {/if}
+        </div>
       </div>
     {/each}
   </div>
 
-  <!-- 흐름 노드 -->
-  <div class="flex items-stretch overflow-x-auto pb-1">
-    {#each mainNodes as n, i (n.def.key)}
-      {#if i > 0}
-        <div class="flex items-center text-faint text-sm px-1.5 shrink-0" aria-hidden="true">→</div>
-      {/if}
-      <div
-        class="relative bg-surface border-[1.5px] rounded-card px-3 py-2.5 min-w-[124px] shrink-0 text-left transition-colors cursor-pointer hover:bg-surface-hover
-          {n.inflowDominant ? 'border-warning' : n.etaMinutes != null && n.def.stages.includes('chunk') ? 'border-success' : 'border-default'}
-          {selected === n.def.key ? 'node-sel' : ''}"
-        role="button"
-        tabindex="0"
-        onclick={() => toggleNode(n.def.key)}
-        onkeydown={(e) => { if (e.key === 'Enter' || e.key === ' ') { e.preventDefault(); toggleNode(n.def.key); } }}
-        title="{n.def.label} — 클릭하면 상세"
-      >
-        {#if n.failed > 0}
-          <button
-            class="absolute -top-2 -right-1.5 text-[9px] font-extrabold bg-error text-white rounded-full px-1.5 py-px shadow cursor-pointer"
-            onclick={(e) => { e.stopPropagation(); openFailures(); }}
-            title="실패 {n.failed}건 — 클릭하면 실패 처리"
-          >{n.failed}</button>
-        {/if}
-        <span class="inline-block text-[9px] font-bold rounded px-1.5 py-px mb-1.5 mtag-{n.def.machine}">
-          {MACHINE_META[n.def.machine].label} · {n.def.engine}
-        </span>
-        <div class="text-xs font-bold text-text flex items-center gap-1.5">
-          {n.def.label}
-          {#if n.processing > 0}
-            <span class="inline-block w-1.5 h-1.5 rounded-full bg-accent animate-pulse" title="처리 중 {n.processing}"></span>
-          {/if}
-          {#if n.inflowDominant}
-            <span class="text-[9px] font-bold text-warning">유입 우세</span>
-          {/if}
-        </div>
-        <div class="text-base font-extrabold tabular-nums tracking-tight leading-tight mt-0.5 text-text">
-          {n.pending.toLocaleString()}
-        </div>
-        <div class="text-[10px] text-dim tabular-nums">
-          {formatRate(n.done1h)}/h · 오늘 {n.doneToday.toLocaleString()}
-          {#if n.etaMinutes != null && !n.inflowDominant && n.pending > 0}
-            · <span class="text-accent font-semibold">{etaShort(n.etaMinutes)}</span>
-          {/if}
-        </div>
+  <!-- 번다운 / ETA 패널 -->
+  {#if burn}
+    <div class="bg-surface border border-default rounded-card px-3.5 py-3 mb-1">
+      <div class="flex items-center gap-2 mb-2">
+        <span class="text-[11px] font-bold text-text">요약 백로그 24시간</span>
+        <span class="text-[9px] text-faint">유입(회색) vs 소화(녹색)</span>
+        {#if offloadActive}<span class="text-[9px] text-warning ml-auto">맥북 합류 {burn.markHour} — 소화 급증</span>{/if}
       </div>
-    {/each}
-  </div>
+      <svg viewBox="0 0 {burn.w} {burn.h}" class="block w-full" style="height:64px" preserveAspectRatio="none" role="img" aria-label="요약 백로그 24시간 번다운">
+        <polygon points={burn.area} fill="currentColor" class="text-success" opacity="0.12" />
+        <polyline points={burn.inflowLine} fill="none" stroke="currentColor" stroke-width="1.2" class="text-faint" />
+        <polyline points={burn.doneLine} fill="none" stroke="currentColor" stroke-width="1.6" class="text-success" />
+        {#if offloadActive}
+          <line x1={burn.markX} y1="0" x2={burn.markX} y2={burn.h} stroke="currentColor" stroke-width="1" stroke-dasharray="2 2" class="text-warning" opacity="0.7" />
+        {/if}
+      </svg>
+      <div class="flex flex-wrap gap-x-4 gap-y-1 mt-2 pt-2 border-t border-default text-[10px] text-dim tabular-nums">
+        {#each mainNodes.filter((n) => n.pending > 0 && n.def.key !== 'summarize') as n (n.def.key)}
+          <span class="whitespace-nowrap">{n.def.label} 대기 <b class="text-text">{n.pending.toLocaleString()}</b>{#if netEtaLabel(n)} · <span class="text-accent font-semibold">{netEtaLabel(n)}</span>{/if}</span>
+        {/each}
+      </div>
+    </div>
+  {/if}
 
   <!-- 보조 라인 -->
   <p class="text-[10px] text-faint mt-1.5 tabular-nums">
@@ -413,6 +518,9 @@
   .mtag-gpu { background: #e7eef6; color: #3b6ea5; }
   .mtag-macmini { background: #efe9f7; color: #8a5fbf; }
   .mtag-macbook { background: #f7eedd; color: #b07a10; }
+  /* 요약 오프로드 분담 막대 채움 (맥미니 보라 / 맥북 황) */
+  .mtag-macmini-bar { background: #8a5fbf; }
+  .mtag-macbook-bar { background: #b07a10; }
   .node-sel { outline: 2px solid #3b6ea5; outline-offset: 1px; }
   .detail-frame { border-color: #3b6ea5; }
   .detail-head { background: #e7eef6; }

frontend/src/lib/components/SectionOutline.svelte

@@ -43,14 +43,17 @@
   {@const open = selectedId === s.chunk_id}
   {@const active = activeKey != null && activeKey === s.chunk_id}
   {@const typeLabel = sectionTypeLabel(s.section_type)}
+  {@const depth = Math.max(0, (s.level ?? 1) - 1)}
   <li>
     <button
       type="button"
       onclick={() => { toggle(item); onJump?.(s.chunk_id); }}
       aria-expanded={open}
       aria-current={active ? 'true' : undefined}
+      style="padding-left:{8 + depth * 13}px"
       class={[
-        'w-full text-left px-2 py-1.5 rounded-md text-xs flex items-start gap-1.5 transition-colors border-l-2',
+        'w-full text-left pr-2 py-1.5 rounded-md text-xs flex items-start gap-1.5 transition-colors border-l-2',
+        depth > 0 ? 'text-[11px]' : '',
         open ? 'bg-surface-active text-text border-accent' : active ? 'bg-surface text-accent-hover border-accent' : 'text-dim hover:bg-surface hover:text-text border-transparent',
       ].join(' ')}
     >

frontend/src/lib/stores/queueOverview.ts

@@ -17,6 +17,11 @@ let pollHandle: ReturnType<typeof setInterval> | null = null;
 let subscriberCount = 0;
 let inFlight: Promise<void> | null = null;
 
+// 마지막 성공 갱신 시각(epoch ms) — 보드 신선도 '갱신 N초 전' + stale 경고용
+// (ds-board-merged B-4). 실패(null 수렴) 시엔 갱신 안 함 → age 가 늘어 stale 로 드러남.
+const updatedAt = writable<number | null>(null);
+export const queueUpdatedAt = { subscribe: updatedAt.subscribe };
+
 const internal = writable<QueueOverview | null>(null, (_set) => {
   subscriberCount += 1;
   if (subscriberCount === 1 && browser) {
@@ -54,7 +59,9 @@ export async function refreshQueueOverview(): Promise<void> {
   if (inFlight) return inFlight;
   inFlight = (async () => {
     try {
-      internal.set(await fetchOverview());
+      const ov = await fetchOverview();
+      internal.set(ov);
+      if (ov) updatedAt.set(Date.now()); // 성공 시에만 신선도 갱신 (실패=stale 유지)
     } finally {
       inFlight = null;
     }

frontend/src/lib/types/queue.ts

@@ -43,13 +43,19 @@ export interface SummarizeEta {
   eta_minutes: number | null;
 }
 
-/** 시간당 유입 vs 소화 (이번 트랙 미렌더 — 후속 추세 위젯 슬롯) */
+/** 시간당 유입 vs 소화 (요약 24h 추이) */
 export interface TrendPoint {
   hour: string;
   inflow: number;
   done: number;
 }
 
+/** summarize 머신별 완료 실적 분담 (오프로드 가시화 — ds-board-merged A-1) */
+export interface SummarizeByMachine {
+  macmini: { done_1h: number; done_today: number };
+  macbook: { done_1h: number; done_today: number };
+}
+
 export interface QueueTotals {
   pending: number;
   processing: number;
@@ -72,6 +78,7 @@ export interface QueueStageRow {
 export interface QueueOverview {
   machines: MachineOverview[];
   summarize_eta: SummarizeEta;
+  summarize_by_machine: SummarizeByMachine;
   trend_24h: TrendPoint[];
   stages: QueueStageRow[];
   totals: QueueTotals;

frontend/src/lib/utils/queueDisplay.ts

@@ -37,10 +37,14 @@ export function etaPhrase(minutes: number): string {
   return `약 ${text}시간 후 소진 예상`;
 }
 
-/** ETA 분 → 칩용 짧은 표기 ("약 4.6시간" / "약 12분") */
+/** ETA 분 → 칩용 짧은 표기 ("약 12분" / "약 4.6시간" / 48h+ = "약 5.5일") */
 export function etaShort(minutes: number): string {
   if (minutes < 60) return `약 ${Math.max(1, Math.round(minutes))}분`;
   const hours = minutes / 60;
+  if (hours >= 48) {
+    const days = hours / 24;
+    return `약 ${days >= 10 ? Math.round(days) : Math.round(days * 10) / 10}일`;
+  }
   const text = hours >= 10 ? String(Math.round(hours)) : String(Math.round(hours * 10) / 10);
   return `약 ${text}시간`;
 }

frontend/src/routes/+page.svelte

@@ -19,7 +19,7 @@
   import EmptyState from '$lib/components/ui/EmptyState.svelte';
   import Skeleton from '$lib/components/ui/Skeleton.svelte';
   import {
-    Scale, FileText, Pin, ChevronRight, GraduationCap, Upload, Newspaper,
+    Scale, FileText, Pin, GraduationCap, Upload, Newspaper,
   } from 'lucide-svelte';
   import { addToast } from '$lib/stores/toast';
 
@@ -133,17 +133,6 @@
   // 백엔드 미배포/실패 시 store=null → 보드 자체가 조용히 생략 (silent 비차단).
   let queue = $derived<QueueOverview | null>($queueOverview);
 
-  // 머신 담당 단계 라벨 — STAGE_LABEL 재사용 + overview 전용 단계 보강
-  // (backend services/queue_overview.py _STAGE_ORDER 와 동기), 미지 키는 raw
-  const QUEUE_STAGE_LABEL: Record<string, string> = {
-    ...STAGE_LABEL,
-    summarize: '요약', chunk: '청크', markdown: '마크다운',
-    fulltext: '전문', deep_summary: '심층분석',
-  };
-  function queueStageLabel(stage: string): string {
-    return QUEUE_STAGE_LABEL[stage] ?? stage;
-  }
-
   onMount(() => {
     void refreshQueueOverview();
     const handle = setInterval(() => void refreshQueueOverview(), 30_000);
@@ -191,35 +180,10 @@
   let pipelineRows = $derived(
     summary ? buildPipelineRows(summary.pipeline_status, summary.queue_lag ?? []) : []
   );
-  let pipelineMax = $derived(Math.max(1, ...pipelineRows.map((r) => r.total)));
   let totalFailed = $derived(summary?.failed_count ?? 0);
   let totalPending = $derived(pipelineRows.reduce((s, r) => s + r.pending, 0));
   let totalProcessing = $derived(pipelineRows.reduce((s, r) => s + r.processing, 0));
 
-  let pipelineManualClosed = $state(false);
-  let pipelineOpen = $derived(
-    pipelineManualClosed ? false : (queue?.totals.failed ?? totalFailed) > 0
-  );
-
-  // 단계별 현황 (2026-06-11 피드백 재설계: 완료가 보여야 한다 — overview.stages 단일 소스)
-  // active = 오늘 움직임이 있는 단계만, idle = 전부 0 인 단계는 한 줄로 숨김.
-  let stageRows = $derived(queue?.stages ?? []);
-  let activeStageRows = $derived(
-    stageRows.filter((r) => r.pending + r.processing + r.failed + r.done_today > 0)
-  );
-  let idleStageRows = $derived(
-    stageRows.filter((r) => r.pending + r.processing + r.failed + r.done_today === 0)
-  );
-  let stageDoneToday = $derived(stageRows.reduce((s, r) => s + r.done_today, 0));
-
-  function formatAge(sec: number | null): string {
-    if (sec == null || sec <= 0) return '';
-    if (sec < 60) return `${sec}초 전`;
-    if (sec < 3600) return `${Math.floor(sec / 60)}분 전`;
-    if (sec < 86400) return `${Math.floor(sec / 3600)}시간 전`;
-    return `${Math.floor(sec / 86400)}일 전`;
-  }
-
   function formatTime(dateStr: string) {
     const d = new Date(dateStr);
     if (isNaN(d.getTime())) return ''; // 빈 문자열/유효하지 않은 created_at → 'Invalid Date' 회피
@@ -463,80 +427,6 @@
         <ProcessingFlowBoard overview={queue} />
       {/if}
 
-      <!-- ═══ 단계 상세 (기존 stage 테이블 — 접힘 강등, 실패 있을 때 자동 펼침) ═══ -->
-      <details
-        class="mt-5"
-        open={pipelineOpen}
-        ontoggle={(e) => { if (!e.currentTarget.open) pipelineManualClosed = true; }}
-      >
-        <summary class="flex items-center justify-between px-5 py-3.5 bg-surface border border-default rounded-card cursor-pointer hover:bg-surface-hover transition-colors select-none list-none">
-          <span class="text-sm font-semibold text-text flex items-center gap-2">
-            <ChevronRight size={14} class="transition-transform details-chevron" />
-            단계별 현황
-          </span>
-          <span class="text-xs text-dim flex items-center gap-2.5">
-            {#if queue}
-              {#if stageDoneToday > 0}<span class="text-success">오늘 {stageDoneToday.toLocaleString()} 완료</span>{/if}
-              {#if queue.totals.failed > 0}<span class="text-error font-medium">실패 {queue.totals.failed}</span>{/if}
-              {#if queue.totals.pending > 0}<span>대기 {queue.totals.pending.toLocaleString()}</span>{/if}
-              {#if stageDoneToday === 0 && queue.totals.failed === 0 && queue.totals.pending === 0}<span>모든 단계 한가함</span>{/if}
-            {:else}
-              {#if totalFailed > 0}<span class="text-error font-medium">실패 {totalFailed}</span>{/if}
-              {#if totalPending > 0}<span>대기 {totalPending}</span>{/if}
-            {/if}
-          </span>
-        </summary>
-
-        <div class="mt-2 px-5 py-4 bg-surface border border-default rounded-card">
-          {#if queue}
-            {#if activeStageRows.length > 0}
-              <div class="space-y-3.5">
-                {#each activeStageRows as row (row.stage)}
-                  {@const total = row.done_today + row.pending + row.processing}
-                  {@const donePct = total > 0 ? (row.done_today / total) * 100 : 0}
-                  {@const procPct = total > 0 ? (row.processing / total) * 100 : 0}
-                  <div>
-                    <div class="flex items-baseline justify-between text-xs mb-1.5 gap-2">
-                      <span class="font-medium text-text flex items-center gap-1.5 whitespace-nowrap">
-                        {queueStageLabel(row.stage)}
-                        {#if row.processing > 0}
-                          <span class="inline-block w-1.5 h-1.5 rounded-full bg-accent animate-pulse"></span>
-                          <span class="text-accent font-normal">처리 중 {row.processing}</span>
-                        {/if}
-                      </span>
-                      <span class="text-dim tabular-nums flex items-center gap-2.5 whitespace-nowrap">
-                        {#if row.done_today > 0}<span class="text-success">오늘 {row.done_today.toLocaleString()} 완료</span>{/if}
-                        {#if row.pending > 0}<span>대기 {row.pending.toLocaleString()}</span>{/if}
-                        {#if row.failed > 0}<span class="text-error font-medium">실패 {row.failed}</span>{/if}
-                      </span>
-                    </div>
-                    <!-- 게이지 = 이 단계의 오늘 진척 (완료 / 완료+대기) — 가득 찬 초록 = 다 끝남 -->
-                    <div class="flex h-1.5 w-full overflow-hidden rounded-sm bg-bg" title="오늘 완료 {row.done_today.toLocaleString()} / 잔여 {row.pending.toLocaleString()}">
-                      {#if donePct > 0}<div class="bg-success/70 h-full" style="width: {donePct}%"></div>{/if}
-                      {#if procPct > 0}<div class="bg-accent h-full" style="width: {Math.max(procPct, 1)}%"></div>{/if}
-                    </div>
-                    {#if row.pending > 0 && row.oldest_pending_age_sec && row.oldest_pending_age_sec > 600}
-                      <p class="text-[10px] mt-1 tabular-nums {row.oldest_pending_age_sec > 21600 ? 'text-warning' : 'text-faint'}">
-                        가장 오래 기다린 항목 {formatAge(row.oldest_pending_age_sec)}
-                      </p>
-                    {/if}
-                  </div>
-                {/each}
-              </div>
-            {:else}
-              <p class="text-xs text-dim text-center py-3">대기·처리·실패 없음 — 모든 단계가 한가합니다</p>
-            {/if}
-            {#if idleStageRows.length > 0}
-              <p class="text-[11px] text-faint mt-4 pt-3 border-t border-default">
-                비어 있음: {idleStageRows.map((r) => queueStageLabel(r.stage)).join(' · ')}
-              </p>
-            {/if}
-          {:else}
-            <p class="text-xs text-dim text-center py-3">현황을 불러오지 못했습니다</p>
-          {/if}
-        </div>
-      </details>
-
     {/if}
   </div>
 </div>
@@ -548,7 +438,3 @@
   </div>
 {/snippet}
 
-<style>
-  details[open] :global(.details-chevron) { transform: rotate(90deg); }
-  details summary::-webkit-details-marker { display: none; }
-</style>

frontend/src/routes/documents/[id]/+page.svelte

@@ -30,7 +30,6 @@
   import AnalysisPanel from '$lib/components/AnalysisPanel.svelte';
   import ReadCounter from '$lib/components/ReadCounter.svelte';
   import SectionOutline from '$lib/components/SectionOutline.svelte';
-  import Tabs from '$lib/components/ui/Tabs.svelte';
 
   marked.use({ mangle: false, headerIds: false });
   function renderMd(text) {
@@ -460,53 +459,68 @@
         {/if}
       </div>
 
-      <!-- 오른쪽 — 메타 Tabs [정보 | AI | 관리] (카드 11개 수직 스프롤 해소) -->
-      <aside class="min-w-0">
-        <Card>
-          <Tabs
-            tabs={[
-              { id: 'info', label: '정보' },
-              { id: 'ai', label: 'AI' },
-              { id: 'manage', label: '관리' },
-            ]}
-          >
-            {#snippet children(tab)}
-              <div class="pt-3 space-y-4">
-                {#if tab === 'info'}
-                  {#if doc.category === 'library'}
-                    <ReadCounter
-                      documentId={doc.id}
-                      initialCount={doc.read_count ?? 0}
-                      initialLastReadAt={doc.last_read_at ?? null}
-                    />
-                  {/if}
-                  <FileInfoView {doc} />
-                  <ProcessingStatusView {doc} />
-                {:else if tab === 'ai'}
-                  <AnalysisPanel docId={doc.id} doc={doc} />
-                  <AIClassificationEditor {doc} />
-                  <div>
-                    <h4 class="text-xs font-semibold text-dim uppercase mb-1.5">관련 문서</h4>
-                    <!-- TODO(backend): GET /documents/{id}/related?limit=10 (벡터 유사도) -->
-                    <EmptyState
-                      icon={FileText}
-                      title="추후 지원"
-                      description="관련 문서 추천은 backend 연동 후 제공됩니다."
-                    />
-                  </div>
-                {:else}
-                  <LibraryPathEditor {doc} />
-                  <NoteEditor {doc} />
-                  <EditUrlEditor {doc} />
-                  <TagsEditor {doc} />
-                  <div class="pt-2 border-t border-default">
-                    <DocumentDangerZone {doc} ondelete={handleDocDelete} />
-                  </div>
-                {/if}
-              </div>
-            {/snippet}
-          </Tabs>
-        </Card>
+      <!-- 오른쪽 — 슬림 전역 인사이트 레일 (D3: 탭 게이트 제거, 요약·심층·불일치 상시 노출).
+           정보/관리는 접이(<details>) — 데스크탑은 인사이트 상시, 모바일은 본문 메인 + 열어서 확인. -->
+      <aside class="min-w-0 space-y-3">
+        {#if doc.category === 'library'}
+          <Card>
+            <ReadCounter
+              documentId={doc.id}
+              initialCount={doc.read_count ?? 0}
+              initialLastReadAt={doc.last_read_at ?? null}
+            />
+          </Card>
+        {/if}
+
+        <!-- 요약·분석 — 기본 펼침(데스크탑 상시감, 모바일 접기 가능) -->
+        <details open class="bg-surface border border-default rounded-card overflow-hidden group">
+          <summary class="cursor-pointer list-none flex items-center justify-between px-3.5 py-2.5 text-xs font-semibold text-dim uppercase tracking-wide select-none">
+            <span>요약 · 분석</span>
+            <ChevronRight size={14} class="transition-transform group-open:rotate-90 text-faint" />
+          </summary>
+          <div class="px-3.5 pb-3.5 space-y-4">
+            <AnalysisPanel docId={doc.id} doc={doc} />
+            <AIClassificationEditor {doc} />
+            <div>
+              <h4 class="text-xs font-semibold text-dim uppercase mb-1.5">관련 문서</h4>
+              <!-- TODO(backend): GET /documents/{id}/related?limit=10 (벡터 유사도) — v1 제외(자리만) -->
+              <EmptyState
+                icon={FileText}
+                title="추후 지원"
+                description="관련 문서 추천은 backend 연동 후 제공됩니다."
+              />
+            </div>
+          </div>
+        </details>
+
+        <!-- 문서 정보 — 접이(기본 닫힘) -->
+        <details class="bg-surface border border-default rounded-card overflow-hidden group">
+          <summary class="cursor-pointer list-none flex items-center justify-between px-3.5 py-2.5 text-xs font-semibold text-dim uppercase tracking-wide select-none">
+            <span>문서 정보</span>
+            <ChevronRight size={14} class="transition-transform group-open:rotate-90 text-faint" />
+          </summary>
+          <div class="px-3.5 pb-3.5 space-y-3">
+            <FileInfoView {doc} />
+            <ProcessingStatusView {doc} />
+          </div>
+        </details>
+
+        <!-- 관리 — 접이(기본 닫힘) -->
+        <details class="bg-surface border border-default rounded-card overflow-hidden group">
+          <summary class="cursor-pointer list-none flex items-center justify-between px-3.5 py-2.5 text-xs font-semibold text-dim uppercase tracking-wide select-none">
+            <span>관리</span>
+            <ChevronRight size={14} class="transition-transform group-open:rotate-90 text-faint" />
+          </summary>
+          <div class="px-3.5 pb-3.5 space-y-3">
+            <LibraryPathEditor {doc} />
+            <NoteEditor {doc} />
+            <EditUrlEditor {doc} />
+            <TagsEditor {doc} />
+            <div class="pt-2 border-t border-default">
+              <DocumentDangerZone {doc} ondelete={handleDocDelete} />
+            </div>
+          </div>
+        </details>
       </aside>
     </div>
 

migrations/340_documents_material_type.sql

@@ -0,0 +1,6 @@
+-- 340_documents_material_type.sql
+-- 안전 자료실 분류 축 A-1 (1/12) — 자료유형 컬럼.
+-- plan: safety-library-1 (PKM plans/2026-06-12-safety-library-plan.html)
+-- TEXT+CHECK 방식 (PG enum 아님 — 152 의 enum ADD VALUE 동일-런 사용 불가 함정 회피).
+-- 값 부여 = 수집기 ingest 시점 deterministic (classify_worker 아님 — classify-skip 경로 다수).
+ALTER TABLE documents ADD COLUMN IF NOT EXISTS material_type TEXT;

migrations/341_documents_material_type_check.sql

@@ -0,0 +1,6 @@
+-- 341_documents_material_type_check.sql
+-- 안전 자료실 분류 축 A-1 (2/12) — material_type 값 공간 named CHECK.
+-- plan: safety-library-1 0-1 확정 7값. 값 추가 시 = 본 제약 DROP + 재ADD 2파일 (named 라 가능).
+-- NULL 은 CHECK 통과 (비안전/일반 문서는 NULL 유지 — 전수 분류 시도 금지).
+ALTER TABLE documents ADD CONSTRAINT chk_documents_material_type
+    CHECK (material_type IN ('law', 'paper', 'book', 'incident', 'manual', 'standard', 'guide'));

migrations/342_documents_jurisdiction.sql

@@ -0,0 +1,5 @@
+-- 342_documents_jurisdiction.sql
+-- 안전 자료실 분류 축 A-1 (3/12) — 관할(나라) 컬럼. 법령 1급 시민 축.
+-- plan: safety-library-1 0-1. 'GB' 표기 (news_sources.country 실측 어휘와 통일, UI 라벨만 UK).
+-- paper 는 NULL 허용 (국제 학술지 — 관할 개념 부적합). INT = ISO 류 국제기구 자료 유보.
+ALTER TABLE documents ADD COLUMN IF NOT EXISTS jurisdiction TEXT;

migrations/343_documents_jurisdiction_check.sql

@@ -0,0 +1,4 @@
+-- 343_documents_jurisdiction_check.sql
+-- 안전 자료실 분류 축 A-1 (4/12) — jurisdiction 값 공간 named CHECK.
+ALTER TABLE documents ADD CONSTRAINT chk_documents_jurisdiction
+    CHECK (jurisdiction IN ('KR', 'US', 'EU', 'JP', 'GB', 'INT'));

migrations/344_documents_law_jurisdiction_check.sql

@@ -0,0 +1,7 @@
+-- 344_documents_law_jurisdiction_check.sql
+-- 안전 자료실 분류 축 A-1 (5/12) — 나라 혼선 금지를 구조로 강제.
+-- 법령(material_type='law')인데 jurisdiction NULL 인 행은 적재 자체가 거부된다.
+-- 업로드 승인 경로는 proposed_jurisdiction 필수 입력 (KR 기본값 오염 금지 — plan A-2).
+-- material_type 이 NULL 이면 식 전체가 NULL = CHECK 통과 (비법령 무영향).
+ALTER TABLE documents ADD CONSTRAINT chk_documents_law_jurisdiction
+    CHECK (material_type <> 'law' OR jurisdiction IS NOT NULL);

migrations/345_documents_published_date.sql

@@ -0,0 +1,5 @@
+-- 345_documents_published_date.sql
+-- 안전 자료실 분류 축 A-1 (6/12) — 유형별 대표 날짜 (패싯 연도·freshness 단일 날짜 축).
+-- 법령 = COALESCE(effective_date, promulgation_date) — plan 0-1 R2-M2 확정.
+-- 논문 = 발행일 / 재해 = 발생일 / 뉴스·크롤 = extract_meta.published_at backfill (A-3).
+ALTER TABLE documents ADD COLUMN IF NOT EXISTS published_date DATE;

migrations/346_legal_acts_table.sql

@@ -0,0 +1,22 @@
+-- 346_legal_acts_table.sql
+-- 안전 자료실 A-1 (7/12) — 법령 레지스트리 = 워치리스트 (news_sources 패턴의 법령판).
+-- plan: safety-library-1 0-2. statute_watchlist 별도 테이블 안 만듦 (R2 blocker — 이중 정의 해소, watermark 흡수).
+-- KOSHA GUIDE / KGS Code 는 비대상 (guide=비법령, KGS=watch-폴더 단독 트랙 R3-M5).
+-- 시드 = B-1 PR① (레거시 law_monitor 26개 superset, watch=true 전부 — R3-B1).
+-- repeal_detected_at: 어댑터(코어)는 폐지 감지 마킹만, 전이는 일일 잡 단일 지점 (R3-M3).
+CREATE TABLE IF NOT EXISTS legal_acts (
+    family_id          TEXT PRIMARY KEY,
+    jurisdiction       TEXT NOT NULL CHECK (jurisdiction IN ('KR', 'US', 'EU', 'JP', 'GB', 'INT')),
+    law_level          TEXT NOT NULL CHECK (law_level IN ('statute', 'decree', 'rule', 'admin_rule', 'code')),
+    title              TEXT NOT NULL,
+    title_ko           TEXT,
+    parent_family_id   TEXT REFERENCES legal_acts(family_id),
+    native_id          TEXT NOT NULL,
+    source_api         TEXT NOT NULL,
+    watch              BOOLEAN NOT NULL DEFAULT TRUE,
+    poll_cycle         TEXT NOT NULL DEFAULT 'daily' CHECK (poll_cycle IN ('daily', 'weekly', 'monthly', 'quarterly')),
+    watermark          TEXT,
+    repeal_detected_at TIMESTAMPTZ,
+    created_at         TIMESTAMPTZ NOT NULL DEFAULT now(),
+    updated_at         TIMESTAMPTZ NOT NULL DEFAULT now()
+);

migrations/347_legal_meta_table.sql

@@ -0,0 +1,20 @@
+-- 347_legal_meta_table.sql
+-- 안전 자료실 A-1 (8/12) — 법령 문서 1건(=1버전 또는 1부속문서)당 1행. documents 1:0..1 위성, 최소형.
+-- plan: safety-library-1 0-2. supersedes 체인 컬럼은 미포함 (개정 이벤트 10건 관찰 후 승격).
+-- version_key: KR primary = MST / annex = 'MST|별표N' 합성 (같은 MST 별표 다건 UNIQUE 충돌 회피)
+--   / interpretation = 소스 native id. dedup 키도 이 합성형 그대로 (R3-M4 silent skip 차단).
+-- version_status 운영 계약 (B-1 PR② 일일 잡이 유일한 전이 지점, R2-B2·R3-M3):
+--   전 버전 pending 적재 → 잡이 KST 기준 시행일 도래분 current 승격 + 직전 current 를 superseded
+--   + 구버전 청크 in_corpus=false 를 한 트랜잭션 처리. repeal 도 잡 경유.
+-- 입법예고 등 신호류 문서는 legal_meta 없음 (legal_meta 존재 = 법령 본문).
+CREATE TABLE IF NOT EXISTS legal_meta (
+    document_id        BIGINT PRIMARY KEY REFERENCES documents(id) ON DELETE CASCADE,
+    family_id          TEXT NOT NULL REFERENCES legal_acts(family_id),
+    law_doc_kind       TEXT NOT NULL DEFAULT 'primary' CHECK (law_doc_kind IN ('primary', 'annex', 'interpretation')),
+    version_key        TEXT NOT NULL,
+    promulgation_date  DATE,
+    effective_date     DATE,
+    version_status     TEXT NOT NULL DEFAULT 'pending' CHECK (version_status IN ('pending', 'current', 'superseded', 'repealed')),
+    created_at         TIMESTAMPTZ NOT NULL DEFAULT now(),
+    CONSTRAINT uq_legal_meta_version UNIQUE (family_id, law_doc_kind, version_key)
+);

migrations/348_documents_material_type_idx.sql

@@ -0,0 +1,4 @@
+-- 348_documents_material_type_idx.sql
+-- 안전 자료실 A-1 (9/12) — material_type partial index (128~131 facet 인덱스 선례).
+CREATE INDEX IF NOT EXISTS idx_documents_material_type
+    ON documents (material_type) WHERE material_type IS NOT NULL;

migrations/349_documents_jurisdiction_idx.sql

@@ -0,0 +1,4 @@
+-- 349_documents_jurisdiction_idx.sql
+-- 안전 자료실 A-1 (10/12) — jurisdiction partial index.
+CREATE INDEX IF NOT EXISTS idx_documents_jurisdiction
+    ON documents (jurisdiction) WHERE jurisdiction IS NOT NULL;

migrations/350_legal_meta_family_idx.sql

@@ -0,0 +1,6 @@
+-- 350_legal_meta_family_idx.sql
+-- 안전 자료실 A-1 (11/12) — point-in-time 조회 축.
+-- 술어 = COALESCE(effective_date, promulgation_date) (KGS 류 시행일 미상 row 침묵 탈락 방지)
+--   이나 인덱스는 effective_date 단순형으로 시작 — COALESCE expression index 는 실측 후.
+CREATE INDEX IF NOT EXISTS idx_legal_meta_family
+    ON legal_meta (family_id, effective_date DESC);

migrations/351_documents_paper_doi_uq.sql

@@ -0,0 +1,9 @@
+-- 351_documents_paper_doi_uq.sql
+-- 안전 자료실 A-1 (12/12) — 논문 DOI dedup 구조 강제 (partial UNIQUE).
+-- doi 보유 계약 (R3 — R2-B1): paper.doi 는 서지 Document 단일 보유.
+--   OA 전문 PDF / 구매분 file Document 는 paper.doi 를 갖지 않고 paper.parent_doi 링크로 연결
+--   → 인덱스 식이 NULL 이라 다중 행 허용, 2-Document 구조와 무충돌.
+-- DOI 정규화(소문자·prefix 제거)는 단일 함수 경유 — 저장=조회 동일 함수 원칙 (B-3).
+CREATE UNIQUE INDEX IF NOT EXISTS uq_documents_paper_doi
+    ON documents (lower(extract_meta #>> '{paper,doi}'))
+    WHERE material_type = 'paper' AND extract_meta #>> '{paper,doi}' IS NOT NULL;

migrations/352_news_sources_material_type.sql

@@ -0,0 +1,8 @@
+-- 352_news_sources_material_type.sql
+-- 안전 자료실 A-2 (1/4) — 소스 레지스트리에 자료유형 기본값.
+-- plan: safety-library-1 A-2. 수집기 ingest 시점 deterministic 부여의 단일 진실 =
+-- 레지스트리 행 (country 와 동일 패턴 — 코드 하드코딩/이름 매칭 회피).
+-- NULL = 자료유형 비대상 (뉴스/철학 등). paper 소스는 country 가 있어도
+-- documents.jurisdiction 은 NULL (국제 학술지 — 코드 레벨 규칙).
+ALTER TABLE news_sources ADD COLUMN IF NOT EXISTS material_type TEXT
+    CHECK (material_type IN ('law', 'paper', 'book', 'incident', 'manual', 'standard', 'guide'));

migrations/353_news_sources_license_scheme.sql

@@ -0,0 +1,6 @@
+-- 353_news_sources_license_scheme.sql
+-- 안전 자료실 A-2 (2/4) — 소스별 라이선스 scheme (0-3 license 메타 deterministic 주입).
+-- kogl(공공누리류) / ogl(UK) / public_domain(미 연방) / proprietary / unknown.
+-- 미확정 소스는 보수적으로 unknown/proprietary + redistribute=false 에서 시작
+-- (갱신은 근거 확보 시 완화 방향 — 보수적=빡빡 원칙).
+ALTER TABLE news_sources ADD COLUMN IF NOT EXISTS license_scheme TEXT;

migrations/354_news_sources_license_redistribute.sql

@@ -0,0 +1,4 @@
+-- 354_news_sources_license_redistribute.sql
+-- 안전 자료실 A-2 (3/4) — 재배포 가능 여부. P3 다이제스트/발행류의 구조 게이트 입력
+-- (redistribute=false 소스 제외 — 사람 기억 의존 차단, 0-3).
+ALTER TABLE news_sources ADD COLUMN IF NOT EXISTS license_redistribute BOOLEAN;

migrations/355_news_sources_material_seed.sql

@@ -0,0 +1,45 @@
+-- 355_news_sources_material_seed.sql
+-- 안전 자료실 A-2 (4/4) — 기존 안전/공학 소스 12행 material_type + license 시드.
+-- 매핑 근거 = plan safety-library-1 0-1 경계 확정 (2026-06-12 prod 레지스트리 실측 대조):
+--   law=입법예고(신호) / incident=HSE·KOSHA사례·CSB·CCPS / guide=KOSHA GUIDE·TWI
+--   / standard=NB·API 공지 / paper=JPVT·arXiv (jurisdiction 은 코드에서 NULL 강제).
+-- 뉴스/철학 소스는 NULL 유지 (자료유형 비대상). 이름 키 = 시드 마이그레이션이 부여한 고정값.
+UPDATE news_sources SET
+    material_type = CASE name
+        WHEN '고용노동부 입법행정예고' THEN 'law'
+        WHEN 'UK HSE Press' THEN 'incident'
+        WHEN 'KOSHA 재해사례' THEN 'incident'
+        WHEN 'US CSB 사고조사보고서' THEN 'incident'
+        WHEN 'CCPS Process Safety Beacon' THEN 'incident'
+        WHEN 'KOSHA GUIDE' THEN 'guide'
+        WHEN 'TWI Job Knowledge' THEN 'guide'
+        WHEN 'National Board 기술 아티클' THEN 'standard'
+        WHEN 'API 표준 공지' THEN 'standard'
+        WHEN 'ASME J. Pressure Vessel Technology' THEN 'paper'
+        WHEN 'arXiv physics.app-ph' THEN 'paper'
+        WHEN 'arXiv cond-mat.mtrl-sci' THEN 'paper'
+    END,
+    license_scheme = CASE name
+        WHEN '고용노동부 입법행정예고' THEN 'kogl'
+        WHEN 'KOSHA 재해사례' THEN 'kogl'
+        WHEN 'KOSHA GUIDE' THEN 'kogl'
+        WHEN 'UK HSE Press' THEN 'ogl'
+        WHEN 'US CSB 사고조사보고서' THEN 'public_domain'
+        WHEN 'TWI Job Knowledge' THEN 'proprietary'
+        WHEN 'National Board 기술 아티클' THEN 'proprietary'
+        WHEN 'API 표준 공지' THEN 'proprietary'
+        WHEN 'CCPS Process Safety Beacon' THEN 'proprietary'
+        WHEN 'ASME J. Pressure Vessel Technology' THEN 'proprietary'
+        WHEN 'arXiv physics.app-ph' THEN 'unknown'
+        WHEN 'arXiv cond-mat.mtrl-sci' THEN 'unknown'
+    END,
+    license_redistribute = CASE name
+        WHEN 'UK HSE Press' THEN TRUE
+        WHEN 'US CSB 사고조사보고서' THEN TRUE
+        ELSE FALSE
+    END
+WHERE name IN ('고용노동부 입법행정예고', 'UK HSE Press', 'KOSHA 재해사례',
+               'US CSB 사고조사보고서', 'CCPS Process Safety Beacon', 'KOSHA GUIDE',
+               'TWI Job Knowledge', 'National Board 기술 아티클', 'API 표준 공지',
+               'ASME J. Pressure Vessel Technology', 'arXiv physics.app-ph',
+               'arXiv cond-mat.mtrl-sci');

migrations/356_seed_legal_acts_kr.sql

@@ -0,0 +1,41 @@
+-- 356_seed_legal_acts_kr.sql
+-- 안전 자료실 B-1 PR① — legal_acts KR 시드 26행 (레거시 law_monitor 26개 superset).
+-- plan: safety-library-1 B-1. watch=true 26개 전부 (R3-B1 ① — '우선순위'는 정렬일 뿐 제외 아님).
+-- 법령ID/공포/시행 = 2026-06-13 lawSearch 라이브 실측 (tests/fixtures/statute_kr/seed_26laws.tsv).
+-- ★ '유해ㆍ위험작업...' = 정식명에 가운뎃점(U+318D) — law_monitor 하드코딩(점 없음)은 exact match
+--   불일치로 이 법령을 영구 미매칭하던 잠복 누락이었음 (R8-m1 의 watchlist 판 실증).
+-- parent 계열: 법률 → 시행령/시행규칙/위임 부령. VALUES 순서 = 부모 선행 (FK).
+INSERT INTO legal_acts (family_id, jurisdiction, law_level, title, parent_family_id, native_id, source_api, watch, poll_cycle)
+SELECT v.family_id, v.jurisdiction, v.law_level, v.title, v.parent_family_id, v.native_id, v.source_api, v.watch, v.poll_cycle
+FROM (VALUES
+    -- 법률 (statute, 14)
+    ('kr-law:001766', 'KR', 'statute', '산업안전보건법',                              NULL,             '001766', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:013993', 'KR', 'statute', '중대재해 처벌 등에 관한 법률',                NULL,             '013993', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:001807', 'KR', 'statute', '건설기술 진흥법',                             NULL,             '001807', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:000237', 'KR', 'statute', '시설물의 안전 및 유지관리에 관한 특별법',     NULL,             '000237', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:009502', 'KR', 'statute', '위험물안전관리법',                            NULL,             '009502', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:000162', 'KR', 'statute', '화학물질관리법',                              NULL,             '000162', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:011857', 'KR', 'statute', '화학물질의 등록 및 평가 등에 관한 법률',      NULL,             '011857', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:009503', 'KR', 'statute', '소방시설 설치 및 관리에 관한 법률',           NULL,             '009503', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:001854', 'KR', 'statute', '전기사업법',                                  NULL,             '001854', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:013718', 'KR', 'statute', '전기안전관리법',                              NULL,             '013718', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:001850', 'KR', 'statute', '고압가스 안전관리법',                         NULL,             '001850', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:001849', 'KR', 'statute', '액화석유가스의 안전관리 및 사업법',           NULL,             '001849', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:001872', 'KR', 'statute', '근로기준법',                                  NULL,             '001872', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:002016', 'KR', 'statute', '환경영향평가법',                              NULL,             '002016', 'law.go.kr', TRUE, 'daily'),
+    -- 대통령령 (decree, 7)
+    ('kr-law:003786', 'KR', 'decree',  '산업안전보건법 시행령',                       'kr-law:001766',  '003786', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:014159', 'KR', 'decree',  '중대재해 처벌 등에 관한 법률 시행령',         'kr-law:013993',  '014159', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:002111', 'KR', 'decree',  '건설기술 진흥법 시행령',                      'kr-law:001807',  '002111', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:009707', 'KR', 'decree',  '위험물안전관리법 시행령',                     'kr-law:009502',  '009707', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:004390', 'KR', 'decree',  '화학물질관리법 시행령',                       'kr-law:000162',  '004390', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:009694', 'KR', 'decree',  '소방시설 설치 및 관리에 관한 법률 시행령',    'kr-law:009503',  '009694', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:002246', 'KR', 'decree',  '고압가스 안전관리법 시행령',                  'kr-law:001850',  '002246', 'law.go.kr', TRUE, 'daily'),
+    -- 부령 (rule, 5)
+    ('kr-law:007364', 'KR', 'rule',    '산업안전보건법 시행규칙',                     'kr-law:001766',  '007364', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:007363', 'KR', 'rule',    '산업안전보건기준에 관한 규칙',                'kr-law:001766',  '007363', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:007844', 'KR', 'rule',    '유해ㆍ위험작업의 취업 제한에 관한 규칙',      'kr-law:001766',  '007844', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:006175', 'KR', 'rule',    '건설기술 진흥법 시행규칙',                    'kr-law:001807',  '006175', 'law.go.kr', TRUE, 'daily'),
+    ('kr-law:009732', 'KR', 'rule',    '위험물안전관리법 시행규칙',                   'kr-law:009502',  '009732', 'law.go.kr', TRUE, 'daily')
+) AS v(family_id, jurisdiction, law_level, title, parent_family_id, native_id, source_api, watch, poll_cycle)
+WHERE NOT EXISTS (SELECT 1 FROM legal_acts la WHERE la.family_id = v.family_id);

scripts/backfill_material_axis.py

@@ -0,0 +1,221 @@
+"""안전 자료실 A-3 백필 — 기존 코퍼스에 material_type/jurisdiction/published_date/license 소급.
+
+plan: safety-library-1 A-3 (PKM plans/2026-06-12-safety-library-plan.html)
+선례: backfill_category.py (one-off 멱등 스크립트 — migration 아님, 152 단일 트랜잭션 제약 회피)
+
+술어 (2026-06-13 prod 실측 교정 — R2 blocker 반영):
+  1. extract_meta.source_id JOIN news_sources  → 레지스트리 material_type/country 전파
+     (KOSHA 사례 본문·CSB 페이지·HSE·MOEL·JPVT·arXiv·NB·TWI·API 공지 전부 커버.
+      paper 는 jurisdiction NULL 강제 — plan 0-1. KOSHA 본문의 kosha.kind='case' 가정은
+      실측 부정됨: kind 는 첨부/GUIDE 에만 존재 → source_id JOIN 이 정본 술어)
+  2. kosha.kind='case_attachment'              → incident/KR
+  3. kosha.kind='guide'                        → guide/KR (+ ofancYmd 'YYYY-MM-DD' 실측)
+  4. csb.kind='report_pdf'                     → incident/US (source_id 없음 — JOIN 비대상)
+  5. source_channel='law_monitor'              → law/KR (243건. legal_meta 생략 — MST 미보존,
+     버전 체인은 B-1 가동 시점부터. published_date = title 의 '(YYYYMMDD)' 공포일 추출 —
+     extract_meta 빈값 실측, R3-m1 의 'NULL 허용' 보다 1줄 정규식이 저렴해 채움)
+  6. file_path LIKE '%KGS_Code%'               → law/KR (frontmatter 키 = 'code' 실측 117/118,
+     'kgs_code' 0건. 경로 술어가 더 단순·전수. license 는 B-4 소관 — 미주입)
+
+불변식:
+  - 전 UPDATE 에 material_type IS NULL 가드 (멱등 — 재실행 안전, A-2 신규 유입분 무접촉)
+  - material_type + jurisdiction 동일 statement (law CHECK chk_documents_law_jurisdiction 충족)
+  - published_date / license 는 각자 필드 부재 가드 (이미 값 있으면 무접촉)
+  - 업로드 Industrial_Safety 문서 = 대상 아님 (LLM 제안+승인 경로만 — 자동 전이 금지)
+  - 코퍼스(청크/임베딩) 무접촉 — 검색 지표 무변동이 정상
+
+실행:
+  docker compose exec -T fastapi python /app/scripts/backfill_material_axis.py --dry-run
+  docker compose exec -T fastapi python /app/scripts/backfill_material_axis.py --apply
+"""
+
+import argparse
+import asyncio
+import os
+import sys
+
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", "app"))
+
+# text() 미사용 — exec_driver_sql (정규식 콜론 함정)
+from sqlalchemy.ext.asyncio import create_async_engine
+
+# ─── 술어별 (라벨, 카운트 SQL, 적용 SQL) ───────────────────────────────────────
+
+_KOSHA_LICENSE = ("kogl", "false", "한국산업안전보건공단(KOSHA)")
+_CSB_LICENSE = ("public_domain", "true", "U.S. Chemical Safety Board")
+_LAW_LICENSE = ("public_domain", "true", "국가법령정보센터")
+
+
+def _license_obj(scheme: str, redistribute: str, attribution: str) -> str:
+    return (
+        f"jsonb_build_object('license', jsonb_build_object("
+        f"'scheme', '{scheme}', 'redistribute', {redistribute}::boolean, "
+        f"'attribution', '{attribution}'))"
+    )
+
+
+STEPS: list[tuple[str, str]] = [
+    # 1) 레지스트리 전파 (source_id JOIN)
+    ("1. src_join material/jurisdiction", """
+        UPDATE documents d SET
+            material_type = ns.material_type,
+            jurisdiction = CASE WHEN ns.material_type = 'paper' THEN NULL ELSE ns.country END
+        FROM news_sources ns
+        WHERE d.material_type IS NULL AND d.deleted_at IS NULL
+          AND d.extract_meta->>'source_id' ~ '^[0-9]+$'
+          AND ns.id = (d.extract_meta->>'source_id')::int
+          AND ns.material_type IS NOT NULL
+    """),
+    # 2) KOSHA 첨부
+    ("2. kosha 첨부 incident/KR", """
+        UPDATE documents SET material_type = 'incident', jurisdiction = 'KR'
+        WHERE material_type IS NULL AND deleted_at IS NULL
+          AND extract_meta#>>'{kosha,kind}' = 'case_attachment'
+    """),
+    # 3) KOSHA GUIDE
+    ("3. kosha GUIDE guide/KR", """
+        UPDATE documents SET material_type = 'guide', jurisdiction = 'KR'
+        WHERE material_type IS NULL AND deleted_at IS NULL
+          AND extract_meta#>>'{kosha,kind}' = 'guide'
+    """),
+    # 4) CSB 보고서 PDF
+    ("4. csb PDF incident/US", """
+        UPDATE documents SET material_type = 'incident', jurisdiction = 'US'
+        WHERE material_type IS NULL AND deleted_at IS NULL
+          AND extract_meta#>>'{csb,kind}' = 'report_pdf'
+    """),
+    # 5) 레거시 law_monitor
+    ("5. law_monitor law/KR", """
+        UPDATE documents SET material_type = 'law', jurisdiction = 'KR'
+        WHERE material_type IS NULL AND deleted_at IS NULL
+          AND source_channel = 'law_monitor'
+    """),
+    # 6) KGS Code watch 폴더
+    ("6. KGS law/KR", """
+        UPDATE documents SET material_type = 'law', jurisdiction = 'KR'
+        WHERE material_type IS NULL AND deleted_at IS NULL
+          AND file_path LIKE '%KGS_Code%'
+    """),
+    # 7) published_date — crawl/news 공통 (extract_meta.published_at ISO)
+    ("7. published_date (published_at)", """
+        UPDATE documents SET published_date = (extract_meta->>'published_at')::date
+        WHERE published_date IS NULL AND deleted_at IS NULL
+          AND extract_meta->>'published_at' ~ '^\\d{4}-\\d{2}-\\d{2}'
+    """),
+    # 8) published_date — KOSHA GUIDE 공표일자 ('YYYY-MM-DD' 실측)
+    ("8. published_date (GUIDE ofancYmd)", """
+        UPDATE documents SET published_date = (extract_meta#>>'{kosha,ofancYmd}')::date
+        WHERE published_date IS NULL AND deleted_at IS NULL
+          AND extract_meta#>>'{kosha,ofancYmd}' ~ '^\\d{4}-\\d{2}-\\d{2}$'
+    """),
+    # 9) published_date — 레거시 law title 공포일 '(YYYYMMDD)'
+    ("9. published_date (law title 공포일)", """
+        UPDATE documents
+        SET published_date = to_date(substring(title from '\\((20\\d{2}(?:0[1-9]|1[0-2])(?:0[1-9]|[12]\\d|3[01]))\\)'), 'YYYYMMDD')
+        WHERE published_date IS NULL AND deleted_at IS NULL
+          AND source_channel = 'law_monitor'
+          AND title ~ '\\((20\\d{2}(?:0[1-9]|1[0-2])(?:0[1-9]|[12]\\d|3[01]))\\)'
+    """),
+    # 10) license — 레지스트리 전파 (scheme 있는 소스만)
+    ("10. license (src_join)", """
+        UPDATE documents d SET
+            extract_meta = COALESCE(d.extract_meta, '{}'::jsonb)
+                || jsonb_build_object('license', jsonb_build_object(
+                       'scheme', ns.license_scheme,
+                       'redistribute', COALESCE(ns.license_redistribute, false),
+                       'attribution', ns.name))
+        FROM news_sources ns
+        WHERE d.deleted_at IS NULL AND NOT (COALESCE(d.extract_meta, '{}'::jsonb) ? 'license')
+          AND d.extract_meta->>'source_id' ~ '^[0-9]+$'
+          AND ns.id = (d.extract_meta->>'source_id')::int
+          AND ns.license_scheme IS NOT NULL
+    """),
+    # 11) license — KOSHA 첨부/GUIDE (source_id 없음)
+    ("11. license (kosha kinds)", f"""
+        UPDATE documents SET
+            extract_meta = COALESCE(extract_meta, '{{}}'::jsonb) || {_license_obj(*_KOSHA_LICENSE)}
+        WHERE deleted_at IS NULL AND NOT (COALESCE(extract_meta, '{{}}'::jsonb) ? 'license')
+          AND extract_meta#>>'{{kosha,kind}}' IN ('case_attachment', 'guide')
+    """),
+    # 12) license — CSB PDF
+    ("12. license (csb PDF)", f"""
+        UPDATE documents SET
+            extract_meta = COALESCE(extract_meta, '{{}}'::jsonb) || {_license_obj(*_CSB_LICENSE)}
+        WHERE deleted_at IS NULL AND NOT (COALESCE(extract_meta, '{{}}'::jsonb) ? 'license')
+          AND extract_meta#>>'{{csb,kind}}' = 'report_pdf'
+    """),
+    # 13) license — 레거시 법령 (저작권법 제7조 비보호)
+    ("13. license (law_monitor)", f"""
+        UPDATE documents SET
+            extract_meta = COALESCE(extract_meta, '{{}}'::jsonb) || {_license_obj(*_LAW_LICENSE)}
+        WHERE deleted_at IS NULL AND NOT (COALESCE(extract_meta, '{{}}'::jsonb) ? 'license')
+          AND source_channel = 'law_monitor'
+    """),
+]
+
+VERIFY_SQL = [
+    ("축 전수표 (material_type x jurisdiction)", """
+        SELECT material_type, jurisdiction, count(*) AS docs,
+               count(published_date) AS with_date,
+               count(*) FILTER (WHERE extract_meta ? 'license') AS with_license
+        FROM documents WHERE material_type IS NOT NULL AND deleted_at IS NULL
+        GROUP BY 1, 2 ORDER BY 1, 2
+    """),
+    ("law & jurisdiction NULL (0 이어야 함 — hard)", """
+        SELECT count(*) FROM documents
+        WHERE material_type = 'law' AND jurisdiction IS NULL AND deleted_at IS NULL
+    """),
+    ("잔여 미분류 안전 후보 (kosha/csb 메타 보유인데 NULL — 0 이어야 함)", """
+        SELECT count(*) FROM documents
+        WHERE material_type IS NULL AND deleted_at IS NULL
+          AND (extract_meta ? 'kosha' OR extract_meta ? 'csb')
+    """),
+]
+
+
+async def main() -> None:
+    parser = argparse.ArgumentParser()
+    mode = parser.add_mutually_exclusive_group(required=True)
+    mode.add_argument("--dry-run", action="store_true",
+                      help="전 UPDATE 를 트랜잭션 안에서 실행해 정확한 rowcount + 검증표를 보여주고 ROLLBACK (변경 0)")
+    mode.add_argument("--apply", action="store_true", help="백필 실행 (단일 트랜잭션 커밋)")
+    args = parser.parse_args()
+
+    db_url = os.getenv(
+        "DATABASE_URL", "postgresql+asyncpg://pkm:pkm@localhost:5432/pkm"
+    )
+    engine = create_async_engine(db_url)
+    tag = "apply" if args.apply else "dry-run"
+
+    async with engine.connect() as conn:
+        trans = await conn.begin()
+        try:
+            for label, sql in STEPS:
+                # text() 는 정규식의 '(?:' 콜론을 bind param 으로 오인 (migration 러너와
+                # 동일 함정) → driver 직결 실행
+                result = await conn.exec_driver_sql(sql)
+                print(f"[{tag}] {label}: {result.rowcount}행")
+
+            print("\n─── 검증 (트랜잭션 내 미리보기) ───")
+            for label, sql in VERIFY_SQL:
+                result = await conn.exec_driver_sql(sql)
+                rows = result.fetchall()
+                print(f"\n{label}:")
+                for row in rows:
+                    print("  ", tuple(row))
+
+            if args.apply:
+                await trans.commit()
+                print("\n[apply] 커밋 완료")
+            else:
+                await trans.rollback()
+                print("\n[dry-run] 전체 롤백 — 변경 0")
+        except Exception:
+            await trans.rollback()
+            raise
+
+    await engine.dispose()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

scripts/verify_statute_chain.py

@@ -0,0 +1,138 @@
+"""B-1 PR③ — 법령 버전 체인 검증 3술어 (plan safety-library-1).
+
+read-only 진단 — E-1 관찰의 법령 게이트 도구로도 재사용 (반복 실행 안전).
+
+검증 3술어 (R7-M1, B-1 단일 정본):
+  ① 존재성 — watch family 각각 primary 시리즈 current 정확 1건(0건도 위반)
+              + annex 시리즈당 current ≤ 1
+  ② 노출 유일성 — primary current 보유 family당 primary 노출(체인+레거시 매핑 합산) 정확 1건
+                  (모집단 = primary current 보유 family 한정 — R8-M2)
+  ③ 고아 그물 — law_monitor in_corpus=true 레거시 중:
+                (a) current 보유 family 에 매핑되는데 안 flip 된 것(flip 누락) = 0
+                (b) 어느 watch family 에도 매핑 안 되는 것(제명 개정 등 매핑 구멍) = 0
+                repealed family·primary current 미보유 family 의 레거시 보존은 위반 아님
+
+repealed family 는 ①② 기대값 0 으로 면제.
+
+실행:
+  docker compose exec -T fastapi python /app/scripts/verify_statute_chain.py
+종료코드: 0 = 전건 PASS, 1 = 위반 (CI/관찰 게이트 용)
+"""
+
+import asyncio
+import os
+import sys
+
+# 컨테이너: /app/scripts → /app (workers/core/models 패키지 루트). 로컬: repo/scripts → repo/app
+_here = os.path.dirname(os.path.abspath(__file__))
+for _cand in (os.path.join(_here, ".."), os.path.join(_here, "..", "app")):
+    if os.path.isdir(os.path.join(_cand, "workers")):
+        sys.path.insert(0, os.path.abspath(_cand))
+        break
+
+from collections import defaultdict
+
+from sqlalchemy import select, text
+from sqlalchemy.ext.asyncio import create_async_engine
+
+from workers.statute_collector import legacy_law_name, normalize_law_name, series_suffix
+
+
+async def main() -> int:
+    db_url = os.getenv("DATABASE_URL", "postgresql+asyncpg://pkm:pkm@localhost:5432/pkm")
+    engine = create_async_engine(db_url)
+    violations: list[str] = []
+
+    async with engine.connect() as conn:
+        # ── 로드 ──
+        acts = (await conn.execute(text(
+            "SELECT family_id, title, repeal_detected_at IS NOT NULL AS repealed "
+            "FROM legal_acts WHERE watch"))).all()
+        metas = (await conn.execute(text(
+            "SELECT family_id, law_doc_kind, version_key, version_status FROM legal_meta"))).all()
+
+        act_title = {a.family_id: a.title for a in acts}
+        repealed = {a.family_id for a in acts if a.repealed}
+        active = [a for a in acts if not a.repealed]
+
+        # family → primary current 수 / annex 시리즈별 current 수
+        prim_current = defaultdict(int)
+        annex_series_current = defaultdict(int)
+        for m in metas:
+            if m.version_status != "current":
+                continue
+            if m.law_doc_kind == "primary":
+                prim_current[m.family_id] += 1
+            else:
+                annex_series_current[(m.family_id, series_suffix(m.version_key))] += 1
+
+        # ── ① 존재성 ──
+        for a in active:
+            n = prim_current[a.family_id]
+            if n != 1:
+                violations.append(f"① {a.family_id} ({a.title}): primary current {n}건 (정확 1 기대)")
+        for (fid, suf), n in annex_series_current.items():
+            if fid not in repealed and n > 1:
+                violations.append(f"① {fid} annex 시리즈 {suf}: current {n}건 (≤1 기대)")
+
+        # ── ③ 고아 그물 (정규화 동등 매핑) ──
+        # watch family 정규화명 → family_id (current 보유 여부 동반)
+        norm_to_fid = {}
+        for a in active:
+            norm_to_fid[normalize_law_name(a.title)] = a.family_id
+
+        legacy = (await conn.execute(text(
+            "SELECT d.id, d.title, "
+            "  EXISTS(SELECT 1 FROM document_chunks c WHERE c.doc_id=d.id AND c.in_corpus) AS exposed "
+            "FROM documents d WHERE d.source_channel='law_monitor' AND d.deleted_at IS NULL"))).all()
+
+        orphan_flip_miss = 0
+        orphan_unmapped = 0
+        unmapped_names = set()
+        for row in legacy:
+            if not row.exposed:
+                continue  # in_corpus=false = 정상 (스윕됨 or 청크 없음)
+            name = legacy_law_name(row.title or "")
+            norm = normalize_law_name(name) if name else None
+            fid = norm_to_fid.get(norm) if norm else None
+            if fid is None:
+                orphan_unmapped += 1
+                if name:
+                    unmapped_names.add(name)
+            elif prim_current.get(fid, 0) >= 1:
+                # current 보유 family 인데 레거시가 노출 중 = flip 누락
+                orphan_flip_miss += 1
+        if orphan_flip_miss:
+            violations.append(f"③(a) flip 누락: current 보유 family 의 노출 레거시 {orphan_flip_miss}건")
+        if orphan_unmapped:
+            violations.append(
+                f"③(b) 무매핑 노출 레거시 {orphan_unmapped}건 — 매핑 구멍(매핑 보강 신호): "
+                + ", ".join(sorted(unmapped_names))[:200])
+
+        # ── ② 노출 유일성 (primary current 보유 family 한정) ──
+        # 노출 primary = 체인 primary current(=1) + 레거시 매핑 노출분.
+        # ③(a)=0 이면 레거시 노출분 0 → 체인 1건만 = 정확 1. 별도 위반 추출은 ③(a)에 포함됨.
+        # (annex 노출 비동기 일반화는 may — Phase 1 미적용)
+
+        # ── 상태 요약 출력 ──
+        print("=== 법령 체인 검증 (B-1 PR③ 3술어) ===")
+        print(f"watch family: {len(acts)} (active {len(active)}, repealed {len(repealed)})")
+        print(f"primary current 보유 family: {sum(1 for a in active if prim_current[a.family_id]==1)}/{len(active)}")
+        print(f"annex current 시리즈: {len(annex_series_current)}")
+        exposed_legacy = sum(1 for r in legacy if r.exposed)
+        print(f"레거시 law_monitor: {len(legacy)}건 (in_corpus 노출 {exposed_legacy}건)")
+        print()
+
+    await engine.dispose()
+
+    if violations:
+        print(f"[FAIL] 위반 {len(violations)}건:")
+        for v in violations:
+            print("  -", v)
+        return 1
+    print("[PASS] 3술어 전건 통과 (존재성·노출 유일성·고아 그물)")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(asyncio.run(main()))

tests/fixtures/kosha_fatal_response.json

@@ -0,0 +1 @@
+{"header": {"resultCode": "00", "resultMsg": "NORMAL_CODE"}, "body": {"pageNo": 1, "totalCount": 2845, "numOfRows": 3, "items": {"item": [{"contents": "<p><img src='https://portal.kosha.or.kr/api/compn24/auth/stdtboard/getImage.do?bbsId=B2025021314108&pstNo=20260611111536KIZXJ8&bbsAtcflNo=E0802000030001' style='width: 931px;' data-filename='6-9 부산 사상구.jpg' data-tboard-img-cvrt='Y'></p><p><br></p><p>2026. 6. 9. (화), 12:22경부산 사상구 소재 아파트에서</p><p><br></p><p>재해자가 2명이 실외기 설치 작업 중</p><p><br></p><p>베란다 난간이 파손되며 바닥으로 떨어짐</p><p><br></p><p>(사망 2명)</p><p>※ 위 내용은 신고 및 현재 파악된 내용으로 조사결과에 따라 변경될 수 있습니다.</p><div><br></div>", "keyword": "[6/9, 부산 사상구] 실외기 설치 작업 중 베란다 난간이 파손되어 떨어짐", "arno": "20260611111536KIZXJ8"}, {"contents": "<p><br><img src='https://portal.kosha.or.kr/api/compn24/auth/stdtboard/getImage.do?bbsId=B2025021314108&pstNo=20260611111355OZSS9T&bbsAtcflNo=E0802000030001' style='width: 931px;' data-filename='서 울관악구.jpg' data-tboard-img-cvrt='Y'></p><p><br></p><p>2026. 6. 9. (화), 17:26경서울 관악구 철도 공사 현장에서</p><p><br></p><p>재해자가 수직형 케이블 거치대 설치 준비 작업 중</p><p><br></p><p>개구부로 떨어짐(사망 1명)</p><p><br></p><p>※ 위 내용은 신고 및 현재 파악된 내용으로 조사결과에 따라 변경될 수 있습니다.</p><div><br></div><p></p>", "keyword": "[6/9, 서울 관악구] 수직형 케이블 거치대 설치 준비 중 개구부로 떨어짐", "arno": "20260611111355OZSS9T"}, {"contents": "<p><img src='https://portal.kosha.or.kr/api/compn24/auth/stdtboard/getImage.do?bbsId=B2025021314108&pstNo=202606111110595AR9QY&bbsAtcflNo=E0802000030001' style='width: 931px;' data-filename='5-14 전남 광양시.jpg' data-tboard-img-cvrt='Y'><br></p><p><br></p><p>2026. 5. 14. (목), 16:51경전남 광양시 소재 화학물질 제조사업장에서</p><p><br></p><p>재해자가 정제설비 내부에서 플랜지 해체 작업 중</p><p><br></p><p>고온 응축수가 쏟아져 화상을 입음(사망 1명)※ 위 내용은 신고 및 현재 파악된 내용으로 조사결과에 따라 변경될 수 있습니다.<br></p>", "keyword": "[5/14, 전남 광양시] 플랜지 해체 작업 중 고온 응축수가 쏟아져 화상", "arno": "202606111110595AR9QY"}]}}}

tests/fixtures/statute_kr/lawjosub_probe.xml

@@ -0,0 +1,46 @@
+<?xml version="1.0" encoding="utf-8"?>
+<법령 법령키="0017662026021921374">
+<기본정보>
+<법령ID>001766</법령ID>
+<공포일자>20260219</공포일자>
+<공포번호>21374</공포번호>
+<언어>한글</언어>
+<법종구분 법종구분코드="A0002">법률</법종구분>
+<법령명_한글><![CDATA[산업안전보건법]]></법령명_한글>
+<법령명_한자><![CDATA[산업안전보건법]]></법령명_한자>
+<제명변경여부>N</제명변경여부>
+<한글법령여부>Y</한글법령여부>
+<편장절관>40040000</편장절관>
+<소관부처 소관부처코드="1492000">고용노동부</소관부처>
+<전화번호>044-202-8810, 8813, 8815, 8997</전화번호>
+<시행일자>20260601</시행일자>
+<제개정구분>일부개정</제개정구분>
+<조문별시행일자>20260601</조문별시행일자>
+<조문시행일자문자열>20260801:제10조의2, 제23조, 제175조제4항제1호의2</조문시행일자문자열>
+<별표편집여부>N</별표편집여부>
+<공포법령여부>Y</공포법령여부>
+<시행일기준편집여부>Y</시행일기준편집여부>
+</기본정보>
+<조문>
+<조문단위 조문키="0001000">
+<조문번호>1</조문번호>
+<조문여부>전문</조문여부>
+<조문시행일자>20260601</조문시행일자>
+<조문이동이전></조문이동이전>
+<조문이동이후></조문이동이후>
+<조문변경여부>N</조문변경여부>
+</조문단위>
+<조문단위 조문키="0001001">
+<조문번호>1</조문번호>
+<조문여부>조문</조문여부>
+<조문제목><![CDATA[목적]]></조문제목>
+<조문시행일자>20260601</조문시행일자>
+<조문이동이전></조문이동이전>
+<조문이동이후></조문이동이후>
+<조문변경여부>N</조문변경여부>
+<조문내용>
+<![CDATA[제1조(목적) 이 법은 산업 안전 및 보건에 관한 기준을 확립하고 그 책임의 소재를 명확하게 하여 산업재해를 예방하고 쾌적한 작업환경을 조성함으로써 노무를 제공하는 사람의 안전 및 보건을 유지ㆍ증진함을 목적으로 한다. <개정 2020.5.26>]]>
+</조문내용>
+</조문단위>
+</조문>
+</법령>

tests/fixtures/statute_kr/lawsearch_rule.xml

@@ -0,0 +1,2 @@
+<?xml version="1.0" encoding="UTF-8"?><LawSearch><target>law</target><키워드>산업안전보건기준에 관한 규칙</키워드><section>lawNm</section><totalCnt>1</totalCnt><page>1</page><numOfRows>1</numOfRows><resultCode>00</resultCode><resultMsg>success</resultMsg><law id="1"><법령일련번호>273603</법령일련번호><현행연혁코드>현행</현행연혁코드><법령명한글><![CDATA[산업안전보건기준에 관한 규칙]]></법령명한글><법령약칭명><![CDATA[안전보건규칙]]></법령약칭명><법령ID>007363</법령ID><공포일자>20250901</공포일자><공포번호>00450</공포번호><제개정구분명>일부개정</제개정구분명><소관부처코드>1492000</소관부처코드><소관부처명>고용노동부</소관부처명><법령구분명>고용노동부령</법령구분명><공동부령정보></공동부령정보><시행일자>20260302</시행일자><자법타법여부></자법타법여부><법령상세링크>/DRF/lawService.do?OC=__OC_REDACTED__&amp;target=law&amp;MST=273603&amp;type=HTML&amp;mobileYn=&amp;efYd=20260302</법령상세링크></law></LawSearch>
+

tests/fixtures/statute_kr/lawsearch_sanab.xml

@@ -0,0 +1,2 @@
+<?xml version="1.0" encoding="UTF-8"?><LawSearch><target>law</target><키워드>산업안전보건법</키워드><section>lawNm</section><totalCnt>3</totalCnt><page>1</page><numOfRows>3</numOfRows><resultCode>00</resultCode><resultMsg>success</resultMsg><law id="1"><법령일련번호>283449</법령일련번호><현행연혁코드>현행</현행연혁코드><법령명한글><![CDATA[산업안전보건법]]></법령명한글><법령약칭명><![CDATA[]]></법령약칭명><법령ID>001766</법령ID><공포일자>20260219</공포일자><공포번호>21374</공포번호><제개정구분명>일부개정</제개정구분명><소관부처코드>1492000</소관부처코드><소관부처명>고용노동부</소관부처명><법령구분명>법률</법령구분명><공동부령정보></공동부령정보><시행일자>20260601</시행일자><자법타법여부></자법타법여부><법령상세링크>/DRF/lawService.do?OC=__OC_REDACTED__&amp;target=law&amp;MST=283449&amp;type=HTML&amp;mobileYn=&amp;efYd=20260601</법령상세링크></law><law id="2"><법령일련번호>284771</법령일련번호><현행연혁코드>현행</현행연혁코드><법령명한글><![CDATA[산업안전보건법 시행령]]></법령명한글><법령약칭명><![CDATA[]]></법령약칭명><법령ID>003786</법령ID><공포일자>20260324</공포일자><공포번호>36220</공포번호><제개정구분명>타법개정</제개정구분명><소관부처코드>1492000</소관부처코드><소관부처명>고용노동부</소관부처명><법령구분명>대통령령</법령구분명><공동부령정보></공동부령정보><시행일자>20260324</시행일자><자법타법여부></자법타법여부><법령상세링크>/DRF/lawService.do?OC=__OC_REDACTED__&amp;target=law&amp;MST=284771&amp;type=HTML&amp;mobileYn=&amp;efYd=20260324</법령상세링크></law><law id="3"><법령일련번호>286657</법령일련번호><현행연혁코드>현행</현행연혁코드><법령명한글><![CDATA[산업안전보건법 시행규칙]]></법령명한글><법령약칭명><![CDATA[]]></법령약칭명><법령ID>007364</법령ID><공포일자>20260529</공포일자><공포번호>00470</공포번호><제개정구분명>일부개정</제개정구분명><소관부처코드>1492000</소관부처코드><소관부처명>고용노동부</소관부처명><법령구분명>고용노동부령</법령구분명><공동부령정보></공동부령정보><시행일자>20260601</시행일자><자법타법여부></자법타법여부><법령상세링크>/DRF/lawService.do?OC=__OC_REDACTED__&amp;target=law&amp;MST=286657&amp;type=HTML&amp;mobileYn=&amp;efYd=20260601</법령상세링크></law></LawSearch>
+

tests/fixtures/statute_kr/seed_26laws.tsv

@@ -0,0 +1,26 @@
+산업안전보건법	001766	283449	20260219	20260601	법률
+산업안전보건법 시행령	003786	284771	20260324	20260324	대통령령
+산업안전보건법 시행규칙	007364	286657	20260529	20260601	고용노동부령
+산업안전보건기준에 관한 규칙	007363	273603	20250901	20260302	고용노동부령
+유해위험작업의 취업 제한에 관한 규칙	MISS				
+중대재해 처벌 등에 관한 법률	013993	228817	20210126	20220127	법률
+중대재해 처벌 등에 관한 법률 시행령	014159	277417	20251001	20251001	대통령령
+건설기술 진흥법	001807	276921	20251001	20251001	법률
+건설기술 진흥법 시행령	002111	286847	20260609	20260609	대통령령
+건설기술 진흥법 시행규칙	006175	286885	20260611	20260611	국토교통부령
+시설물의 안전 및 유지관리에 관한 특별법	000237	266683	20241203	20251204	법률
+위험물안전관리법	009502	259933	20240206	20250807	법률
+위험물안전관리법 시행령	009707	273077	20250805	20250807	대통령령
+위험물안전관리법 시행규칙	009732	262765	20240520	20250521	행정안전부령
+화학물질관리법	000162	276815	20251001	20251001	법률
+화학물질관리법 시행령	004390	280507	20251223	20251223	대통령령
+화학물질의 등록 및 평가 등에 관한 법률	011857	279805	20251111	20260512	법률
+소방시설 설치 및 관리에 관한 법률	009503	236977	20211130	20241201	법률
+소방시설 설치 및 관리에 관한 법률 시행령	009694	284781	20260324	20260324	대통령령
+전기사업법	001854	283981	20260310	20260310	법률
+전기안전관리법	013718	268805	20250131	20260201	법률
+고압가스 안전관리법	001850	283919	20260310	20260310	법률
+고압가스 안전관리법 시행령	002246	286839	20260609	20260609	대통령령
+액화석유가스의 안전관리 및 사업법	001849	276549	20251001	20251128	법률
+근로기준법	001872	265959	20241022	20251023	법률
+환경영향평가법	002016	276833	20251001	20251023	법률

tests/search_eval/runs/p2a_baseline_exact_20260612_0220.csv

@@ -0,0 +1,52 @@
+label,id,category,legacy_category,intent,domain_hint,language,ocr_derived,failure_expected,query,relevant_ids,graded_relevance,returned_ids_top10,latency_ms,recall_at_10,mrr_at_10,ndcg_at_10,top3_hit,graded_ndcg_at_10,graded_recall_at_10_t2,graded_recall_at_10_t3,dedup_count,error
+single,kw_001,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건법 제6장,3856;3868;3879,3856:3;3868:2;3879:2,30411;30423;30406;30418;3879;3868;3890;3863;3856;3908,925.1,1.000,0.200,0.490,0,0.417,1.000,1.000,0,
+single,kw_002,standards,exact_keyword,fact_lookup,document,ko,0,0,중대재해 처벌 등에 관한 법률 제2장 중대산업재해,3917;3921,3917:3;3921:2,3921;3917;3919;3923;30418;10573;10571;3916;3874;3918,549.1,1.000,1.000,1.000,1,0.834,1.000,1.000,0,
+single,kw_003,standards,exact_keyword,fact_lookup,document,ko,0,0,화학물질관리법 유해화학물질 영업자,3981,3981:3,3981;3985;3980;3984;3993;30412;3857;3978;3986;3983,350.4,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_004,standards,exact_keyword,fact_lookup,document,ko,0,0,근로기준법 안전과 보건,4041,4041:3,4041;3852;30407;3851;30406;30421;3877;30410;30422;3905,657.8,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_005,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 보호구,3888,3888:3,10570;3888;3912;3913;3911;3905;3909;3889;3910;3902,665.5,1.000,0.500,0.631,1,0.631,1.000,1.000,0,
+single,nl_001,korean_only,natural_language_ko,semantic_search,document,ko,0,0,기계로 인한 산업재해 관련 법령,3856;3868;3879;3854,3854:1;3856:3;3868:2;3879:2,5244;30422;30410;3878;5249;3897;3863;30421;3868;5253,697.8,0.250,0.111,0.118,0,0.083,0.333,0.000,0,
+single,nl_002,korean_only,natural_language_ko,semantic_search,document,ko,0,0,사업주가 도급을 줄 때 산업재해를 예방하기 위해 해야 할 일,3855;3867;3878,3855:3;3867:2;3878:2,30410;3855;5227;30422;30409;3917;3854;5244;30406;3867,566.2,0.667,0.500,0.432,1,0.508,0.667,1.000,0,
+single,nl_003,korean_only,natural_language_ko,semantic_search,document,ko,0,0,유해화학물질을 다루는 회사가 지켜야 할 안전 의무,3980;3981;3982,3980:2;3981:2;3982:2,3980;3903;3904;3981;3760;5253;3985;3896;3917;3909,554.8,0.667,1.000,0.671,1,0.671,0.667,0.000,0,
+single,nl_004,korean_only,natural_language_ko,semantic_search,document,ko,0,0,중대재해가 발생했을 때 경영책임자가 처벌받는 기준,3916;3917;3920;3921,3916:2;3917:3;3920:2;3921:2,10572;10573;3917;3918;3916;3923;5244;3919;10571;30409,418.8,0.500,0.333,0.346,1,0.399,0.500,1.000,0,
+single,nl_005,korean_only,natural_language_ko,semantic_search,document,ko,0,0,안전보건교육은 누가 받아야 하고 어떤 내용을 다루는가,3853;3865,3853:3;3865:2,3853;3876;5249;5234;4025;30420;30408;30414;6675;11677,918.5,0.500,1.000,0.613,1,0.787,0.500,1.000,0,
+single,news_001,korean_only,news_ko,semantic_search,news,ko,0,0,이란과 미국의 군사 충돌,4303;4304;4307;4316;4322;4323;4327;4335,4303:2;4304:2;4307:2;4316:2;4322:2;4323:2;4327:2;4335:2,37599;27583;35225;27350;37601;32410;16081;37592;28407;37608,674.1,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_002,korean_only,news_ko,semantic_search,news,ko,0,0,호르무즈 해협 봉쇄,4316;4320;4322;4327,4316:3;4320:2;4322:2;4327:2,41652;41650;41634;31894;35899;22049;16392;9022;17123;11945,411.6,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,misc_001,korean_only,other_domain,fact_lookup,document,ko,0,0,강체의 평면 운동학,4063;4065,4063:3;4065:2,4063;4065;4064;4071;4066;4068;4058;4067;5063;5105,653.8,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,misc_002,korean_only,other_domain,semantic_search,document,ko,0,0,질점의 운동역학,4060;4061;4062,4060:2;4061:2;4062:2,4060;4062;4070;4059;4061;4064;4065;4063;4058;5095,552.8,1.000,1.000,0.947,1,0.947,1.000,0.000,0,
+single,news_003,english_only,news_en,semantic_search,news,en,0,0,Trump Iran ultimatum,4258;4260;4262,4258:2;4260:2;4262:2,21155;4775;4202;4776;4679;4199;4519;4668;37733;25592,462.7,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_001,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,기계 안전 가드 설계 원리,3770;3856,3770:3;3856:2,35778;5239;5244;3770;3817;3774;4540;3762;3789;5249,819.4,0.500,0.250,0.264,0,0.339,0.500,1.000,0,
+single,cl_002,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,산업 안전 입문서,3755;3775;3776;3777,3755:2;3775:2;3776:2;3777:2,5249;3760;5229;3755;3774;5230;10573;3775;3787;3818,644.5,0.500,0.250,0.291,1,0.291,0.500,0.000,0,
+single,cl_003,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,전기 안전 위험,3772;3790,3772:2;3790:2,3772;35779;3790;5260;3897;37326;3755;37216;37011;5235,1193.3,1.000,1.000,0.920,1,0.920,1.000,0.000,0,
+single,news_004,mixed,news_fr,semantic_search,news,mixed,0,0,guerre en Iran,4199;4202;4210;4361;4363;4507;4519;4521,4199:2;4202:2;4210:2;4361:2;4363:2;4507:2;4519:2;4521:2,33101;33368;7740;27078;5925;5840;16010;16457;6945;4507,469.1,0.125,0.100,0.073,1,0.073,0.125,0.000,0,
+single,news_005,mixed,news_crosslingual,semantic_search,news,mixed,0,0,이란 미국 전쟁 글로벌 반응,4202;4258;4262;4536;4303;4304;4316,4202:2;4258:2;4262:2;4303:2;4304:2;4316:2;4536:2,21275;16927;4457;19111;16761;35246;41629;35894;6789;35908,844.5,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_001,failure_expected,failure_expected,semantic_search,document,mixed,0,1,Rust async runtime tokio scheduler 내부 구조,,,5262;23732;36544;35645;37582;3774;30259;28766;31709;35635,533.4,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_002,failure_expected,failure_expected,semantic_search,document,ko,0,1,양자컴퓨터 큐비트 디코히어런스,,,31984;29686;16289;16384;5066;33238;20507;30566;5092;32721,290.9,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_003,failure_expected,failure_expected,semantic_search,news,ko,0,1,재즈 보컬리스트 빌리 홀리데이,,,31643;20470;35249;4634;26593;16059;9102;23082;23336;31187,265.7,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,kw_006,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 작업장 통로,3886;3887,3886:3;3887:2,3898;3886;3908;3887;3881;3911;3912;3892;3900;3909,506.2,1.000,0.500,0.651,1,0.642,1.000,1.000,0,
+single,kw_007,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 폭발 화재 위험물 누출 방지,3896;3766,3766:1;3896:3,3896;3895;3903;13930;13931;13929;3894;3886;13944;3892,477.6,0.500,1.000,0.613,1,0.917,1.000,1.000,0,
+single,kw_008,standards,standards,fact_lookup,document,ko,0,0,고압가스 안전관리법 전문,4025;4026,4025:3;4026:2,11644;11579;35775;4026;4025;11693;11645;4034;13299;13749,538.7,1.000,0.250,0.501,0,0.450,1.000,1.000,0,
+single,kw_009,standards,standards,fact_lookup,document,ko,0,0,KGS FP111 가스설비 배관설비 기준,13305,13305:3,13305;13311;13306;13312;13302;13304;13309;13299;13313;13918,446.0,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_010,standards,standards,fact_lookup,document,ko,0,0,KGS FU551 가스설비 압력조정기 가스계량기,13652,13652:3,13652;11689;13657;13655;13656;13649;13651;13752;13659;13650,312.5,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_011,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 전기로 인한 위험 방지,3897;3772,3772:1;3897:3,3897;3895;3887;13935;3902;3896;3877;3900;3899;3854,587.9,0.500,1.000,0.613,1,0.917,1.000,1.000,0,
+single,en_001,english_only,english_only,semantic_search,document,en,0,0,pressure vessel flange design,5144;5136,5136:2;5144:3,5144;5140;5136;5180;36899;5137;5149;5178;5207;5148,4162.7,1.000,1.000,0.920,1,0.956,1.000,1.000,0,
+single,en_002,english_only,english_only,semantic_search,document,en,0,0,ASME Section VIII introduction,5204;5206,5204:3;5206:2,5204;5212;5180;5208;5210;5143;5141;5206;5137;11634,3038.6,1.000,1.000,0.807,1,0.894,1.000,1.000,0,
+single,en_003,english_only,english_only,semantic_search,document,en,0,0,impact test requirements ASME,5205;5148,5148:1;5205:3,36721;5205;5204;5178;5214;5224;5148;5140;5209;5145,5221.9,1.000,0.500,0.591,1,0.622,1.000,1.000,0,
+single,en_004,english_only,english_only,semantic_search,document,en,0,0,design of vessel supports,5149,5149:3,5149;5186;5141;5137;36901;36869;5140;5139;5136;5178,5732.5,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,en_005,english_only,english_only,semantic_search,document,en,0,0,hydrogen piping ASME code,5178,5178:3,5178;5180;5210;5212;5182;5133;5207;5148;5204;5209,4941.9,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,en_006,english_only,english_only,semantic_search,document,en,0,0,ASME welding qualification requirements,5209;3771,3771:1;5209:3,37097;5224;36877;36721;5208;5209;5205;5178;5204;5180,2432.9,0.500,0.167,0.218,0,0.327,1.000,1.000,0,
+single,en_007,english_only,english_only,semantic_search,document,en,0,0,pressure vessel fabrication and inspection,5208;5145,5145:2;5208:3,5208;38747;38791;36935;5189;36874;5180;36908;5187;36901,3828.4,0.500,1.000,0.613,1,0.787,0.500,1.000,0,
+single,en_008,english_only,english_only,semantic_search,document,en,0,0,Industrial Safety and Health Management ergonomics,3763;3755,3755:1;3763:3,3763;3759;3774;3755;3818;3812;3778;3756;3761;3769,1866.0,1.000,1.000,0.877,1,0.974,1.000,1.000,0,
+single,cl_004,mixed,mixed,semantic_search,document,mixed,0,0,ASME 압력용기 설계 실무,5207;5210;5139;5135,5135:2;5139:2;5207:3;5210:3,5139;5207;5204;5225;5206;5208;5210;5137;5182;5145,805.5,0.750,1.000,0.767,1,0.686,0.750,1.000,0,
+single,cl_005,mixed,mixed,semantic_search,document,mixed,0,0,ASME 용접 코드 해설,5224;5209,5209:2;5224:3,5224;5222;5225;5209;5204;5210;5205;5178;5143;5141,737.6,1.000,1.000,0.877,1,0.932,1.000,1.000,0,
+single,cl_006,mixed,mixed,semantic_search,document,mixed,0,0,pressure vessel Korean industrial safety regulation,4025;3856;5136,3856:2;4025:2;5136:1,38941;4026;36874;5145;5186;38791;39084;5143;3895;5210,1231.3,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_007,mixed,mixed,semantic_search,document,mixed,0,0,KGS 코드 LPG safety standard,11647;11689;11645;4025,4025:1;11645:2;11647:3;11689:2,11647;13760;13674;13669;13774;13773;13675;13755;13924;11688,528.5,0.250,1.000,0.390,1,0.647,0.333,1.000,0,
+single,cl_008,mixed,mixed,semantic_search,document,mixed,0,0,수소 가스 안전 기준 hydrogen safety,5178;5169,5169:2;5178:3,10575;5177;11671;11649;11648;13915;5241;5173;11653;5170,872.1,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,exam_001,exam,exam,fact_lookup,document,ko,0,0,레이놀즈수 정의 공식,11504;11505,11504:3;11505:2,11504;11533;5090;4544;11509;11505;11476;11486;5132;5140,628.8,1.000,1.000,0.832,1,0.907,1.000,1.000,0,
+single,exam_002,exam,exam,fact_lookup,document,ko,0,0,탱크 바닥 구멍 유체 유속 토리첼리,11500;11495;11496,11495:2;11496:2;11500:3,11500;11495;11501;5139;5090;5140;5178;11515;11517;5210,338.1,0.667,1.000,0.765,1,0.856,0.667,1.000,0,
+single,exam_003,exam,exam,fact_lookup,document,ko,0,0,이상기체 음속 마하수,11514;11515,11514:3;11515:2,11514;5083;11475;5090;5084;11531;11476;11473;11479;5124,675.8,0.500,1.000,0.613,1,0.787,0.500,1.000,0,
+single,exam_004,exam,exam,fact_lookup,document,ko,0,0,고압가스 용기 내압시험 영구증가량,11591;11644;11691,11591:3;11644:2;11691:2,11591;11664;13660;5177;11691;11665;13652;13942;13917;11660,357.3,0.667,1.000,0.651,1,0.785,0.667,1.000,0,
+single,exam_005,exam,exam,fact_lookup,document,ko,0,0,도시가스 배관 매설 이격거리,11627;11625;11646,11625:2;11627:3;11646:1,11627;11658;11600;11625;13918;11692;13751;5177;11655;13653,400.9,0.667,1.000,0.671,1,0.883,1.000,1.000,0,
+single,exam_006,exam,exam,fact_lookup,document,ko,0,0,LPG 저장탱크 안전거리 분말소화기,11617;11669;11620,11617:3;11620:1;11669:2,11595;11616;13669;11617;11649;11605;11655;11690;11658;13915,349.0,0.333,0.250,0.202,0,0.321,0.500,1.000,0,
+single,exam_007,exam,exam,fact_lookup,document,ko,0,0,오리피스 차압식 유량계,11712;11711;11503,11503:2;11711:2;11712:3,11711;11712;11503;11500;11714;11713;13930;11717;11701;11715,416.5,1.000,1.000,1.000,1,0.858,1.000,1.000,0,
+single,fail_004,failure_expected,failure_expected,fact_lookup,document,ko,0,1,KGS AC999 임의 가스 코드,,,11691;11647;11693;11692;13665;13661;13664;13666;13670;13773,404.6,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_005,failure_expected,failure_expected,fact_lookup,document,ko,0,1,초전도 안전 관리법 시행규칙,,,4026;35775;5236;3966;3977;5260;3971;4018;3972;3973,569.1,0.000,0.000,0.000,1,0.000,0.000,0.000,0,

tests/search_eval/runs/p2a_qwen06_20260612_0220.csv

@@ -0,0 +1,52 @@
+label,id,category,legacy_category,intent,domain_hint,language,ocr_derived,failure_expected,query,relevant_ids,graded_relevance,returned_ids_top10,latency_ms,recall_at_10,mrr_at_10,ndcg_at_10,top3_hit,graded_ndcg_at_10,graded_recall_at_10_t2,graded_recall_at_10_t3,dedup_count,error
+single,kw_001,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건법 제6장,3856;3868;3879,3856:3;3868:2;3879:2,30411;30423;3879;3868;3856;3851;4041;10573;5163;30406,659.8,1.000,0.333,0.618,0,0.529,1.000,1.000,0,
+single,kw_002,standards,exact_keyword,fact_lookup,document,ko,0,0,중대재해 처벌 등에 관한 법률 제2장 중대산업재해,3917;3921,3917:3;3921:2,3921;3917;10573;5244;3919;3923;30418;5227;3916;3874,517.4,1.000,1.000,1.000,1,0.834,1.000,1.000,0,
+single,kw_003,standards,exact_keyword,fact_lookup,document,ko,0,0,화학물질관리법 유해화학물질 영업자,3981,3981:3,3981;3985;3980;3984;3993;3869;30412;3857;3978;3986,336.1,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_004,standards,exact_keyword,fact_lookup,document,ko,0,0,근로기준법 안전과 보건,4041,4041:3,4041;3852;30407;5227;30421;30409;30406;10573;3851;3858,599.6,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_005,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 보호구,3888,3888:3,10570;3888;5249;3893;3887;3904;3899;3890;3895;3902,629.5,1.000,0.500,0.631,1,0.631,1.000,1.000,0,
+single,nl_001,korean_only,natural_language_ko,semantic_search,document,ko,0,0,기계로 인한 산업재해 관련 법령,3856;3868;3879;3854,3854:1;3856:3;3868:2;3879:2,5244;30422;30410;3878;5249;3855;3897;30421;3868;3863,664.7,0.250,0.111,0.118,0,0.083,0.333,0.000,0,
+single,nl_002,korean_only,natural_language_ko,semantic_search,document,ko,0,0,사업주가 도급을 줄 때 산업재해를 예방하기 위해 해야 할 일,3855;3867;3878,3855:3;3867:2;3878:2,30410;3855;5227;30409;3917;3854;5244;3867;30421;3878,545.5,1.000,0.500,0.580,1,0.599,1.000,1.000,0,
+single,nl_003,korean_only,natural_language_ko,semantic_search,document,ko,0,0,유해화학물질을 다루는 회사가 지켜야 할 안전 의무,3980;3981;3982,3980:2;3981:2;3982:2,3980;3904;3903;3981;3905;35780;3985;10569;11648;3857,477.8,0.667,1.000,0.671,1,0.671,0.667,0.000,0,
+single,nl_004,korean_only,natural_language_ko,semantic_search,document,ko,0,0,중대재해가 발생했을 때 경영책임자가 처벌받는 기준,3916;3917;3920;3921,3916:2;3917:3;3920:2;3921:2,10573;3917;3916;3918;5227;5244;3919;10571;30409;3854,391.5,0.500,0.500,0.441,1,0.506,0.500,1.000,0,
+single,nl_005,korean_only,natural_language_ko,semantic_search,document,ko,0,0,안전보건교육은 누가 받아야 하고 어떤 내용을 다루는가,3853;3865,3853:3;3865:2,3853;3876;5249;5234;30420;30408;30414;10573;11677;3867,708.8,0.500,1.000,0.613,1,0.787,0.500,1.000,0,
+single,news_001,korean_only,news_ko,semantic_search,news,ko,0,0,이란과 미국의 군사 충돌,4303;4304;4307;4316;4322;4323;4327;4335,4303:2;4304:2;4307:2;4316:2;4322:2;4323:2;4327:2;4335:2,37599;27583;35225;27350;34850;32177;15924;15976;28316;16378,601.4,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_002,korean_only,news_ko,semantic_search,news,ko,0,0,호르무즈 해협 봉쇄,4316;4320;4322;4327,4316:3;4320:2;4322:2;4327:2,41652;41650;41634;31894;35899;22055;9022;17123;11945;9105,402.3,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,misc_001,korean_only,other_domain,fact_lookup,document,ko,0,0,강체의 평면 운동학,4063;4065,4063:3;4065:2,4063;4065;4066;4071;4067;4058;5105;4068;4064;5083,578.4,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,misc_002,korean_only,other_domain,semantic_search,document,ko,0,0,질점의 운동역학,4060;4061;4062,4060:2;4061:2;4062:2,4070;4060;4062;4059;4064;4066;4058;4068;5083;4065,509.9,0.667,0.500,0.531,1,0.531,0.667,0.000,0,
+single,news_003,english_only,news_en,semantic_search,news,en,0,0,Trump Iran ultimatum,4258;4260;4262,4258:2;4260:2;4262:2,21155;21441;4775;4202;4776;4679;4519;37733;25592;28001,339.8,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_001,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,기계 안전 가드 설계 원리,3770;3856,3770:3;3856:2,35778;5239;5244;3758;3770;3763;4540;3787;3810;3791,699.5,0.500,0.200,0.237,0,0.305,0.500,1.000,0,
+single,cl_002,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,산업 안전 입문서,3755;3775;3776;3777,3755:2;3775:2;3776:2;3777:2,5230;5249;3755;3774;5229;10573;3851;3863;30406;3802,543.0,0.250,0.333,0.195,1,0.195,0.250,0.000,0,
+single,cl_003,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,전기 안전 위험,3772;3790,3772:2;3790:2,35779;5260;3772;3897;3790;37326;3755;37216;37011;37330,1114.4,1.000,0.333,0.544,1,0.544,1.000,0.000,0,
+single,news_004,mixed,news_fr,semantic_search,news,mixed,0,0,guerre en Iran,4199;4202;4210;4361;4363;4507;4519;4521,4199:2;4202:2;4210:2;4361:2;4363:2;4507:2;4519:2;4521:2,33101;33368;7740;33374;27078;5925;26631;5840;16010;16457,426.3,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_005,mixed,news_crosslingual,semantic_search,news,mixed,0,0,이란 미국 전쟁 글로벌 반응,4202;4258;4262;4536;4303;4304;4316,4202:2;4258:2;4262:2;4303:2;4304:2;4316:2;4536:2,23242;4457;19111;41629;26741;6789;17242;9469;4765;4345,839.7,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_001,failure_expected,failure_expected,semantic_search,document,mixed,0,1,Rust async runtime tokio scheduler 내부 구조,,,23732;36544;25771;35645;37582;17810;28766;40739;40432;40679,465.6,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_002,failure_expected,failure_expected,semantic_search,document,ko,0,1,양자컴퓨터 큐비트 디코히어런스,,,32096;31984;16289;16384;32043;32035;28233;33551;32721;32028,270.7,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_003,failure_expected,failure_expected,semantic_search,news,ko,0,1,재즈 보컬리스트 빌리 홀리데이,,,18567;20022;20470;35249;26594;16232;23336;35498;18286;16694,258.9,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,kw_006,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 작업장 통로,3886;3887,3886:3;3887:2,3898;3886;3887;3900;3881;3889;3912;3894;3897;3888,480.6,1.000,0.500,0.693,1,0.665,1.000,1.000,0,
+single,kw_007,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 폭발 화재 위험물 누출 방지,3896;3766,3766:1;3896:3,3896;3895;13930;3903;3897;13769;11679;11562;11567;10571,513.9,0.500,1.000,0.613,1,0.917,1.000,1.000,0,
+single,kw_008,standards,standards,fact_lookup,document,ko,0,0,고압가스 안전관리법 전문,4025;4026,4025:3;4026:2,11644;11579;4025;4026;35775;11693;4034;13750;13299;13941,493.9,1.000,0.333,0.571,1,0.539,1.000,1.000,0,
+single,kw_009,standards,standards,fact_lookup,document,ko,0,0,KGS FP111 가스설비 배관설비 기준,13305,13305:3,13305;13306;13311;13312;13302;13304;11688;13309;13313;13918,434.6,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_010,standards,standards,fact_lookup,document,ko,0,0,KGS FU551 가스설비 압력조정기 가스계량기,13652,13652:3,13652;11689;13657;13655;13649;13752;13648;13659;13650;13656,307.8,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_011,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 전기로 인한 위험 방지,3897;3772,3772:1;3897:3,3897;3886;3887;13935;3895;3902;3896;3899;3877;3900,583.1,0.500,1.000,0.613,1,0.917,1.000,1.000,0,
+single,en_001,english_only,english_only,semantic_search,document,en,0,0,pressure vessel flange design,5144;5136,5136:2;5144:3,5140;5136;5186;5144;5180;5212;5137;5193;36899;5182,4040.7,1.000,0.500,0.651,0,0.552,1.000,1.000,0,
+single,en_002,english_only,english_only,semantic_search,document,en,0,0,ASME Section VIII introduction,5204;5206,5204:3;5206:2,5204;5212;5208;5210;5206;11634;5207;5183;5205;5182,2996.1,1.000,1.000,0.850,1,0.918,1.000,1.000,0,
+single,en_003,english_only,english_only,semantic_search,document,en,0,0,impact test requirements ASME,5205;5148,5148:1;5205:3,36721;5205;5204;5178;5214;5224;5148;5186;5192;5185,5084.5,1.000,0.500,0.591,1,0.622,1.000,1.000,0,
+single,en_004,english_only,english_only,semantic_search,document,en,0,0,design of vessel supports,5149,5149:3,5149;5186;5141;5136;5137;36901;36869;5139;5178;5145,5563.0,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,en_005,english_only,english_only,semantic_search,document,en,0,0,hydrogen piping ASME code,5178,5178:3,5178;5180;5212;5182;5207;5210;5148;5204;5209;5224,4870.2,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,en_006,english_only,english_only,semantic_search,document,en,0,0,ASME welding qualification requirements,5209;3771,3771:1;5209:3,5180;5178;5224;36721;36877;37097;5205;5204;36878;37051,2374.8,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,en_007,english_only,english_only,semantic_search,document,en,0,0,pressure vessel fabrication and inspection,5208;5145,5145:2;5208:3,5186;5208;38747;38791;36935;36874;5187;5191;5182;5175,3752.4,0.500,0.500,0.387,1,0.497,0.500,1.000,0,
+single,en_008,english_only,english_only,semantic_search,document,en,0,0,Industrial Safety and Health Management ergonomics,3763;3755,3755:1;3763:3,3759;3763;3774;3755;3818;5230;3812;3770;3756;3760,1817.5,1.000,0.500,0.651,1,0.635,1.000,1.000,0,
+single,cl_004,mixed,mixed,semantic_search,document,mixed,0,0,ASME 압력용기 설계 실무,5207;5210;5139;5135,5135:2;5139:2;5207:3;5210:3,5139;5207;5204;5206;5225;5210;5133;5143;5141;5140,795.3,0.750,1.000,0.776,1,0.697,0.750,1.000,0,
+single,cl_005,mixed,mixed,semantic_search,document,mixed,0,0,ASME 용접 코드 해설,5224;5209,5209:2;5224:3,5224;5222;5209;5225;5210;5208;5180;5204;5195;5205,724.0,1.000,1.000,0.920,1,0.956,1.000,1.000,0,
+single,cl_006,mixed,mixed,semantic_search,document,mixed,0,0,pressure vessel Korean industrial safety regulation,4025;3856;5136,3856:2;4025:2;5136:1,38941;3762;4026;36889;36874;38791;5182;39084;5143;5186,1217.1,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_007,mixed,mixed,semantic_search,document,mixed,0,0,KGS 코드 LPG safety standard,11647;11689;11645;4025,4025:1;11645:2;11647:3;11689:2,11647;13674;13675;11688;11645;13310;13773;13759;13668;11652,512.6,0.500,1.000,0.541,1,0.754,0.667,1.000,0,
+single,cl_008,mixed,mixed,semantic_search,document,mixed,0,0,수소 가스 안전 기준 hydrogen safety,5178;5169,5169:2;5178:3,10575;5177;13924;13925;5241;11562;5178;5173;11568;11563,838.4,0.500,0.143,0.204,0,0.262,0.500,1.000,0,
+single,exam_001,exam,exam,fact_lookup,document,ko,0,0,레이놀즈수 정의 공식,11504;11505,11504:3;11505:2,11504;11533;5081;11482;5088;5090;5082;5077;11509;11505,590.3,1.000,1.000,0.790,1,0.885,1.000,1.000,0,
+single,exam_002,exam,exam,fact_lookup,document,ko,0,0,탱크 바닥 구멍 유체 유속 토리첼리,11500;11495;11496,11495:2;11496:2;11500:3,11500;11495;11501;5085;5071;11515;5090;5139;5086;5088,387.2,0.667,1.000,0.765,1,0.856,0.667,1.000,0,
+single,exam_003,exam,exam,fact_lookup,document,ko,0,0,이상기체 음속 마하수,11514;11515,11514:3;11515:2,11479;5083;5084;11476;11514;11516;11515;11518;11517;11472,645.5,1.000,0.200,0.442,0,0.417,1.000,1.000,0,
+single,exam_004,exam,exam,fact_lookup,document,ko,0,0,고압가스 용기 내압시험 영구증가량,11591;11644;11691,11591:3;11644:2;11691:2,11591;11664;5177;13652;13660;13942;11660;13653;13752;11579,338.3,0.333,1.000,0.469,1,0.674,0.333,1.000,0,
+single,exam_005,exam,exam,fact_lookup,document,ko,0,0,도시가스 배관 매설 이격거리,11627;11625;11646,11625:2;11627:3;11646:1,11627;11658;11625;11692;13753;13751;11655;13653;13752;11624,399.8,0.667,1.000,0.704,1,0.905,1.000,1.000,0,
+single,exam_006,exam,exam,fact_lookup,document,ko,0,0,LPG 저장탱크 안전거리 분말소화기,11617;11669;11620,11617:3;11620:1;11669:2,11595;11616;13669;11617;11649;11605;11655;11690;11658;11653,353.2,0.333,0.250,0.202,0,0.321,0.500,1.000,0,
+single,exam_007,exam,exam,fact_lookup,document,ko,0,0,오리피스 차압식 유량계,11712;11711;11503,11503:2;11711:2;11712:3,11712;11711;11500;11503;11713;13930;11701;11502;11715;11717,408.6,1.000,1.000,0.967,1,0.980,1.000,1.000,0,
+single,fail_004,failure_expected,failure_expected,fact_lookup,document,ko,0,1,KGS AC999 임의 가스 코드,,,11647;11668;5177;11583;11692;13665;13661;11680;11644;13666,417.6,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_005,failure_expected,failure_expected,fact_lookup,document,ko,0,1,초전도 안전 관리법 시행규칙,,,35775;3966;3875;5227;4026;4025;30419;3971;4018;4021,897.2,0.000,0.000,0.000,1,0.000,0.000,0.000,0,

tests/search_eval/runs/p2a_qwen4_20260612_0220.csv

@@ -0,0 +1,52 @@
+label,id,category,legacy_category,intent,domain_hint,language,ocr_derived,failure_expected,query,relevant_ids,graded_relevance,returned_ids_top10,latency_ms,recall_at_10,mrr_at_10,ndcg_at_10,top3_hit,graded_ndcg_at_10,graded_recall_at_10_t2,graded_recall_at_10_t3,dedup_count,error
+single,kw_001,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건법 제6장,3856;3868;3879,3856:3;3868:2;3879:2,30417;30418;30406;10573;3863;3871;3862;3874;3851;3859,437.7,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,kw_002,standards,exact_keyword,fact_lookup,document,ko,0,0,중대재해 처벌 등에 관한 법률 제2장 중대산업재해,3917;3921,3917:3;3921:2,10573;3923;30418;3916;3921;3874;3917;3918;3920;3854,449.2,1.000,0.200,0.442,0,0.393,1.000,1.000,0,
+single,kw_003,standards,exact_keyword,fact_lookup,document,ko,0,0,화학물질관리법 유해화학물질 영업자,3981,3981:3,3981;3985;3980;3978;3984;3869;3983;30424;3982;3903,320.2,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_004,standards,exact_keyword,fact_lookup,document,ko,0,0,근로기준법 안전과 보건,4041,4041:3,4041;30406;3851;10573;30419;30407;4036;3915;3852;3875,530.2,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_005,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 보호구,3888,3888:3,3905;3888;3885;3910;3895;3911;3890;3906;3913;3889,605.2,1.000,0.500,0.631,1,0.631,1.000,1.000,0,
+single,nl_001,korean_only,natural_language_ko,semantic_search,document,ko,0,0,기계로 인한 산업재해 관련 법령,3856;3868;3879;3854,3854:1;3856:3;3868:2;3879:2,3856;30411;30423;3879;5249;5244;3868;30421;30406;3851,644.8,0.750,1.000,0.689,1,0.859,1.000,1.000,0,
+single,nl_002,korean_only,natural_language_ko,semantic_search,document,ko,0,0,사업주가 도급을 줄 때 산업재해를 예방하기 위해 해야 할 일,3855;3867;3878,3855:3;3867:2;3878:2,3855;3867;30410;5227;3917;3878;30418;30422;3874;30421,538.4,1.000,1.000,0.933,1,0.958,1.000,1.000,0,
+single,nl_003,korean_only,natural_language_ko,semantic_search,document,ko,0,0,유해화학물질을 다루는 회사가 지켜야 할 안전 의무,3980;3981;3982,3980:2;3981:2;3982:2,3981;3985;3903;3980;30424;3867;3904;3905;3880;3760,474.4,0.667,1.000,0.671,1,0.671,0.667,0.000,0,
+single,nl_004,korean_only,natural_language_ko,semantic_search,document,ko,0,0,중대재해가 발생했을 때 경영책임자가 처벌받는 기준,3916;3917;3920;3921,3916:2;3917:3;3920:2;3921:2,3917;10573;3918;3916;10572;3872;10571;3923;30418;30417,403.9,0.500,1.000,0.559,1,0.710,0.500,1.000,0,
+single,nl_005,korean_only,natural_language_ko,semantic_search,document,ko,0,0,안전보건교육은 누가 받아야 하고 어떤 내용을 다루는가,3853;3865,3853:3;3865:2,5249;10573;30420;30408;3876;3853;3867;5234;30424;4842,711.7,0.500,0.167,0.218,0,0.280,0.500,1.000,0,
+single,news_001,korean_only,news_ko,semantic_search,news,ko,0,0,이란과 미국의 군사 충돌,4303;4304;4307;4316;4322;4323;4327;4335,4303:2;4304:2;4307:2;4316:2;4322:2;4323:2;4327:2;4335:2,27583;35225;34298;30806;30679;27350;27974;32155;26534;26734,566.4,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_002,korean_only,news_ko,semantic_search,news,ko,0,0,호르무즈 해협 봉쇄,4316;4320;4322;4327,4316:3;4320:2;4322:2;4327:2,35904;35247;22055;16532;35893;26204;35899;16392;17123;4327,333.3,0.250,0.100,0.113,0,0.074,0.250,0.000,0,
+single,misc_001,korean_only,other_domain,fact_lookup,document,ko,0,0,강체의 평면 운동학,4063;4065,4063:3;4065:2,4063;4071;4065;4064;4066;4058;5083;4067;5105;4068,468.5,1.000,1.000,0.920,1,0.956,1.000,1.000,0,
+single,misc_002,korean_only,other_domain,semantic_search,document,ko,0,0,질점의 운동역학,4060;4061;4062,4060:2;4061:2;4062:2,4059;4060;4070;4062;4058;4061;5113;5083;4064;4066,382.4,1.000,0.500,0.665,1,0.665,1.000,0.000,0,
+single,news_003,english_only,news_en,semantic_search,news,en,0,0,Trump Iran ultimatum,4258;4260;4262,4258:2;4260:2;4262:2,4775;37733;4776;4658;4679;4519;4202;4668;4333;25974,321.8,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_001,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,기계 안전 가드 설계 원리,3770;3856,3770:3;3856:2,35778;5239;5244;3758;3770;3763;3774;4540;5253;3787,776.2,0.500,0.200,0.237,0,0.305,0.500,1.000,0,
+single,cl_002,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,산업 안전 입문서,3755;3775;3776;3777,3755:2;3775:2;3776:2;3777:2,5230;5249;3756;3882;3774;37259;3863;3755;10573;3802,599.2,0.250,0.125,0.123,1,0.123,0.250,0.000,0,
+single,cl_003,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,전기 안전 위험,3772;3790,3772:2;3790:2,35779;3790;3772;5260;3897;37326;3755;37216;37011;3774,1205.5,1.000,0.500,0.693,1,0.693,1.000,0.000,0,
+single,news_004,mixed,news_fr,semantic_search,news,mixed,0,0,guerre en Iran,4199;4202;4210;4361;4363;4507;4519;4521,4199:2;4202:2;4210:2;4361:2;4363:2;4507:2;4519:2;4521:2,33101;33368;7740;33374;27078;5925;26631;5840;16010;16457,494.7,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_005,mixed,news_crosslingual,semantic_search,news,mixed,0,0,이란 미국 전쟁 글로벌 반응,4202;4258;4262;4536;4303;4304;4316,4202:2;4258:2;4262:2;4303:2;4304:2;4316:2;4536:2,23242;15924;4457;36648;36490;19111;35556;41629;26631;26741,875.2,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_001,failure_expected,failure_expected,semantic_search,document,mixed,0,1,Rust async runtime tokio scheduler 내부 구조,,,23732;36544;35645;37582;17810;28766;40739;40432;40679;40700,502.6,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_002,failure_expected,failure_expected,semantic_search,document,ko,0,1,양자컴퓨터 큐비트 디코히어런스,,,32096;31984;33521;16289;16384;32043;32035;22202;28233;33551,342.0,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_003,failure_expected,failure_expected,semantic_search,news,ko,0,1,재즈 보컬리스트 빌리 홀리데이,,,18567;31643;20022;20470;16059;27146;26559;26637;18286;31187,289.1,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,kw_006,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 작업장 통로,3886;3887,3886:3;3887:2,5236;3898;3886;3902;3895;3887;5244;3889;3912;3894,537.9,1.000,0.333,0.525,1,0.514,1.000,1.000,0,
+single,kw_007,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 폭발 화재 위험물 누출 방지,3896;3766,3766:1;3896:3,3896;3895;3903;13930;3897;13769;3766;13935;10571;3867,583.1,1.000,1.000,0.818,1,0.961,1.000,1.000,0,
+single,kw_008,standards,standards,fact_lookup,document,ko,0,0,고압가스 안전관리법 전문,4025;4026,4025:3;4026:2,11644;11579;4026;4025;11693;35775;11645;4034;13299;13941,539.9,1.000,0.333,0.571,0,0.508,1.000,1.000,0,
+single,kw_009,standards,standards,fact_lookup,document,ko,0,0,KGS FP111 가스설비 배관설비 기준,13305,13305:3,13305;13311;13306;13312;13302;13304;11688;13309;13313;13918,488.1,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_010,standards,standards,fact_lookup,document,ko,0,0,KGS FU551 가스설비 압력조정기 가스계량기,13652,13652:3,13652;11689;13657;13655;13649;13656;13648;13659;13752;13650,362.0,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_011,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 전기로 인한 위험 방지,3897;3772,3772:1;3897:3,3897;3758;3886;3755;5260;3772;3887;13935;3895;3774,716.8,1.000,1.000,0.832,1,0.964,1.000,1.000,0,
+single,en_001,english_only,english_only,semantic_search,document,en,0,0,pressure vessel flange design,5144;5136,5136:2;5144:3,5136;5178;5140;5180;5144;5193;36899;5182;5207;5206,4109.2,1.000,1.000,0.850,0,0.642,1.000,1.000,0,
+single,en_002,english_only,english_only,semantic_search,document,en,0,0,ASME Section VIII introduction,5204;5206,5204:3;5206:2,5204;5212;5208;5206;5207;5183;5210;5205;5182;5148,3037.9,1.000,1.000,0.877,1,0.932,1.000,1.000,0,
+single,en_003,english_only,english_only,semantic_search,document,en,0,0,impact test requirements ASME,5205;5148,5148:1;5205:3,5205;5204;36721;5225;5178;5224;37029;5214;5210;5190,5183.0,0.500,1.000,0.613,1,0.917,1.000,1.000,0,
+single,en_004,english_only,english_only,semantic_search,document,en,0,0,design of vessel supports,5149,5149:3,5149;5186;5141;5136;5140;5137;36901;36869;5139;5178,5657.1,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,en_005,english_only,english_only,semantic_search,document,en,0,0,hydrogen piping ASME code,5178,5178:3,5143;5178;5180;5212;5182;5133;5207;5210;5148;5204,4902.9,1.000,0.500,0.631,1,0.631,1.000,1.000,0,
+single,en_006,english_only,english_only,semantic_search,document,en,0,0,ASME welding qualification requirements,5209;3771,3771:1;5209:3,5204;5178;5224;36721;37097;5205;37051;36963;36951;5208,2455.9,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,en_007,english_only,english_only,semantic_search,document,en,0,0,pressure vessel fabrication and inspection,5208;5145,5145:2;5208:3,5208;5186;38747;38791;36935;5180;5182;36874;5187;36889,3808.4,0.500,1.000,0.613,1,0.787,0.500,1.000,0,
+single,en_008,english_only,english_only,semantic_search,document,en,0,0,Industrial Safety and Health Management ergonomics,3763;3755,3755:1;3763:3,3759;3763;3818;3774;3812;3760;3756;3778;3757;3775,1870.9,0.500,0.500,0.387,1,0.579,1.000,1.000,0,
+single,cl_004,mixed,mixed,semantic_search,document,mixed,0,0,ASME 압력용기 설계 실무,5207;5210;5139;5135,5135:2;5139:2;5207:3;5210:3,5207;5204;5139;5225;5210;5182;5140;5148;5212;5144,832.1,0.750,1.000,0.737,1,0.789,0.750,1.000,0,
+single,cl_005,mixed,mixed,semantic_search,document,mixed,0,0,ASME 용접 코드 해설,5224;5209,5209:2;5224:3,5224;5222;5204;5225;5210;5205;36877;5209;37051;36721,737.6,1.000,1.000,0.807,1,0.894,1.000,1.000,0,
+single,cl_006,mixed,mixed,semantic_search,document,mixed,0,0,pressure vessel Korean industrial safety regulation,4025;3856;5136,3856:2;4025:2;5136:1,38941;3762;11644;4026;5186;38791;3895;39084;13299;5185,1267.9,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_007,mixed,mixed,semantic_search,document,mixed,0,0,KGS 코드 LPG safety standard,11647;11689;11645;4025,4025:1;11645:2;11647:3;11689:2,11647;13674;13669;13774;13675;11688;13667;13670;11645;13310,558.8,0.500,1.000,0.508,1,0.730,0.667,1.000,0,
+single,cl_008,mixed,mixed,semantic_search,document,mixed,0,0,수소 가스 안전 기준 hydrogen safety,5178;5169,5169:2;5178:3,5177;10572;11671;5241;11688;11653;3774;5178;5173;38701,879.9,0.500,0.125,0.193,0,0.248,0.500,1.000,0,
+single,exam_001,exam,exam,fact_lookup,document,ko,0,0,레이놀즈수 정의 공식,11504;11505,11504:3;11505:2,11504;11533;5081;5087;11482;5088;5090;4544;11509;11505,629.4,1.000,1.000,0.790,1,0.885,1.000,1.000,0,
+single,exam_002,exam,exam,fact_lookup,document,ko,0,0,탱크 바닥 구멍 유체 유속 토리첼리,11500;11495;11496,11495:2;11496:2;11500:3,11500;11495;11501;5085;5090;3788;5086;5071;11515;11503,405.7,0.667,1.000,0.765,1,0.856,0.667,1.000,0,
+single,exam_003,exam,exam,fact_lookup,document,ko,0,0,이상기체 음속 마하수,11514;11515,11514:3;11515:2,11479;5083;11514;5084;11476;11527;11516;11515;5124;11472,691.4,1.000,0.333,0.500,1,0.500,1.000,1.000,0,
+single,exam_004,exam,exam,fact_lookup,document,ko,0,0,고압가스 용기 내압시험 영구증가량,11591;11644;11691,11591:3;11644:2;11691:2,11591;13660;11668;5177;13652;13943;13942;13917;11592;13752,400.9,0.333,1.000,0.469,1,0.674,0.333,1.000,0,
+single,exam_005,exam,exam,fact_lookup,document,ko,0,0,도시가스 배관 매설 이격거리,11627;11625;11646,11625:2;11627:3;11646:1,11627;11658;11600;11625;11692;13918;13751;5177;11655;13653,435.4,0.667,1.000,0.671,1,0.883,1.000,1.000,0,
+single,exam_006,exam,exam,fact_lookup,document,ko,0,0,LPG 저장탱크 안전거리 분말소화기,11617;11669;11620,11617:3;11620:1;11669:2,11595;11616;11617;11649;13304;11655;11690;11658;13669;13915,417.2,0.333,0.333,0.235,1,0.373,0.500,1.000,0,
+single,exam_007,exam,exam,fact_lookup,document,ko,0,0,오리피스 차압식 유량계,11712;11711;11503,11503:2;11711:2;11712:3,11712;11711;11503;11500;11713;13930;11502;11716;11612;11715,436.7,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,fail_004,failure_expected,failure_expected,fact_lookup,document,ko,0,1,KGS AC999 임의 가스 코드,,,11647;11668;5177;11688;11583;11692;13665;13661;11680;13666,447.7,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_005,failure_expected,failure_expected,fact_lookup,document,ko,0,1,초전도 안전 관리법 시행규칙,,,35775;11644;3966;4026;3977;4022;3972;13299;3974;3971,586.4,0.000,0.000,0.000,1,0.000,0.000,0.000,0,

tests/search_eval/runs/p2a_qwen4m_20260612_0220.csv

@@ -0,0 +1,52 @@
+label,id,category,legacy_category,intent,domain_hint,language,ocr_derived,failure_expected,query,relevant_ids,graded_relevance,returned_ids_top10,latency_ms,recall_at_10,mrr_at_10,ndcg_at_10,top3_hit,graded_ndcg_at_10,graded_recall_at_10_t2,graded_recall_at_10_t3,dedup_count,error
+single,kw_001,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건법 제6장,3856;3868;3879,3856:3;3868:2;3879:2,30411;3879;3868;3856;3851;10573;5249;3867;3873;3853,697.4,1.000,0.500,0.733,0,0.617,1.000,1.000,0,
+single,kw_002,standards,exact_keyword,fact_lookup,document,ko,0,0,중대재해 처벌 등에 관한 법률 제2장 중대산업재해,3917;3921,3917:3;3921:2,3921;3917;10573;3919;3923;30418;5227;10571;3916;3874,513.6,1.000,1.000,1.000,1,0.834,1.000,1.000,0,
+single,kw_003,standards,exact_keyword,fact_lookup,document,ko,0,0,화학물질관리법 유해화학물질 영업자,3981,3981:3,3981;3985;3984;3983;3904;3980;3903;3905;3978;3869,319.1,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_004,standards,exact_keyword,fact_lookup,document,ko,0,0,근로기준법 안전과 보건,4041,4041:3,4041;3852;30407;3851;30406;3915;10573;3858;5249;3863,603.5,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_005,standards,exact_keyword,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 보호구,3888,3888:3,10570;3888;3912;3913;3908;3911;3905;5249;3910;3893,642.7,1.000,0.500,0.631,1,0.631,1.000,1.000,0,
+single,nl_001,korean_only,natural_language_ko,semantic_search,document,ko,0,0,기계로 인한 산업재해 관련 법령,3856;3868;3879;3854,3854:1;3856:3;3868:2;3879:2,5244;5227;5249;3867;3878;3863;30411;3868;3856;3879,664.3,0.750,0.125,0.354,0,0.362,1.000,1.000,0,
+single,nl_002,korean_only,natural_language_ko,semantic_search,document,ko,0,0,사업주가 도급을 줄 때 산업재해를 예방하기 위해 해야 할 일,3855;3867;3878,3855:3;3867:2;3878:2,30410;3855;30409;3917;5244;30406;3851;3867;5249;30421,547.9,0.667,0.500,0.444,1,0.516,0.667,1.000,0,
+single,nl_003,korean_only,natural_language_ko,semantic_search,document,ko,0,0,유해화학물질을 다루는 회사가 지켜야 할 안전 의무,3980;3981;3982,3980:2;3981:2;3982:2,3903;3982;3760;3905;3981;3904;3980;5249;3985;10569,466.6,1.000,0.500,0.634,1,0.634,1.000,0.000,0,
+single,nl_004,korean_only,natural_language_ko,semantic_search,document,ko,0,0,중대재해가 발생했을 때 경영책임자가 처벌받는 기준,3916;3917;3920;3921,3916:2;3917:3;3920:2;3921:2,10572;10573;3917;3918;3916;5227;3919;10571;30409;3854,392.0,0.500,0.333,0.346,1,0.399,0.500,1.000,0,
+single,nl_005,korean_only,natural_language_ko,semantic_search,document,ko,0,0,안전보건교육은 누가 받아야 하고 어떤 내용을 다루는가,3853;3865,3853:3;3865:2,3853;5249;5234;4025;30420;30408;10573;11677;3876;3867,726.2,0.500,1.000,0.613,1,0.787,0.500,1.000,0,
+single,news_001,korean_only,news_ko,semantic_search,news,ko,0,0,이란과 미국의 군사 충돌,4303;4304;4307;4316;4322;4323;4327;4335,4303:2;4304:2;4307:2;4316:2;4322:2;4323:2;4327:2;4335:2,37599;27583;35225;27350;34850;30518;15924;32410;16081;28407,593.2,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_002,korean_only,news_ko,semantic_search,news,ko,0,0,호르무즈 해협 봉쇄,4316;4320;4322;4327,4316:3;4320:2;4322:2;4327:2,41652;41650;41634;31894;35899;22055;16392;9022;17123;11945,399.4,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,misc_001,korean_only,other_domain,fact_lookup,document,ko,0,0,강체의 평면 운동학,4063;4065,4063:3;4065:2,4063;4065;4066;4071;4064;4068;4067;4058;5083;4059,592.2,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,misc_002,korean_only,other_domain,semantic_search,document,ko,0,0,질점의 운동역학,4060;4061;4062,4060:2;4061:2;4062:2,4070;4061;4060;4062;4059;4065;4066;4058;4068;4071,512.9,1.000,0.500,0.733,1,0.733,1.000,0.000,0,
+single,news_003,english_only,news_en,semantic_search,news,en,0,0,Trump Iran ultimatum,4258;4260;4262,4258:2;4260:2;4262:2,21155;4775;4202;4776;4679;4199;4519;4668;37733;4658,342.4,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_001,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,기계 안전 가드 설계 원리,3770;3856,3770:3;3856:2,35778;5239;5244;3758;3791;3770;3763;3774;4540;5253,683.2,0.500,0.167,0.218,0,0.280,0.500,1.000,0,
+single,cl_002,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,산업 안전 입문서,3755;3775;3776;3777,3755:2;3775:2;3776:2;3777:2,5230;5249;3756;3761;3757;3774;3863;3755;10573;3802,565.0,0.250,0.125,0.123,1,0.123,0.250,0.000,0,
+single,cl_003,mixed,crosslingual_ko_en,semantic_search,document,mixed,0,0,전기 안전 위험,3772;3790,3772:2;3790:2,35779;3790;5260;3772;3897;37326;3755;37216;37011;3774,1136.0,1.000,0.500,0.651,1,0.651,1.000,0.000,0,
+single,news_004,mixed,news_fr,semantic_search,news,mixed,0,0,guerre en Iran,4199;4202;4210;4361;4363;4507;4519;4521,4199:2;4202:2;4210:2;4361:2;4363:2;4507:2;4519:2;4521:2,33101;33368;7740;33374;27078;5925;5840;16010;16457;6945,423.2,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,news_005,mixed,news_crosslingual,semantic_search,news,mixed,0,0,이란 미국 전쟁 글로벌 반응,4202;4258;4262;4536;4303;4304;4316,4202:2;4258:2;4262:2;4303:2;4304:2;4316:2;4536:2,23242;4688;15924;4457;36648;36490;19111;35556;41629;26631,823.5,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_001,failure_expected,failure_expected,semantic_search,document,mixed,0,1,Rust async runtime tokio scheduler 내부 구조,,,36544;35645;37582;17810;28766;40739;40432;40679;40700;40734,461.6,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_002,failure_expected,failure_expected,semantic_search,document,ko,0,1,양자컴퓨터 큐비트 디코히어런스,,,32096;31984;33521;16289;16384;32043;32035;22202;28233;33551,253.5,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_003,failure_expected,failure_expected,semantic_search,news,ko,0,1,재즈 보컬리스트 빌리 홀리데이,,,18567;20022;20470;27146;26559;26637;18286;31187;26759;38352,245.5,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,kw_006,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준에 관한 규칙 작업장 통로,3886;3887,3886:3;3887:2,5236;3898;3886;3902;3895;3887;5244;3889;3912;3894,498.7,1.000,0.333,0.525,1,0.514,1.000,1.000,0,
+single,kw_007,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 폭발 화재 위험물 누출 방지,3896;3766,3766:1;3896:3,3896;3895;3903;13930;3897;13769;3766;13935;10571;3867,532.8,1.000,1.000,0.818,1,0.961,1.000,1.000,0,
+single,kw_008,standards,standards,fact_lookup,document,ko,0,0,고압가스 안전관리법 전문,4025;4026,4025:3;4026:2,11644;11579;4026;4025;11693;35775;11645;4034;13299;13941,505.4,1.000,0.333,0.571,0,0.508,1.000,1.000,0,
+single,kw_009,standards,standards,fact_lookup,document,ko,0,0,KGS FP111 가스설비 배관설비 기준,13305,13305:3,13305;13306;13311;13312;13302;13304;11688;13309;13313;13918,439.9,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_010,standards,standards,fact_lookup,document,ko,0,0,KGS FU551 가스설비 압력조정기 가스계량기,13652,13652:3,13652;11689;13657;13655;13649;13656;13651;13648;13752;13659,324.4,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,kw_011,standards,standards,fact_lookup,document,ko,0,0,산업안전보건기준 전기로 인한 위험 방지,3897;3772,3772:1;3897:3,3897;3758;3886;3755;5260;3772;3887;13935;3895;3774,873.6,1.000,1.000,0.832,1,0.964,1.000,1.000,0,
+single,en_001,english_only,english_only,semantic_search,document,en,0,0,pressure vessel flange design,5144;5136,5136:2;5144:3,5136;5178;5140;5180;5144;36899;5182;5207;5206;5149,4311.0,1.000,1.000,0.850,0,0.642,1.000,1.000,0,
+single,en_002,english_only,english_only,semantic_search,document,en,0,0,ASME Section VIII introduction,5204;5206,5204:3;5206:2,5204;5212;5208;5210;5206;5141;5207;5205;5182;5148,11067.7,1.000,1.000,0.850,1,0.918,1.000,1.000,0,
+single,en_003,english_only,english_only,semantic_search,document,en,0,0,impact test requirements ASME,5205;5148,5148:1;5205:3,5205;5204;36721;5225;5178;5224;37029;5214;5210;5190,5144.1,0.500,1.000,0.613,1,0.917,1.000,1.000,0,
+single,en_004,english_only,english_only,semantic_search,document,en,0,0,design of vessel supports,5149,5149:3,5149;5186;5141;5140;5136;5137;36901;36869;5139;5210,5560.7,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,en_005,english_only,english_only,semantic_search,document,en,0,0,hydrogen piping ASME code,5178,5178:3,5143;5178;5180;5212;5182;5207;5210;5148;5204;5209,4909.7,1.000,0.500,0.631,1,0.631,1.000,1.000,0,
+single,en_006,english_only,english_only,semantic_search,document,en,0,0,ASME welding qualification requirements,5209;3771,3771:1;5209:3,5180;5204;5224;36721;37097;5205;37051;36963;5178;36951,2380.2,0.000,0.000,0.000,0,0.000,0.000,0.000,0,
+single,en_007,english_only,english_only,semantic_search,document,en,0,0,pressure vessel fabrication and inspection,5208;5145,5145:2;5208:3,5186;5208;38747;38791;36935;5180;5182;36874;5187;36897,3795.7,0.500,0.500,0.387,1,0.497,0.500,1.000,0,
+single,en_008,english_only,english_only,semantic_search,document,en,0,0,Industrial Safety and Health Management ergonomics,3763;3755,3755:1;3763:3,3759;3763;3818;3774;3812;3757;3756;3775;3762;37280,1851.3,0.500,0.500,0.387,1,0.579,1.000,1.000,0,
+single,cl_004,mixed,mixed,semantic_search,document,mixed,0,0,ASME 압력용기 설계 실무,5207;5210;5139;5135,5135:2;5139:2;5207:3;5210:3,5207;5204;5139;5225;5210;5182;5140;5212;5144;5224,783.3,0.750,1.000,0.737,1,0.789,0.750,1.000,0,
+single,cl_005,mixed,mixed,semantic_search,document,mixed,0,0,ASME 용접 코드 해설,5224;5209,5209:2;5224:3,5224;5222;5204;5225;5210;5182;5195;5205;36877;5209,704.0,1.000,1.000,0.790,1,0.885,1.000,1.000,0,
+single,cl_006,mixed,mixed,semantic_search,document,mixed,0,0,pressure vessel Korean industrial safety regulation,4025;3856;5136,3856:2;4025:2;5136:1,38941;3762;11644;4026;5186;38791;3895;39084;5204;13299,1227.3,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,cl_007,mixed,mixed,semantic_search,document,mixed,0,0,KGS 코드 LPG safety standard,11647;11689;11645;4025,4025:1;11645:2;11647:3;11689:2,11647;13674;13669;13774;13675;11688;13769;13667;13670;11645,525.2,0.500,1.000,0.503,1,0.727,0.667,1.000,0,
+single,cl_008,mixed,mixed,semantic_search,document,mixed,0,0,수소 가스 안전 기준 hydrogen safety,5178;5169,5169:2;5178:3,10575;5177;10572;5241;11688;11653;5178;11693;5173;11563,838.0,0.500,0.143,0.204,0,0.262,0.500,1.000,0,
+single,exam_001,exam,exam,fact_lookup,document,ko,0,0,레이놀즈수 정의 공식,11504;11505,11504:3;11505:2,11504;11533;5081;5087;11482;5088;5090;5082;11509;11505,576.2,1.000,1.000,0.790,1,0.885,1.000,1.000,0,
+single,exam_002,exam,exam,fact_lookup,document,ko,0,0,탱크 바닥 구멍 유체 유속 토리첼리,11500;11495;11496,11495:2;11496:2;11500:3,11500;11495;11501;5090;5085;3788;5086;5071;5075;11515,377.2,0.667,1.000,0.765,1,0.856,0.667,1.000,0,
+single,exam_003,exam,exam,fact_lookup,document,ko,0,0,이상기체 음속 마하수,11514;11515,11514:3;11515:2,11479;5083;11514;5084;11527;11516;11515;5124;11518;11471,650.0,1.000,0.333,0.511,1,0.506,1.000,1.000,0,
+single,exam_004,exam,exam,fact_lookup,document,ko,0,0,고압가스 용기 내압시험 영구증가량,11591;11644;11691,11591:3;11644:2;11691:2,11591;11664;11668;5177;13943;13942;13917;11592;13752;4026,354.7,0.333,1.000,0.469,1,0.674,0.333,1.000,0,
+single,exam_005,exam,exam,fact_lookup,document,ko,0,0,도시가스 배관 매설 이격거리,11627;11625;11646,11625:2;11627:3;11646:1,11627;11658;11600;11625;11692;13918;13751;5177;11655;13653,393.6,0.667,1.000,0.671,1,0.883,1.000,1.000,0,
+single,exam_006,exam,exam,fact_lookup,document,ko,0,0,LPG 저장탱크 안전거리 분말소화기,11617;11669;11620,11617:3;11620:1;11669:2,11595;11616;11617;11649;13304;11655;11690;11658;13669;13915,393.7,0.333,0.333,0.235,1,0.373,0.500,1.000,0,
+single,exam_007,exam,exam,fact_lookup,document,ko,0,0,오리피스 차압식 유량계,11712;11711;11503,11503:2;11711:2;11712:3,11712;11711;11503;11500;11713;11502;11716;11612;11715;11717,397.8,1.000,1.000,1.000,1,1.000,1.000,1.000,0,
+single,fail_004,failure_expected,failure_expected,fact_lookup,document,ko,0,1,KGS AC999 임의 가스 코드,,,11647;11668;5177;11688;11583;11692;13665;13661;13664;11680,405.8,0.000,0.000,0.000,1,0.000,0.000,0.000,0,
+single,fail_005,failure_expected,failure_expected,fact_lookup,document,ko,0,1,초전도 안전 관리법 시행규칙,,,35775;11644;3966;4026;3977;3972;13299;3974;3971;4025,548.8,0.000,0.000,0.000,1,0.000,0.000,0.000,0,

tests/test_b4_license_watch.py

@@ -0,0 +1,80 @@
+"""B-4 — licensed_restricted 차단 술어 + watch 타깃 (material/jurisdiction/license) 매핑 순수 테스트.
+
+차단 술어(_license_sql)는 retrieval 3-leg + digest 가 공유하는 단일 술어. 실제 제외 동작은
+GPU 라이브(합성 restricted doc 검색 제외)로 검증 — 여기선 술어 형태 + 매핑 표 계약만.
+[[feedback_external_api_fixture_first]] / [[feedback_structural_integrity_over_path_discipline]]
+"""
+
+import sys
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).parent.parent / "app"))
+
+from services.search.license_filter import (  # noqa: E402
+    restricted_exclude_orm,
+    restricted_exclude_sql,
+)
+from services.search.retrieval_service import _license_sql  # noqa: E402
+from workers.file_watcher import _TARGET_AXIS  # noqa: E402
+
+
+def test_shared_predicate_single_source():
+    # retrieval/digest/briefing 가 같은 술어 정의를 공유 — drift 방지(단일 source 계약)
+    assert _license_sql("d") == " AND " + restricted_exclude_sql("d")
+    assert _license_sql("") == " AND " + restricted_exclude_sql("")
+    assert restricted_exclude_sql("d").startswith("COALESCE(d.extract_meta")
+
+
+def test_restricted_exclude_orm_constructs():
+    # study 풀이(explanation_rag)용 ORM 표현 — 컴파일 SQL 이 raw 술어와 동일 구조인지
+    from sqlalchemy.dialects import postgresql
+
+    clause = restricted_exclude_orm()
+    sql = str(clause.compile(dialect=postgresql.dialect(),
+                             compile_kwargs={"literal_binds": True}))
+    assert "extract_meta" in sql
+    assert "'license'" in sql and "'restricted'" in sql  # JSONB 경로 키
+    assert "'false'" in sql and "'true'" in sql           # COALESCE 기본 + 비교값
+
+
+def test_license_sql_shape_with_alias():
+    sql = _license_sql("d")
+    assert sql.startswith(" AND ")                       # 항상 ' AND ...' (WHERE 합성용)
+    assert "COALESCE(d.extract_meta -> 'license' ->> 'restricted', 'false')" in sql
+    assert "<> 'true'" in sql                            # restricted=true 만 제외
+
+
+def test_license_sql_shape_no_alias():
+    # alias='' = 단일 FROM documents (컬럼 직접 참조)
+    sql = _license_sql("")
+    assert "COALESCE(extract_meta -> 'license' ->> 'restricted', 'false')" in sql
+    assert ".extract_meta" not in sql                    # 점 없는 컬럼 직접
+
+
+def test_axis_books_papers_are_restricted():
+    for folder, mt in (("Books", "book"), ("Papers_Purchased", "paper")):
+        material, jur, lic = _TARGET_AXIS[folder]
+        assert material == mt
+        assert jur is None                               # 책/논문 = 관할 없음(A-2 paper NULL 강제와 정합)
+        assert lic["scheme"] == "proprietary"
+        assert lic["restricted"] is True                 # RAG/digest 차단 대상
+        assert lic["redistribute"] is False
+
+
+def test_axis_manuals_proprietary_but_not_restricted():
+    material, jur, lic = _TARGET_AXIS["Manuals"]
+    assert material == "manual"
+    assert lic["scheme"] == "proprietary"
+    assert lic["restricted"] is False                    # 사용자 결정 2026-06-13 (검색·RAG 활용)
+
+
+def test_axis_kgs_law_kr_public_not_restricted():
+    material, jur, lic = _TARGET_AXIS["KGS_Code"]
+    assert (material, jur) == ("law", "KR")
+    assert lic["scheme"] == "kogl"
+    assert lic["restricted"] is False                    # 법정 위임 공공 → 차단 아님
+
+
+def test_axis_non_target_folder_yields_none():
+    # Inbox/Recordings 등 비대상 = (None, None, None) → material/license 미주입
+    assert _TARGET_AXIS.get("Inbox", (None, None, None)) == (None, None, None)

tests/test_c1_decorate.py

@@ -0,0 +1,57 @@
+"""C-1 후속 — facets 집계 + version_status decorate 순수 로직 테스트.
+
+version_status 의 실제 legal_meta 조회는 GPU 라이브(법령 검색)로 검증 — 여기선 facets 분포
+계약 + decorate 의 law 무결과 skip 경로(DB 미접촉)만.
+"""
+
+import asyncio
+import sys
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).parent.parent / "app"))
+
+from services.search.result_decorate import (  # noqa: E402
+    compute_facets,
+    decorate_version_status,
+)
+
+
+class _R:
+    """SearchResult 흉내 — 분류 축 속성만."""
+
+    def __init__(self, material_type=None, jurisdiction=None, version_status=None, id=1):
+        self.material_type = material_type
+        self.jurisdiction = jurisdiction
+        self.version_status = version_status
+        self.id = id
+
+
+def test_compute_facets_distribution():
+    results = [
+        _R("law", "KR", "current"),
+        _R("law", "KR", "superseded"),
+        _R("incident", "KR", None),
+        _R("paper", None, None),
+    ]
+    f = compute_facets(results)
+    assert f["material_type"] == {"law": 2, "incident": 1, "paper": 1}
+    assert f["jurisdiction"] == {"KR": 3}                  # paper jurisdiction None 제외
+    assert f["version_status"] == {"current": 1, "superseded": 1}  # None 제외
+
+
+def test_compute_facets_empty_and_all_none():
+    assert compute_facets([]) == {}
+    assert compute_facets([_R(), _R()]) == {}              # 모든 축 None → 빈 축 미포함
+
+
+def test_compute_facets_excludes_empty_axes():
+    f = compute_facets([_R(jurisdiction="US"), _R(jurisdiction="EU")])
+    assert f == {"jurisdiction": {"US": 1, "EU": 1}}
+    assert "material_type" not in f
+
+
+def test_decorate_version_status_skips_without_law():
+    # law 결과 없으면 legal_meta 조회 skip → session 미사용(None 으로도 무오류)
+    results = [_R("incident", "KR"), _R("paper")]
+    asyncio.run(decorate_version_status(None, results))
+    assert all(r.version_status is None for r in results)

tests/test_kosha_fatal.py

@@ -0,0 +1,67 @@
+"""B-2 KOSHA 사망사고 속보(callApiId=1040) — 순수 파서 fixture 테스트 (plan safety-library-1).
+
+fixture = 2026-06-13 data.go.kr 라이브 박제 (serviceKey 응답 본문 미포함 확인,
+tests/fixtures/kosha_fatal_response.json). _fatal_fields/_items 는 순수 함수라 DB/httpx
+호출 없이 검증 — [[feedback_external_api_fixture_first]].
+"""
+
+import json
+from datetime import date
+from pathlib import Path
+
+from workers.kosha_collector import _fatal_fields, _items
+
+FIXTURE = Path(__file__).parent / "fixtures" / "kosha_fatal_response.json"
+
+
+def _payload() -> dict:
+    return json.loads(FIXTURE.read_text(encoding="utf-8"))
+
+
+def test_items_envelope_parse():
+    """body.items.item 봉투 파싱 — 재해사례와 동일 envelope."""
+    items = _items(_payload())
+    assert len(items) == 3
+    assert all({"arno", "keyword", "contents"} <= set(it) for it in items)
+
+
+def test_fatal_fields_basic_mapping():
+    item = _items(_payload())[0]
+    f = _fatal_fields(item)
+    assert f is not None
+    assert f["arno"] == "20260611111536KIZXJ8"
+    assert f["title"].startswith("[6/9, 부산 사상구]")
+    # HTML 태그 + 이미지 서버 URL 노이즈 완전 제거 (검색/임베딩 본문 정화)
+    assert "<" not in f["text"]
+    assert "portal.kosha.or.kr" not in f["text"]
+    assert "data-filename" not in f["text"]
+    # 본문 텍스트는 보존
+    assert "(사망 2명)" in f["text"]
+    assert "베란다 난간" in f["text"]
+    # published_date = arno 접두 8자리(KST 등록일), reg_dt = 14자리 등록시각 원문
+    assert f["published_date"] == date(2026, 6, 11)
+    assert f["reg_dt"] == "20260611111536"
+
+
+def test_fatal_fields_all_three_items_well_formed():
+    for item in _items(_payload()):
+        f = _fatal_fields(item)
+        assert f is not None
+        assert f["published_date"] == date(2026, 6, 11)  # 3건 모두 06-11 등록
+        assert f["reg_dt"] is not None
+        assert f["text"] and "<" not in f["text"]
+
+
+def test_fatal_fields_skips_missing_required():
+    assert _fatal_fields({"arno": "20260611111536XX", "contents": "x"}) is None  # keyword 부재
+    assert _fatal_fields({"keyword": "제목만", "contents": "x"}) is None           # arno 부재
+    assert _fatal_fields({"arno": "  ", "keyword": "  ", "contents": "x"}) is None  # 공백뿐
+
+
+def test_fatal_fields_malformed_arno_date_is_fail_quiet():
+    # arno 접두가 8자리 날짜로 안 풀리면 published_date/reg_dt = None (보조 축이라 fail-quiet)
+    f = _fatal_fields({"arno": "ABC123", "keyword": "제목", "contents": "<p>본문</p>"})
+    assert f is not None
+    assert f["published_date"] is None
+    assert f["reg_dt"] is None
+    assert f["text"] == "본문"

tests/test_queue_overview.py

@@ -103,6 +103,32 @@ def test_summarize_pool_split_attribution():
     assert macbook["pending"] == 0              # 풀 pending 은 macmini 만
 
 
+def test_summarize_by_machine_projection():
+    """build_summarize_by_machine = split 의 done_1h/done_today 를 머신별로 투영
+    (done_15m 은 제외 — 내부 state 판정 전용)."""
+    from services.queue_overview import build_summarize_by_machine
+    split = _split(
+        macbook={"done_1h": 226, "done_today": 312, "done_15m": 60},
+        macmini={"done_1h": 37, "done_today": 94, "done_15m": 9},
+    )
+    sbm = build_summarize_by_machine(split)
+    assert sbm == {
+        "macmini": {"done_1h": 37, "done_today": 94},
+        "macbook": {"done_1h": 226, "done_today": 312},
+    }
+    assert "done_15m" not in sbm["macbook"]
+
+
+def test_compose_overview_includes_summarize_by_machine():
+    """compose_overview 응답 계약에 summarize_by_machine 포함 (FE 레인 분담 재료)."""
+    now_kst = datetime(2026, 6, 13, 13, 0, tzinfo=KST)
+    stats = {"summarize": _stage(pending=1317, done_1h=264)}
+    split = _split(macbook={"done_1h": 226, "done_today": 312}, macmini={"done_1h": 37, "done_today": 94})
+    ov = compose_overview(stats, split, {}, {}, [], deep_enabled=True, now_kst=now_kst)
+    assert ov["summarize_by_machine"]["macbook"]["done_1h"] == 226
+    assert ov["summarize_by_machine"]["macmini"]["done_today"] == 94
+
+
 def test_deep_disabled_deep_summary_counts_to_macmini():
     stats = {"deep_summary": _stage(pending=2, processing=1, done_1h=3, done_today=4)}
     machines = build_machines(stats, _split(), [], deep_enabled=False)

tests/test_statute_kr_adapter.py

@@ -0,0 +1,89 @@
+"""B-1 PR① — KR 어댑터 순수 파서 fixture 테스트 (plan safety-library-1).
+
+fixture = 2026-06-13 law.go.kr 라이브 박제 (OC 새니타이즈, tests/fixtures/statute_kr/).
+파서는 순수 함수라 httpx/DB 불요 — 컨테이너 밖 로컬 실행.
+"""
+
+import gzip
+import sys
+import xml.etree.ElementTree as ET
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).parent.parent / "app"))
+
+from workers.statute_adapters import ChangeEvent  # noqa: E402
+from workers.statute_adapters.kr import detect_change, parse_search_hit  # noqa: E402
+
+FIX = Path(__file__).parent / "fixtures" / "statute_kr"
+
+
+def _read(name: str) -> str:
+    p = FIX / name
+    if name.endswith(".gz"):
+        return gzip.decompress(p.read_bytes()).decode("utf-8")
+    return p.read_text(encoding="utf-8")
+
+
+def test_parse_search_hit_exact_match():
+    hit = parse_search_hit(_read("lawsearch_sanab.xml"), "산업안전보건법")
+    assert hit is not None
+    assert hit["law_id"] == "001766"
+    assert hit["mst"] == "283449"
+    assert hit["promulgation_date"] == "20260219"
+    assert hit["effective_date"] == "20260601"
+    assert hit["status_code"] == "현행"
+
+
+def test_parse_search_hit_rejects_partial_name():
+    # totalCnt 3 인 응답에서 '산업안전보건법 시행령' 등 부분 일치는 비매칭이어야 함
+    hit = parse_search_hit(_read("lawsearch_sanab.xml"), "산업안전보건")
+    assert hit is None
+
+
+def test_detect_change_same_watermark_is_silent():
+    hit = parse_search_hit(_read("lawsearch_sanab.xml"), "산업안전보건법")
+    assert detect_change(hit, "kr-law:001766", "산업안전보건법", watermark="283449") is None
+
+
+def test_detect_change_new_mst_is_amend():
+    hit = parse_search_hit(_read("lawsearch_sanab.xml"), "산업안전보건법")
+    ev = detect_change(hit, "kr-law:001766", "산업안전보건법", watermark="283448")
+    assert isinstance(ev, ChangeEvent)
+    assert ev.kind == "amend"
+    assert ev.new_version_key == "283449"
+    assert ev.effective_date == "20260601"
+
+
+def test_detect_change_empty_watermark_is_amend():
+    # 첫 폴링(워터마크 부재) = 변경으로 감지 — PR② 부트스트랩 전 관찰 모드의 기대 동작
+    hit = parse_search_hit(_read("lawsearch_sanab.xml"), "산업안전보건법")
+    ev = detect_change(hit, "kr-law:001766", "산업안전보건법", watermark=None)
+    assert ev is not None and ev.kind == "amend"
+
+
+def test_detect_change_repeal_keyword():
+    hit = {"mst": "9", "revision_type": "폐지", "promulgation_date": None,
+           "effective_date": None, "law_id": "x", "status_code": None}
+    ev = detect_change(hit, "kr-law:x", "x", watermark="1")
+    assert ev is not None and ev.kind == "repeal"
+
+
+def test_lawservice_snapshot_semantics_rule():
+    """R7-M3 판정 박제: 전문 1콜 XML = 조문+별표 전체 스냅샷 (PR② payload 계약의 전제)."""
+    root = ET.fromstring(_read("lawservice_rule.xml.gz"))
+    articles = root.findall(".//조문단위")
+    annexes = root.findall(".//별표단위")
+    assert len(articles) >= 800, "산안기준규칙 조문 853 기대 — 전문 1콜 판정 근거"
+    assert len(annexes) == 23, "별표 23 전부 본문 XML 포함 = 스냅샷 의미론"
+    # R7-M3 ②: 별표 식별 = 구조화 필드 (suffix 문자열 파싱 불요)
+    first = annexes[0]
+    assert first.findtext("별표번호") is not None
+    assert first.findtext("별표가지번호") is not None
+
+
+def test_lawservice_sanab_basic_info():
+    root = ET.fromstring(_read("lawservice_sanab.xml.gz"))
+    assert root.findtext(".//법령ID") == "001766"
+    assert len(root.findall(".//조문단위")) >= 200
+    # 별표 없는 법령 = 별표단위 0 (스냅샷 의미론의 반대쪽 케이스)
+    assert len(root.findall(".//별표단위")) == 0

tests/test_statute_lifecycle_units.py

@@ -0,0 +1,87 @@
+"""B-1 PR② — 매핑·시리즈·payload 순수 단위 테스트 (plan safety-library-1).
+
+법령명 매핑 단위 테스트 = R8-B1 동반 계약 (검증(PR③) 전에 스윕이 도는 만큼
+매핑은 코드 레벨 선고정). 실 title 표본 = 2026-06-13 prod documents 실측 형태.
+"""
+
+import gzip
+import sys
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).parent.parent / "app"))
+
+from workers.statute_adapters.kr import parse_service_payloads  # noqa: E402
+from workers.statute_collector import (  # noqa: E402
+    legacy_law_name,
+    normalize_law_name,
+    series_suffix,
+)
+
+FIX = Path(__file__).parent / "fixtures" / "statute_kr"
+
+
+# ─── 법령명 매핑 (실 title 표본) ───
+
+def test_legacy_law_name_extraction():
+    # prod 실측 형태: '법령명 (YYYYMMDD) 섹션'
+    assert legacy_law_name("건설기술 진흥법 시행규칙 (20260611) 제6장_보칙") == "건설기술 진흥법 시행규칙"
+    assert legacy_law_name("산업안전보건법 (20260219) 전문") == "산업안전보건법"
+    assert legacy_law_name("패턴 불일치 제목") is None
+
+
+def test_mapping_equality_not_prefix():
+    """prefix 비교 금지 — '산업안전보건법' family 가 시행령 레거시를 오폭하면 안 됨."""
+    name = legacy_law_name("산업안전보건법 시행령 (20260324) 제1장")
+    assert name == "산업안전보건법 시행령"
+    assert normalize_law_name(name) != normalize_law_name("산업안전보건법")
+    assert normalize_law_name(name) == normalize_law_name("산업안전보건법 시행령")
+
+
+def test_mapping_absorbs_middle_dot_and_space():
+    """가운뎃점·공백 변형 흡수 — 유해ㆍ위험작업(정식) vs 유해위험작업(law_monitor 표기)."""
+    assert (normalize_law_name("유해ㆍ위험작업의 취업 제한에 관한 규칙")
+            == normalize_law_name("유해위험작업의 취업 제한에 관한 규칙"))
+    assert (normalize_law_name("산업안전보건기준에 관한 규칙")
+            == normalize_law_name("산업안전보건기준에관한규칙"))
+
+
+# ─── 버전 시리즈 식별자 (R7-B1 a) ───
+
+def test_series_suffix():
+    assert series_suffix("283449") is None                       # primary
+    assert series_suffix("273603|별표0001-00") == "별표0001-00"  # annex (구분 차원 포함)
+    assert series_suffix("273603|서식0003-00") == "서식0003-00"
+
+
+# ─── fetch_version payload (fixture — R4-M4 리스트 계약) ───
+
+def _read_gz(name: str) -> str:
+    return gzip.decompress((FIX / name).read_bytes()).decode("utf-8")
+
+
+def test_parse_service_payloads_rule_with_annexes():
+    payloads = parse_service_payloads(
+        _read_gz("lawservice_rule.xml.gz"), "산업안전보건기준에 관한 규칙", "273603")
+    assert payloads[0].law_doc_kind == "primary"
+    assert payloads[0].version_key == "273603"
+    assert len(payloads[0].content) > 100_000          # 853조 본문
+    annexes = [p for p in payloads if p.law_doc_kind == "annex"]
+    # 별표단위 23 중 삭제 tombstone 3 skip(별표10 '삭제 <2023.11.14>'·서식1·2 '삭제 <2012.3.5>')
+    # — KR 별표/서식 삭제 = absence 아닌 명시 tombstone (R7-M3 absence 추론 불요의 fixture 증거)
+    assert len(annexes) == 20
+    keys = [p.version_key for p in annexes]
+    assert len(keys) == len(set(keys)), "annex version_key 유일성 (uq_legal_meta_version 전제)"
+    assert all(k.startswith("273603|") for k in keys)
+    # 구분 차원 — 별표1 vs 서식N 공존 (fixture 실측: (번호,가지)만으로는 4건 충돌)
+    assert any("별표" in k for k in keys) and any("서식" in k for k in keys)
+
+
+def test_parse_service_payloads_sanab_no_annex():
+    payloads = parse_service_payloads(
+        _read_gz("lawservice_sanab.xml.gz"), "산업안전보건법", "283449")
+    assert len(payloads) == 1                          # 별표 없는 법령 = primary 단독
+    p = payloads[0]
+    assert p.promulgation_date == "20260219"
+    assert p.effective_date == "20260601"
+    assert "제2조(정의)" in p.content                  # 조문내용 보존
+    assert p.content.startswith("# 산업안전보건법")