Two low-risk Pydantic V1 idioms that will be hard errors in V3:
- Query(regex=...) -> Query(pattern=...) (audit_routes, control_generator_routes)
- class Config: from_attributes=True -> model_config = ConfigDict(...)
in source_policy_router.py (schemas.py is intentionally skipped — it is
the Phase 1 schema-split target and the ConfigDict conversion is most
efficient to do during that split).
Naive -> aware datetime sweep across 47 files:
- datetime.utcnow() -> datetime.now(timezone.utc)
- default=datetime.utcnow -> default=lambda: datetime.now(timezone.utc)
- onupdate=datetime.utcnow -> onupdate=lambda: datetime.now(timezone.utc)
All SQLAlchemy DateTime columns in the project already declare
timezone=True, so the DB schema expects aware datetimes. Before this
commit, the in-Python side was generating naive values and the driver
was silently coercing them. This is a latent-bug fix, not a behavior
change at the DB boundary.
Verified:
- 173/173 pytest compliance/tests/ pass (same as baseline)
- tests/contracts/test_openapi_baseline.py passes (360 paths,
484 operations unchanged)
- DeprecationWarning count dropped from 158 -> 35
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Sucht alle RAG-Kollektionen nach Prüfaspekten und legt automatisch
Anforderungen in der DB an. Kernfeatures:
- Durchsucht alle 6 RAG-Kollektionen parallel (bp_compliance_ce,
bp_compliance_recht, bp_compliance_gesetze, bp_compliance_datenschutz,
bp_dsfa_corpus, bp_legal_templates)
- Erkennt BSI Prüfaspekte (O.Purp_6) im Artikel-Feld und per Regex
- Dedupliziert nach (regulation_code, article) — safe to call many times
- Auto-erstellt Regulations-Stubs für unbekannte regulation_codes
- dry_run=true zeigt was erstellt würde ohne DB-Schreibzugriff
- Optionale Filter: collections, regulation_codes, search_queries
- 18 Tests (alle bestanden)
- Frontend: "Aus RAG extrahieren" Button auf /sdk/requirements
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
