Initial Aurora Core 0.1 implementation

Production-grade media metadata backend per docs/aurora-core-spec.md:
Aurora-owned UUIDv7 identity system, normalized movie/TV/people/credits
schema, provider abstraction with a TMDb adapter, an idempotent async
ingestion pipeline (normalize -> match -> upsert -> provenance), a
Postgres FTS/trigram search layer, a stateless FastAPI v1 API with
Redis caching/rate-limiting, arq workers, dev/prod Docker Compose
topologies, unit + integration + load tests, CI, and architecture/
database/API/ERD documentation.

Co-Authored-By: Claude Sonnet 5 <[email protected]>

app/models/provider.py

@@ -0,0 +1,108 @@
+"""Provider abstraction and provenance tables.
+
+`Provider` identifies a metadata source (TMDb, IMDb dataset export, ...).
+`ProviderRecord` is the provenance ledger: one row per (provider, entity
+type, provider-native ID) linking to the Aurora entity it was matched to,
+retaining the raw payload and match confidence so ambiguous matches stay
+reviewable instead of being silently merged (see docs/architecture.md).
+`IngestionJob` tracks async ingestion work so jobs are observable, retryable,
+and protected against duplicate processing via `idempotency_key`.
+"""
+
+from __future__ import annotations
+
+import uuid
+from datetime import datetime
+
+from sqlalchemy import JSON, DateTime, ForeignKey, Integer, String, Text, UniqueConstraint, func
+from sqlalchemy.dialects.postgresql import JSONB
+from sqlalchemy.orm import Mapped, mapped_column, relationship
+
+from app.models.base import Base, TimestampMixin, UUIDPrimaryKeyMixin
+from app.models.enums import IngestionJobStatus, MatchConfidence
+
+JSONVariant = JSONB().with_variant(JSON(), "sqlite")
+
+
+class Provider(Base, UUIDPrimaryKeyMixin, TimestampMixin):
+    """A metadata data source. Providers are replaceable adapters, never
+    Aurora's identity system — see app/providers/base.py."""
+
+    __tablename__ = "providers"
+
+    slug: Mapped[str] = mapped_column(String(50), unique=True, nullable=False)  # "tmdb", "imdb"
+    name: Mapped[str] = mapped_column(String(200), nullable=False)
+    base_url: Mapped[str | None] = mapped_column(String(500), nullable=True)
+    is_active: Mapped[bool] = mapped_column(default=True, nullable=False)
+    config: Mapped[dict] = mapped_column(JSONVariant, default=dict)
+
+
+class ProviderRecord(Base, UUIDPrimaryKeyMixin, TimestampMixin):
+    """Provenance ledger row: this provider's native record, and what Aurora
+    entity (if any) it has been matched to."""
+
+    __tablename__ = "provider_records"
+    __table_args__ = (
+        UniqueConstraint(
+            "provider_id", "entity_type", "external_id", name="uq_provider_record_source"
+        ),
+    )
+
+    provider_id: Mapped[uuid.UUID] = mapped_column(
+        ForeignKey("providers.id", ondelete="CASCADE"), nullable=False
+    )
+    entity_type: Mapped[str] = mapped_column(String(30), nullable=False)
+    external_id: Mapped[str] = mapped_column(String(200), nullable=False)
+
+    # Polymorphic reference to the Aurora entity this record resolved to.
+    # No DB-level FK (the target table varies by entity_type); integrity is
+    # enforced by the entity matcher / repositories, not the schema.
+    aurora_entity_id: Mapped[uuid.UUID | None] = mapped_column(nullable=True)
+
+    raw_data: Mapped[dict] = mapped_column(JSONVariant, default=dict)
+    source_version: Mapped[str | None] = mapped_column(String(100), nullable=True)
+    match_confidence: Mapped[MatchConfidence] = mapped_column(
+        default=MatchConfidence.NEW, nullable=False
+    )
+    match_notes: Mapped[str | None] = mapped_column(Text, nullable=True)
+
+    ingestion_job_id: Mapped[uuid.UUID | None] = mapped_column(
+        ForeignKey("ingestion_jobs.id", ondelete="SET NULL"), nullable=True
+    )
+    last_synced_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+
+    provider: Mapped[Provider] = relationship()
+
+
+class IngestionJob(Base, UUIDPrimaryKeyMixin, TimestampMixin):
+    """A unit of asynchronous ingestion work processed by workers."""
+
+    __tablename__ = "ingestion_jobs"
+    __table_args__ = (UniqueConstraint("idempotency_key", name="uq_ingestion_job_idempotency"),)
+
+    provider_id: Mapped[uuid.UUID] = mapped_column(
+        ForeignKey("providers.id", ondelete="CASCADE"), nullable=False
+    )
+    job_type: Mapped[str] = mapped_column(String(50), nullable=False)  # "movie_sync", "full_sync"
+    entity_type: Mapped[str | None] = mapped_column(String(30), nullable=True)
+    external_id: Mapped[str | None] = mapped_column(String(200), nullable=True)
+
+    idempotency_key: Mapped[str] = mapped_column(String(300), nullable=False)
+    status: Mapped[IngestionJobStatus] = mapped_column(
+        default=IngestionJobStatus.PENDING, nullable=False
+    )
+    attempts: Mapped[int] = mapped_column(Integer, default=0, nullable=False)
+    max_attempts: Mapped[int] = mapped_column(Integer, default=5, nullable=False)
+
+    started_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    finished_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+
+    error: Mapped[str | None] = mapped_column(Text, nullable=True)
+    result_aurora_id: Mapped[uuid.UUID | None] = mapped_column(nullable=True)
+
+    provider: Mapped[Provider] = relationship()
+
+
+__all__ = ["Provider", "ProviderRecord", "IngestionJob"]