Add app\models\extraction.py

app//models//extraction.py

@@ -0,0 +1,130 @@
+from datetime import datetime
+from typing import Optional, Dict, Any, TYPE_CHECKING
+import enum
+
+from sqlalchemy import (
+    String,
+    Integer,
+    DateTime,
+    ForeignKey,
+    Text,
+    Enum,
+    Index,
+    JSON  # FIX: Reviewer 1 #30 - Native JSON support
+)
+from sqlalchemy.orm import Mapped, mapped_column, relationship
+from sqlalchemy.sql import func
+
+from app.models.base import Base
+
+if TYPE_CHECKING:
+    from app.models.paper import Paper
+    from app.models.user import User
+
+
+class ExtractionStatus(str, enum.Enum):
+    """State machine for extraction job lifecycle."""
+    QUEUED = "queued"
+    PROCESSING = "processing"
+    COMPLETED = "completed"
+    FAILED = "failed"
+
+
+class Extraction(Base):
+    """
+    Stores structured AI extractions from research papers.
+    
+    System Role:
+    - Provides a native JSON data store for multi-dimensional research metadata.
+    - Orchestrates async AI tasks via status-tracking and job_id.
+    """
+
+    __tablename__ = "extractions"
+
+    # ------------------------------------------------------------------
+    # Identifiers & Ownership
+    # ------------------------------------------------------------------
+    id: Mapped[int] = mapped_column(Integer, primary_key=True)
+
+    paper_id: Mapped[int] = mapped_column(
+        Integer,
+        ForeignKey("papers.id", ondelete="CASCADE"),
+        nullable=False,
+    )
+
+    user_id: Mapped[Optional[int]] = mapped_column(
+        Integer,
+        ForeignKey("users.id", ondelete="SET NULL"),
+        nullable=True,
+    )
+
+    # ------------------------------------------------------------------
+    # Job Coordination
+    # ------------------------------------------------------------------
+    job_id: Mapped[str] = mapped_column(
+        String(100),
+        unique=True,
+        nullable=False,
+        index=True
+    )
+
+    status: Mapped[ExtractionStatus] = mapped_column(
+        Enum(ExtractionStatus, name="extraction_status"),
+        default=ExtractionStatus.QUEUED,
+        server_default="queued",
+    )
+
+    error_message: Mapped[Optional[str]] = mapped_column(Text)
+
+    # ------------------------------------------------------------------
+    # Native JSON Storage (FIX: Reviewer 1 #30)
+    # ------------------------------------------------------------------
+    # PICO is now a structured dictionary for native DB-level querying
+    pico_data: Mapped[Optional[Dict[str, Any]]] = mapped_column(
+        JSON, 
+        nullable=True,
+        comment="Structured PICO (Population, Intervention, Comparison, Outcome)"
+    )
+
+    # Risk-of-Bias (RoB) assessment
+    risk_of_bias: Mapped[Optional[Dict[str, Any]]] = mapped_column(
+        JSON, 
+        nullable=True,
+        comment="Native JSON structure for domain-specific bias metrics"
+    )
+
+    # ------------------------------------------------------------------
+    # Audit & Metadata
+    # ------------------------------------------------------------------
+    model_version: Mapped[str] = mapped_column(
+        String(50),
+        default="llama-3.1-70b",
+        server_default="llama-3.1-70b",
+    )
+
+    created_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True),
+        server_default=func.now(),
+    )
+
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True),
+        server_default=func.now(),
+        onupdate=func.now(),
+    )
+
+    completed_at: Mapped[Optional[datetime]] = mapped_column(
+        DateTime(timezone=True)
+    )
+
+    # ------------------------------------------------------------------
+    # Relationships
+    # ------------------------------------------------------------------
+    paper: Mapped["Paper"] = relationship("Paper", lazy="joined")
+    user: Mapped[Optional["User"]] = relationship("User", lazy="select")
+
+    # ------------------------------------------------------------------
+    # Representation
+    # ------------------------------------------------------------------
+    def __repr__(self) -> str:
+        return f"<Extraction(job_id={self.job_id}, status={self.status})>"