src/personalization_module/preference_acquisition/training_data/models.py

from __future__ import annotations

from datetime import datetime

from sqlalchemy import (
    Boolean,
    Column,
    DateTime,
    Float,
    ForeignKey,
    Index,
    Integer,
    String,
    Text,
)

from src.personalization_module.preference_acquisition.preference_memory.db import Base


class DecisionEvent(Base):
    __tablename__ = "decision_events"

    id = Column(Integer, primary_key=True)
    request_id = Column(String, unique=True, nullable=False)
    user_node_id = Column(Integer, ForeignKey("nodes.id"), nullable=False, index=True)
    task_id = Column(String, nullable=True)
    action_input_json = Column(Text, nullable=False, default="{}")
    context_input_json = Column(Text, nullable=False, default="{}")
    preference_snapshot_json = Column(Text, nullable=False, default="[]")

    predicted_decision = Column(String, nullable=False)
    class_scores_json = Column(Text, nullable=False, default="{}")
    confidence = Column(Float, nullable=True)
    runner_up_json = Column(Text, nullable=False, default="{}")

    model_name = Column(String, nullable=True)
    model_version = Column(String, nullable=True)
    feature_schema_version = Column(String, nullable=True)

    timestamp_utc = Column(DateTime, nullable=False, default=datetime.utcnow, index=True)
    created_at = Column(DateTime, nullable=False, default=datetime.utcnow)


Index(
    "ix_decision_events_user_created_at",
    DecisionEvent.user_node_id,
    DecisionEvent.created_at,
)


class FeedbackEvent(Base):
    __tablename__ = "feedback_events"

    id = Column(Integer, primary_key=True)
    request_id = Column(String, nullable=True, index=True)
    user_node_id = Column(Integer, ForeignKey("nodes.id"), nullable=False, index=True)
    feedback_id = Column(String, nullable=True, unique=True)

    predicted_decision = Column(String, nullable=True)
    feedback_type = Column(String, nullable=False)
    correct_decision = Column(String, nullable=True)
    free_text = Column(Text, nullable=True)

    preference_updates_json = Column(Text, nullable=False, default="[]")
    graph_update_summary_json = Column(Text, nullable=False, default="{}")
    timestamp_utc = Column(DateTime, nullable=False, default=datetime.utcnow, index=True)
    applied_to_graph = Column(Boolean, nullable=False, default=False)

    created_at = Column(DateTime, nullable=False, default=datetime.utcnow)


Index(
    "ix_feedback_events_user_created_at",
    FeedbackEvent.user_node_id,
    FeedbackEvent.created_at,
)


class TrainingSample(Base):
    __tablename__ = "training_samples"

    id = Column(Integer, primary_key=True)
    decision_event_id = Column(
        Integer,
        ForeignKey("decision_events.id"),
        nullable=False,
        index=True,
    )
    feedback_event_id = Column(
        Integer,
        ForeignKey("feedback_events.id"),
        nullable=True,
        index=True,
    )
    user_node_id = Column(Integer, ForeignKey("nodes.id"), nullable=False, index=True)

    label_action = Column(String, nullable=False)
    sample_payload_json = Column(Text, nullable=False, default="{}")
    feature_vector_json = Column(Text, nullable=False, default="{}")

    split = Column(String, nullable=True)
    sample_source = Column(String, nullable=False, default="feedback_builder_v1")

    created_at = Column(DateTime, nullable=False, default=datetime.utcnow)


Index(
    "ix_training_samples_user_created_at",
    TrainingSample.user_node_id,
    TrainingSample.created_at,
)