first commit

backend/app/models/__init__.py

@@ -0,0 +1,21 @@
+from app.models.database import (
+    Base,
+    Defect,
+    Experiment,
+    ExperimentRun,
+    PromptTemplate,
+    PromptTemplateVersion,
+    Result,
+    Sample,
+)
+
+__all__ = [
+    "Base",
+    "Sample",
+    "Defect",
+    "PromptTemplate",
+    "PromptTemplateVersion",
+    "Experiment",
+    "ExperimentRun",
+    "Result",
+]

backend/app/models/database.py

@@ -0,0 +1,146 @@
+import uuid
+from datetime import datetime, timezone
+from typing import Optional
+
+from sqlalchemy import (
+    JSON,
+    Column,
+    DateTime,
+    Float,
+    ForeignKey,
+    Integer,
+    String,
+    Text,
+    UniqueConstraint,
+)
+from sqlalchemy.dialects.postgresql import UUID
+from sqlalchemy.orm import relationship
+
+from app.db import Base
+
+
+def now_utc() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+class Sample(Base):
+    __tablename__ = "samples"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    repo = Column(String(255), nullable=False)
+    commit_sha = Column(String(40), nullable=False)
+    language = Column(String(50), nullable=False)
+    diff = Column(Text, nullable=False)
+    before_context = Column(JSON, nullable=True)
+    after_context = Column(JSON, nullable=True)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+
+    defects = relationship("Defect", back_populates="sample", cascade="all, delete-orphan")
+    runs = relationship("ExperimentRun", back_populates="sample")
+
+
+class Defect(Base):
+    __tablename__ = "defects"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    sample_id = Column(UUID(as_uuid=True), ForeignKey("samples.id"), nullable=False)
+    defect_type = Column(String(100), nullable=False)
+    language = Column(String(50), nullable=False)
+    line_start = Column(Integer, nullable=True)
+    line_end = Column(Integer, nullable=True)
+    description = Column(Text, nullable=True)
+    reference_fix = Column(Text, nullable=True)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+
+    sample = relationship("Sample", back_populates="defects")
+
+
+class PromptTemplate(Base):
+    __tablename__ = "prompt_templates"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    strategy_id = Column(String(100), nullable=False)
+    level = Column(String(20), nullable=False)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+
+    versions = relationship(
+        "PromptTemplateVersion",
+        back_populates="template",
+        cascade="all, delete-orphan",
+        order_by="PromptTemplateVersion.version_number",
+    )
+
+    __table_args__ = (UniqueConstraint("strategy_id", "level", name="uix_strategy_level"),)
+
+
+class PromptTemplateVersion(Base):
+    __tablename__ = "prompt_template_versions"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    template_id = Column(UUID(as_uuid=True), ForeignKey("prompt_templates.id"), nullable=False)
+    version_number = Column(Integer, nullable=False)
+    body = Column(Text, nullable=False)
+    variables_schema = Column(JSON, nullable=False, default=dict)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+
+    template = relationship("PromptTemplate", back_populates="versions")
+    runs = relationship("ExperimentRun", back_populates="template_version")
+
+
+class Experiment(Base):
+    __tablename__ = "experiments"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    name = Column(String(255), nullable=False)
+    models = Column(JSON, nullable=False)
+    levels = Column(JSON, nullable=False)
+    sample_ids = Column(JSON, nullable=False)
+    repeats = Column(Integer, nullable=False, default=3)
+    sampling_params = Column(JSON, nullable=False, default=dict)
+    status = Column(String(20), default="pending", nullable=False)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+
+    runs = relationship("ExperimentRun", back_populates="experiment", cascade="all, delete-orphan")
+
+
+class ExperimentRun(Base):
+    __tablename__ = "experiment_runs"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    run_id = Column(String(64), unique=True, nullable=False, default=lambda: str(uuid.uuid4()))
+    experiment_id = Column(UUID(as_uuid=True), ForeignKey("experiments.id"), nullable=False)
+    sample_id = Column(UUID(as_uuid=True), ForeignKey("samples.id"), nullable=False)
+    model_id = Column(String(100), nullable=False)
+    template_version_id = Column(UUID(as_uuid=True), ForeignKey("prompt_template_versions.id"), nullable=False)
+    repeat_index = Column(Integer, nullable=False)
+    status = Column(String(20), default="pending", nullable=False)
+    retry_count = Column(Integer, default=0, nullable=False)
+    sampling_params = Column(JSON, nullable=False, default=dict)
+    started_at = Column(DateTime, nullable=True)
+    completed_at = Column(DateTime, nullable=True)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+
+    experiment = relationship("Experiment", back_populates="runs")
+    sample = relationship("Sample", back_populates="runs")
+    template_version = relationship("PromptTemplateVersion", back_populates="runs")
+    result = relationship("Result", back_populates="run", uselist=False, cascade="all, delete-orphan")
+
+
+class Result(Base):
+    __tablename__ = "results"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    run_id = Column(UUID(as_uuid=True), ForeignKey("experiment_runs.id"), nullable=False, unique=True)
+    raw_output = Column(Text, nullable=True)
+    token_usage = Column(JSON, nullable=True)
+    latency_ms = Column(Float, nullable=True)
+    parsed_findings = Column(JSON, nullable=True)
+    detection_rate = Column(Float, nullable=True)
+    false_positive_rate = Column(Float, nullable=True)
+    coverage_rate = Column(Float, nullable=True)
+    stability_score = Column(Float, nullable=True)
+    likert_score = Column(Integer, nullable=True)
+    created_at = Column(DateTime, default=now_utc, nullable=False)
+    updated_at = Column(DateTime, default=now_utc, onupdate=now_utc, nullable=False)
+
+    run = relationship("ExperimentRun", back_populates="result")