datadoom 0.1.0.dev0__py3-none-any.whl

datadoom/store/migrations/env.py

@@ -0,0 +1,53 @@
+"""Alembic environment (07 §5).
+
+Driven programmatically from ``store.db.init_database`` — the URL is injected via
+``Config.set_main_option("sqlalchemy.url", ...)``. We bind ``target_metadata`` to
+the ORM ``Base`` so ``alembic revision --autogenerate`` stays usable, but the
+checked-in linear history is authoritative.
+"""
+
+from __future__ import annotations
+
+from alembic import context
+from sqlalchemy import engine_from_config, pool
+
+from datadoom.store.models import Base
+
+config = context.config
+target_metadata = Base.metadata
+
+
+def run_migrations_offline() -> None:
+    url = config.get_main_option("sqlalchemy.url")
+    context.configure(
+        url=url,
+        target_metadata=target_metadata,
+        literal_binds=True,
+        dialect_opts={"paramstyle": "named"},
+        render_as_batch=True,
+    )
+    with context.begin_transaction():
+        context.run_migrations()
+
+
+def run_migrations_online() -> None:
+    connectable = engine_from_config(
+        config.get_section(config.config_ini_section, {}),
+        prefix="sqlalchemy.",
+        poolclass=pool.NullPool,
+    )
+    with connectable.connect() as connection:
+        context.configure(
+            connection=connection,
+            target_metadata=target_metadata,
+            render_as_batch=True,
+        )
+        with context.begin_transaction():
+            context.run_migrations()
+    connectable.dispose()
+
+
+if context.is_offline_mode():
+    run_migrations_offline()
+else:
+    run_migrations_online()

datadoom/store/migrations/script.py.mako

@@ -0,0 +1,24 @@
+"""${message}
+
+Revision ID: ${up_revision}
+Revises: ${down_revision | comma,n}
+Create Date: ${create_date}
+"""
+from __future__ import annotations
+
+from alembic import op
+import sqlalchemy as sa
+${imports if imports else ""}
+
+revision = ${repr(up_revision)}
+down_revision = ${repr(down_revision)}
+branch_labels = ${repr(branch_labels)}
+depends_on = ${repr(depends_on)}
+
+
+def upgrade() -> None:
+    ${upgrades if upgrades else "pass"}
+
+
+def downgrade() -> None:
+    ${downgrades if downgrades else "pass"}

datadoom/store/migrations/versions/0001_init.py

@@ -0,0 +1,149 @@
+"""0001 init — core tables (07 §2, no team-mode ``users``).
+
+Revision ID: 0001_init
+Revises:
+Create Date: 2026-06-01
+"""
+from __future__ import annotations
+
+import sqlalchemy as sa
+from alembic import op
+
+revision = "0001_init"
+down_revision = None
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "datasets",
+        sa.Column("dataset_id", sa.String(length=36), primary_key=True),
+        sa.Column("name", sa.String(), nullable=False),
+        sa.Column("description", sa.String(), nullable=True),
+        # Soft references (app-maintained) to break the datasets⇄specs cycle.
+        sa.Column("current_spec_id", sa.String(length=36), nullable=True),
+        sa.Column("latest_run_id", sa.String(length=36), nullable=True),
+        sa.Column("status", sa.String(), nullable=False, server_default="draft"),
+        sa.Column("owner_id", sa.String(length=36), nullable=True),
+        sa.Column("created_at", sa.String(), nullable=False),
+        sa.Column("updated_at", sa.String(), nullable=False),
+    )
+    op.create_index("ux_datasets_owner_name", "datasets", ["owner_id", "name"], unique=True)
+    op.create_index("ix_datasets_status", "datasets", ["status"])
+
+    op.create_table(
+        "specs",
+        sa.Column("spec_id", sa.String(length=36), primary_key=True),
+        sa.Column(
+            "dataset_id",
+            sa.String(length=36),
+            sa.ForeignKey("datasets.dataset_id", ondelete="CASCADE"),
+            nullable=False,
+        ),
+        sa.Column("spec_hash", sa.String(length=64), nullable=False),
+        sa.Column("body", sa.JSON(), nullable=False),
+        sa.Column("datadoom_version", sa.String(), nullable=False),
+        sa.Column("version", sa.Integer(), nullable=False),
+        sa.Column("created_at", sa.String(), nullable=False),
+    )
+    op.create_index("ix_specs_dataset", "specs", ["dataset_id"])
+    op.create_index("ix_specs_hash", "specs", ["spec_hash"])
+    op.create_index(
+        "ux_specs_dataset_version", "specs", ["dataset_id", "version"], unique=True
+    )
+
+    op.create_table(
+        "generation_runs",
+        sa.Column("run_id", sa.String(length=36), primary_key=True),
+        sa.Column(
+            "dataset_id",
+            sa.String(length=36),
+            sa.ForeignKey("datasets.dataset_id", ondelete="CASCADE"),
+            nullable=False,
+        ),
+        sa.Column(
+            "spec_id",
+            sa.String(length=36),
+            sa.ForeignKey("specs.spec_id", ondelete="CASCADE"),
+            nullable=False,
+        ),
+        sa.Column("seed", sa.Integer(), nullable=False),
+        sa.Column("status", sa.String(), nullable=False, server_default="queued"),
+        sa.Column("stage", sa.String(), nullable=True),
+        sa.Column("progress_pct", sa.Integer(), nullable=False, server_default="0"),
+        sa.Column("error", sa.JSON(), nullable=True),
+        sa.Column("metrics", sa.JSON(), nullable=True),
+        sa.Column("started_at", sa.String(), nullable=True),
+        sa.Column("finished_at", sa.String(), nullable=True),
+        sa.Column("created_at", sa.String(), nullable=False),
+    )
+    op.create_index("ix_runs_dataset", "generation_runs", ["dataset_id"])
+    op.create_index("ix_runs_status", "generation_runs", ["status"])
+    op.create_index("ix_runs_repro", "generation_runs", ["spec_id", "seed"])
+
+    op.create_table(
+        "artifacts",
+        sa.Column("artifact_id", sa.String(length=36), primary_key=True),
+        sa.Column(
+            "run_id",
+            sa.String(length=36),
+            sa.ForeignKey("generation_runs.run_id", ondelete="CASCADE"),
+            nullable=False,
+        ),
+        sa.Column("version", sa.String(), nullable=False),
+        sa.Column("split", sa.String(), nullable=True),
+        sa.Column("format", sa.String(), nullable=False),
+        sa.Column("storage_uri", sa.String(), nullable=False),
+        sa.Column("checksum_sha256", sa.String(length=64), nullable=False),
+        sa.Column("size_bytes", sa.Integer(), nullable=False),
+        sa.Column("created_at", sa.String(), nullable=False),
+    )
+    op.create_index("ix_artifacts_run", "artifacts", ["run_id"])
+
+    op.create_table(
+        "reports",
+        sa.Column("report_id", sa.String(length=36), primary_key=True),
+        sa.Column(
+            "run_id",
+            sa.String(length=36),
+            sa.ForeignKey("generation_runs.run_id", ondelete="CASCADE"),
+            nullable=False,
+            unique=True,
+        ),
+        sa.Column("compliance_score", sa.Float(), nullable=True),
+        sa.Column("distribution", sa.JSON(), nullable=True),
+        sa.Column("correlation", sa.JSON(), nullable=True),
+        sa.Column("causal_truth", sa.JSON(), nullable=True),
+        sa.Column("difficulty", sa.JSON(), nullable=True),
+        sa.Column("failures", sa.JSON(), nullable=True),
+        sa.Column("determinism", sa.JSON(), nullable=True),
+        sa.Column("created_at", sa.String(), nullable=False),
+    )
+
+    op.create_table(
+        "plugins",
+        sa.Column("name", sa.String(), primary_key=True),
+        sa.Column("kind", sa.String(), nullable=False),
+        sa.Column("version", sa.String(), nullable=True),
+        sa.Column("schema", sa.JSON(), nullable=True),
+        sa.Column("enabled", sa.Integer(), nullable=False, server_default="1"),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("plugins")
+    op.drop_table("reports")
+    op.drop_index("ix_artifacts_run", table_name="artifacts")
+    op.drop_table("artifacts")
+    op.drop_index("ix_runs_repro", table_name="generation_runs")
+    op.drop_index("ix_runs_status", table_name="generation_runs")
+    op.drop_index("ix_runs_dataset", table_name="generation_runs")
+    op.drop_table("generation_runs")
+    op.drop_index("ux_specs_dataset_version", table_name="specs")
+    op.drop_index("ix_specs_hash", table_name="specs")
+    op.drop_index("ix_specs_dataset", table_name="specs")
+    op.drop_table("specs")
+    op.drop_index("ix_datasets_status", table_name="datasets")
+    op.drop_index("ux_datasets_owner_name", table_name="datasets")
+    op.drop_table("datasets")

datadoom/store/migrations/versions/0002_report_mutual_information.py

@@ -0,0 +1,23 @@
+"""0002 — add ``reports.mutual_information`` (P2 causal: MI matrix, 05 §7).
+
+Adds the JSON column that ``ReportRepository.upsert`` writes the mutual-information
+matrix into. Separate from ``0001_init`` so databases created before this column
+existed are upgraded in place on startup (``alembic upgrade head``).
+"""
+from __future__ import annotations
+
+import sqlalchemy as sa
+from alembic import op
+
+revision = "0002_report_mutual_information"
+down_revision = "0001_init"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column("reports", sa.Column("mutual_information", sa.JSON(), nullable=True))
+
+
+def downgrade() -> None:
+    op.drop_column("reports", "mutual_information")

datadoom/store/migrations/versions/0003_run_name.py

@@ -0,0 +1,23 @@
+"""0003 — add ``generation_runs.name`` (named generations).
+
+Each generation gets an optional human label. Nullable so databases created
+before naming existed upgrade in place; the UI requires a name on new runs and
+falls back to the run id for older, unnamed rows.
+"""
+from __future__ import annotations
+
+import sqlalchemy as sa
+from alembic import op
+
+revision = "0003_run_name"
+down_revision = "0002_report_mutual_information"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column("generation_runs", sa.Column("name", sa.String(), nullable=True))
+
+
+def downgrade() -> None:
+    op.drop_column("generation_runs", "name")

datadoom/store/migrations/versions/0004_report_profile.py

@@ -0,0 +1,24 @@
+"""0004 — add ``reports.profile`` (per-column data profile / "Column Guide").
+
+Adds the JSON column ``ReportRepository.upsert`` writes the per-column profile
+into: summary statistics, role, causal parents, failure attribution, and
+ML-handling advice per column. Separate from earlier revisions so databases
+created before this column existed upgrade in place on startup.
+"""
+from __future__ import annotations
+
+import sqlalchemy as sa
+from alembic import op
+
+revision = "0004_report_profile"
+down_revision = "0003_run_name"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column("reports", sa.Column("profile", sa.JSON(), nullable=True))
+
+
+def downgrade() -> None:
+    op.drop_column("reports", "profile")

datadoom/store/models.py

@@ -0,0 +1,170 @@
+"""SQLAlchemy ORM models — mirror docs 06 §3 and 07 §2.
+
+Portable DDL (SQLite default, Postgres-compatible): UUID string PKs, ISO-8601 UTC
+timestamps stored as TEXT, JSON columns via SQLAlchemy's generic ``JSON`` type
+(SQLite ``JSON1`` / Postgres ``JSONB``).
+
+Circular references (``datasets.current_spec_id`` ⇄ ``specs.dataset_id`` and
+``datasets.latest_run_id``) are kept as plain indexed columns — "soft" FKs the
+application maintains — exactly as doc 07 notes, to avoid SQLite's
+create-order / ALTER-FK limitations. The hard, cascade-enforced FKs all point
+"downward" (spec → dataset, run → dataset/spec, artifact/report → run).
+"""
+
+from __future__ import annotations
+
+from typing import Any
+
+from sqlalchemy import JSON as SAJSON
+from sqlalchemy import ForeignKey, Index, Integer, String
+from sqlalchemy.orm import DeclarativeBase, Mapped, mapped_column, relationship
+
+
+class Base(DeclarativeBase):
+    pass
+
+
+class DatasetRow(Base):
+    __tablename__ = "datasets"
+
+    dataset_id: Mapped[str] = mapped_column(String(36), primary_key=True)
+    name: Mapped[str] = mapped_column(String, nullable=False)
+    description: Mapped[str | None] = mapped_column(String, nullable=True)
+    # Soft references (app-maintained; see module docstring).
+    current_spec_id: Mapped[str | None] = mapped_column(String(36), nullable=True)
+    latest_run_id: Mapped[str | None] = mapped_column(String(36), nullable=True)
+    status: Mapped[str] = mapped_column(String, nullable=False, default="draft")
+    owner_id: Mapped[str | None] = mapped_column(String(36), nullable=True)
+    created_at: Mapped[str] = mapped_column(String, nullable=False)
+    updated_at: Mapped[str] = mapped_column(String, nullable=False)
+
+    # passive_deletes lets the DB's ON DELETE CASCADE do the work in one statement,
+    # avoiding ORM delete-ordering issues across the specs⇄runs FK.
+    specs: Mapped[list[SpecRow]] = relationship(
+        back_populates="dataset", cascade="all, delete-orphan", passive_deletes=True
+    )
+    runs: Mapped[list[GenerationRunRow]] = relationship(
+        back_populates="dataset", cascade="all, delete-orphan", passive_deletes=True
+    )
+
+    __table_args__ = (
+        Index("ux_datasets_owner_name", "owner_id", "name", unique=True),
+        Index("ix_datasets_status", "status"),
+    )
+
+
+class SpecRow(Base):
+    __tablename__ = "specs"
+
+    spec_id: Mapped[str] = mapped_column(String(36), primary_key=True)
+    dataset_id: Mapped[str] = mapped_column(
+        String(36), ForeignKey("datasets.dataset_id", ondelete="CASCADE"), nullable=False
+    )
+    spec_hash: Mapped[str] = mapped_column(String(64), nullable=False)
+    body: Mapped[dict[str, Any]] = mapped_column(SAJSON, nullable=False)
+    datadoom_version: Mapped[str] = mapped_column(String, nullable=False)
+    version: Mapped[int] = mapped_column(Integer, nullable=False)
+    created_at: Mapped[str] = mapped_column(String, nullable=False)
+
+    dataset: Mapped[DatasetRow] = relationship(back_populates="specs")
+
+    __table_args__ = (
+        Index("ix_specs_dataset", "dataset_id"),
+        Index("ix_specs_hash", "spec_hash"),
+        Index("ux_specs_dataset_version", "dataset_id", "version", unique=True),
+    )
+
+
+class GenerationRunRow(Base):
+    __tablename__ = "generation_runs"
+
+    run_id: Mapped[str] = mapped_column(String(36), primary_key=True)
+    dataset_id: Mapped[str] = mapped_column(
+        String(36), ForeignKey("datasets.dataset_id", ondelete="CASCADE"), nullable=False
+    )
+    spec_id: Mapped[str] = mapped_column(
+        String(36), ForeignKey("specs.spec_id", ondelete="CASCADE"), nullable=False
+    )
+    # Human label for the generation. Optional for rows created before naming
+    # existed; the UI mandates one on new runs (falls back to the id for display).
+    name: Mapped[str | None] = mapped_column(String, nullable=True)
+    seed: Mapped[int] = mapped_column(Integer, nullable=False)
+    status: Mapped[str] = mapped_column(String, nullable=False, default="queued")
+    stage: Mapped[str | None] = mapped_column(String, nullable=True)
+    progress_pct: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    error: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    metrics: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    started_at: Mapped[str | None] = mapped_column(String, nullable=True)
+    finished_at: Mapped[str | None] = mapped_column(String, nullable=True)
+    created_at: Mapped[str] = mapped_column(String, nullable=False)
+
+    dataset: Mapped[DatasetRow] = relationship(back_populates="runs")
+    # Read-only handle to the immutable spec snapshot this run was generated from,
+    # so callers can surface its ``spec_hash`` (the version-control anchor).
+    spec: Mapped[SpecRow] = relationship(viewonly=True)
+    artifacts: Mapped[list[ArtifactRow]] = relationship(
+        back_populates="run", cascade="all, delete-orphan", passive_deletes=True
+    )
+    report: Mapped[ReportRow | None] = relationship(
+        back_populates="run", cascade="all, delete-orphan", uselist=False, passive_deletes=True
+    )
+
+    __table_args__ = (
+        Index("ix_runs_dataset", "dataset_id"),
+        Index("ix_runs_status", "status"),
+        Index("ix_runs_repro", "spec_id", "seed"),
+    )
+
+
+class ArtifactRow(Base):
+    __tablename__ = "artifacts"
+
+    artifact_id: Mapped[str] = mapped_column(String(36), primary_key=True)
+    run_id: Mapped[str] = mapped_column(
+        String(36), ForeignKey("generation_runs.run_id", ondelete="CASCADE"), nullable=False
+    )
+    version: Mapped[str] = mapped_column(String, nullable=False)  # clean | injected
+    split: Mapped[str | None] = mapped_column(String, nullable=True)
+    format: Mapped[str] = mapped_column(String, nullable=False)
+    storage_uri: Mapped[str] = mapped_column(String, nullable=False)
+    checksum_sha256: Mapped[str] = mapped_column(String(64), nullable=False)
+    size_bytes: Mapped[int] = mapped_column(Integer, nullable=False)
+    created_at: Mapped[str] = mapped_column(String, nullable=False)
+
+    run: Mapped[GenerationRunRow] = relationship(back_populates="artifacts")
+
+    __table_args__ = (Index("ix_artifacts_run", "run_id"),)
+
+
+class ReportRow(Base):
+    __tablename__ = "reports"
+
+    report_id: Mapped[str] = mapped_column(String(36), primary_key=True)
+    run_id: Mapped[str] = mapped_column(
+        String(36),
+        ForeignKey("generation_runs.run_id", ondelete="CASCADE"),
+        nullable=False,
+        unique=True,
+    )
+    compliance_score: Mapped[float | None] = mapped_column(nullable=True)
+    distribution: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    correlation: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    mutual_information: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    causal_truth: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    difficulty: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    failures: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    profile: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    determinism: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    created_at: Mapped[str] = mapped_column(String, nullable=False)
+
+    run: Mapped[GenerationRunRow] = relationship(back_populates="report")
+
+
+class PluginRow(Base):
+    __tablename__ = "plugins"
+
+    name: Mapped[str] = mapped_column(String, primary_key=True)
+    kind: Mapped[str] = mapped_column(String, nullable=False)
+    version: Mapped[str | None] = mapped_column(String, nullable=True)
+    schema: Mapped[dict[str, Any] | None] = mapped_column(SAJSON, nullable=True)
+    enabled: Mapped[int] = mapped_column(Integer, nullable=False, default=1)