feat: add temporal Mol explorer

2026-07-14 15:19:42 +02:00
parent 0ec1ab4970
commit c0943fe7d4
39 changed files with 3065 additions and 163 deletions
@@ -0,0 +1,72 @@
+"""Add temporal dataset metadata and durable dataset-version provenance."""
+
+from alembic import op
+import sqlalchemy as sa
+
+
+revision = "202607140001"
+down_revision = "202606120900"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column("datasets", sa.Column("temporal_series_key", sa.String(length=255), nullable=True))
+    op.add_column("datasets", sa.Column("observed_at", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("datasets", sa.Column("valid_from", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("datasets", sa.Column("valid_to", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("datasets", sa.Column("temporal_granularity", sa.String(length=32), nullable=True))
+    op.add_column("datasets", sa.Column("source_version", sa.String(length=120), nullable=True))
+
+    op.add_column("dataset_versions", sa.Column("source_version", sa.String(length=120), nullable=True))
+    op.add_column("dataset_versions", sa.Column("observed_at", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("dataset_versions", sa.Column("valid_from", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("dataset_versions", sa.Column("valid_to", sa.DateTime(timezone=True), nullable=True))
+    op.add_column("dataset_versions", sa.Column("checksum_sha256", sa.String(length=64), nullable=True))
+    op.add_column("dataset_versions", sa.Column("source_metadata", sa.JSON(), nullable=True))
+    op.add_column("dataset_versions", sa.Column("provenance_metadata", sa.JSON(), nullable=True))
+
+    op.create_index(
+        "ix_datasets_project_temporal_series_observed",
+        "datasets",
+        ["project_id", "temporal_series_key", "observed_at"],
+    )
+    op.create_index("ix_dataset_versions_dataset_version", "dataset_versions", ["dataset_id", "version"], unique=True)
+    op.create_index(
+        "ix_vector_features_dataset_source_feature",
+        "vector_features",
+        ["dataset_id", "source_feature_id"],
+    )
+    op.create_check_constraint(
+        "ck_datasets_temporal_valid_range",
+        "datasets",
+        "valid_to IS NULL OR valid_from IS NULL OR valid_to >= valid_from",
+    )
+    op.create_check_constraint(
+        "ck_dataset_versions_temporal_valid_range",
+        "dataset_versions",
+        "valid_to IS NULL OR valid_from IS NULL OR valid_to >= valid_from",
+    )
+
+
+def downgrade() -> None:
+    op.drop_constraint("ck_dataset_versions_temporal_valid_range", "dataset_versions", type_="check")
+    op.drop_constraint("ck_datasets_temporal_valid_range", "datasets", type_="check")
+    op.drop_index("ix_vector_features_dataset_source_feature", table_name="vector_features")
+    op.drop_index("ix_dataset_versions_dataset_version", table_name="dataset_versions")
+    op.drop_index("ix_datasets_project_temporal_series_observed", table_name="datasets")
+
+    op.drop_column("dataset_versions", "provenance_metadata")
+    op.drop_column("dataset_versions", "source_metadata")
+    op.drop_column("dataset_versions", "checksum_sha256")
+    op.drop_column("dataset_versions", "valid_to")
+    op.drop_column("dataset_versions", "valid_from")
+    op.drop_column("dataset_versions", "observed_at")
+    op.drop_column("dataset_versions", "source_version")
+
+    op.drop_column("datasets", "source_version")
+    op.drop_column("datasets", "temporal_granularity")
+    op.drop_column("datasets", "valid_to")
+    op.drop_column("datasets", "valid_from")
+    op.drop_column("datasets", "observed_at")
+    op.drop_column("datasets", "temporal_series_key")