adding database_cli.py and alembic

alembic/data_science_dev/versions/2026_03_24-adding_failedingestion_2f43120e3ffc.py

@@ -0,0 +1,50 @@
+"""adding FailedIngestion.
+
+Revision ID: 2f43120e3ffc
+Revises: f99be864fe69
+Create Date: 2026-03-24 23:46:17.277897
+
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import sqlalchemy as sa
+from alembic import op
+
+from python.orm import DataScienceDevBase
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+# revision identifiers, used by Alembic.
+revision: str = "2f43120e3ffc"
+down_revision: str | None = "f99be864fe69"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+schema = DataScienceDevBase.schema_name
+
+
+def upgrade() -> None:
+    """Upgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table(
+        "failed_ingestion",
+        sa.Column("raw_line", sa.Text(), nullable=False),
+        sa.Column("error", sa.Text(), nullable=False),
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("created", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.Column("updated", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_failed_ingestion")),
+        schema=schema,
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    """Downgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_table("failed_ingestion", schema=schema)
+    # ### end Alembic commands ###

alembic/data_science_dev/versions/2026_03_25-attach_partitions_to_posts.py

@@ -0,0 +1,72 @@
+"""Attach all partition tables to the posts parent table.
+
+Alembic autogenerate creates partition tables as standalone tables but does not
+emit the ALTER TABLE ... ATTACH PARTITION statements needed for PostgreSQL to
+route inserts to the correct partition.
+
+Revision ID: a1b2c3d4e5f6
+Revises: 605b1794838f
+Create Date: 2026-03-25 10:00:00.000000
+
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+from alembic import op
+from sqlalchemy import text
+
+from python.orm import DataScienceDevBase
+from python.orm.data_science_dev.posts.partitions import (
+    PARTITION_END_YEAR,
+    PARTITION_START_YEAR,
+    iso_weeks_in_year,
+    week_bounds,
+)
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+# revision identifiers, used by Alembic.
+revision: str = "a1b2c3d4e5f6"
+down_revision: str | None = "605b1794838f"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+schema = DataScienceDevBase.schema_name
+
+ALREADY_ATTACHED_QUERY = text("""
+    SELECT inhrelid::regclass::text
+    FROM pg_inherits
+    WHERE inhparent = :parent::regclass
+""")
+
+
+def upgrade() -> None:
+    """Attach all weekly partition tables to the posts parent table."""
+    connection = op.get_bind()
+    already_attached = {row[0] for row in connection.execute(ALREADY_ATTACHED_QUERY, {"parent": f"{schema}.posts"})}
+
+    for year in range(PARTITION_START_YEAR, PARTITION_END_YEAR + 1):
+        for week in range(1, iso_weeks_in_year(year) + 1):
+            table_name = f"posts_{year}_{week:02d}"
+            qualified_name = f"{schema}.{table_name}"
+            if qualified_name in already_attached:
+                continue
+            start, end = week_bounds(year, week)
+            start_str = start.strftime("%Y-%m-%d %H:%M:%S")
+            end_str = end.strftime("%Y-%m-%d %H:%M:%S")
+            op.execute(
+                f"ALTER TABLE {schema}.posts "
+                f"ATTACH PARTITION {qualified_name} "
+                f"FOR VALUES FROM ('{start_str}') TO ('{end_str}')"
+            )
+
+
+def downgrade() -> None:
+    """Detach all weekly partition tables from the posts parent table."""
+    for year in range(PARTITION_START_YEAR, PARTITION_END_YEAR + 1):
+        for week in range(1, iso_weeks_in_year(year) + 1):
+            table_name = f"posts_{year}_{week:02d}"
+            op.execute(f"ALTER TABLE {schema}.posts DETACH PARTITION {schema}.{table_name}")

alembic/data_science_dev/versions/2026_03_27-adding_congress_data_83bfc8af92d8.py

@@ -0,0 +1,153 @@
+"""adding congress data.
+
+Revision ID: 83bfc8af92d8
+Revises: a1b2c3d4e5f6
+Create Date: 2026-03-27 10:43:02.324510
+
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import sqlalchemy as sa
+from alembic import op
+
+from python.orm import DataScienceDevBase
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+# revision identifiers, used by Alembic.
+revision: str = "83bfc8af92d8"
+down_revision: str | None = "a1b2c3d4e5f6"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+schema = DataScienceDevBase.schema_name
+
+
+def upgrade() -> None:
+    """Upgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table(
+        "bill",
+        sa.Column("congress", sa.Integer(), nullable=False),
+        sa.Column("bill_type", sa.String(), nullable=False),
+        sa.Column("number", sa.Integer(), nullable=False),
+        sa.Column("title", sa.String(), nullable=True),
+        sa.Column("title_short", sa.String(), nullable=True),
+        sa.Column("official_title", sa.String(), nullable=True),
+        sa.Column("status", sa.String(), nullable=True),
+        sa.Column("status_at", sa.Date(), nullable=True),
+        sa.Column("sponsor_bioguide_id", sa.String(), nullable=True),
+        sa.Column("subjects_top_term", sa.String(), nullable=True),
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("created", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.Column("updated", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_bill")),
+        sa.UniqueConstraint("congress", "bill_type", "number", name="uq_bill_congress_type_number"),
+        schema=schema,
+    )
+    op.create_index("ix_bill_congress", "bill", ["congress"], unique=False, schema=schema)
+    op.create_table(
+        "legislator",
+        sa.Column("bioguide_id", sa.Text(), nullable=False),
+        sa.Column("thomas_id", sa.String(), nullable=True),
+        sa.Column("lis_id", sa.String(), nullable=True),
+        sa.Column("govtrack_id", sa.Integer(), nullable=True),
+        sa.Column("opensecrets_id", sa.String(), nullable=True),
+        sa.Column("fec_ids", sa.String(), nullable=True),
+        sa.Column("first_name", sa.String(), nullable=False),
+        sa.Column("last_name", sa.String(), nullable=False),
+        sa.Column("official_full_name", sa.String(), nullable=True),
+        sa.Column("nickname", sa.String(), nullable=True),
+        sa.Column("birthday", sa.Date(), nullable=True),
+        sa.Column("gender", sa.String(), nullable=True),
+        sa.Column("current_party", sa.String(), nullable=True),
+        sa.Column("current_state", sa.String(), nullable=True),
+        sa.Column("current_district", sa.Integer(), nullable=True),
+        sa.Column("current_chamber", sa.String(), nullable=True),
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("created", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.Column("updated", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_legislator")),
+        schema=schema,
+    )
+    op.create_index(op.f("ix_legislator_bioguide_id"), "legislator", ["bioguide_id"], unique=True, schema=schema)
+    op.create_table(
+        "bill_text",
+        sa.Column("bill_id", sa.Integer(), nullable=False),
+        sa.Column("version_code", sa.String(), nullable=False),
+        sa.Column("version_name", sa.String(), nullable=True),
+        sa.Column("text_content", sa.String(), nullable=True),
+        sa.Column("date", sa.Date(), nullable=True),
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("created", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.Column("updated", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["bill_id"], [f"{schema}.bill.id"], name=op.f("fk_bill_text_bill_id_bill"), ondelete="CASCADE"
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_bill_text")),
+        sa.UniqueConstraint("bill_id", "version_code", name="uq_bill_text_bill_id_version_code"),
+        schema=schema,
+    )
+    op.create_table(
+        "vote",
+        sa.Column("congress", sa.Integer(), nullable=False),
+        sa.Column("chamber", sa.String(), nullable=False),
+        sa.Column("session", sa.Integer(), nullable=False),
+        sa.Column("number", sa.Integer(), nullable=False),
+        sa.Column("vote_type", sa.String(), nullable=True),
+        sa.Column("question", sa.String(), nullable=True),
+        sa.Column("result", sa.String(), nullable=True),
+        sa.Column("result_text", sa.String(), nullable=True),
+        sa.Column("vote_date", sa.Date(), nullable=False),
+        sa.Column("yea_count", sa.Integer(), nullable=True),
+        sa.Column("nay_count", sa.Integer(), nullable=True),
+        sa.Column("not_voting_count", sa.Integer(), nullable=True),
+        sa.Column("present_count", sa.Integer(), nullable=True),
+        sa.Column("bill_id", sa.Integer(), nullable=True),
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("created", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.Column("updated", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.ForeignKeyConstraint(["bill_id"], [f"{schema}.bill.id"], name=op.f("fk_vote_bill_id_bill")),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_vote")),
+        sa.UniqueConstraint("congress", "chamber", "session", "number", name="uq_vote_congress_chamber_session_number"),
+        schema=schema,
+    )
+    op.create_index("ix_vote_congress_chamber", "vote", ["congress", "chamber"], unique=False, schema=schema)
+    op.create_index("ix_vote_date", "vote", ["vote_date"], unique=False, schema=schema)
+    op.create_table(
+        "vote_record",
+        sa.Column("vote_id", sa.Integer(), nullable=False),
+        sa.Column("legislator_id", sa.Integer(), nullable=False),
+        sa.Column("position", sa.String(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["legislator_id"],
+            [f"{schema}.legislator.id"],
+            name=op.f("fk_vote_record_legislator_id_legislator"),
+            ondelete="CASCADE",
+        ),
+        sa.ForeignKeyConstraint(
+            ["vote_id"], [f"{schema}.vote.id"], name=op.f("fk_vote_record_vote_id_vote"), ondelete="CASCADE"
+        ),
+        sa.PrimaryKeyConstraint("vote_id", "legislator_id", name=op.f("pk_vote_record")),
+        schema=schema,
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    """Downgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_table("vote_record", schema=schema)
+    op.drop_index("ix_vote_date", table_name="vote", schema=schema)
+    op.drop_index("ix_vote_congress_chamber", table_name="vote", schema=schema)
+    op.drop_table("vote", schema=schema)
+    op.drop_table("bill_text", schema=schema)
+    op.drop_index(op.f("ix_legislator_bioguide_id"), table_name="legislator", schema=schema)
+    op.drop_table("legislator", schema=schema)
+    op.drop_index("ix_bill_congress", table_name="bill", schema=schema)
+    op.drop_table("bill", schema=schema)
+    # ### end Alembic commands ###

alembic/data_science_dev/versions/2026_03_29-adding_legislatorsocialmedia_5cd7eee3549d.py

@@ -0,0 +1,58 @@
+"""adding LegislatorSocialMedia.
+
+Revision ID: 5cd7eee3549d
+Revises: 83bfc8af92d8
+Create Date: 2026-03-29 11:53:44.224799
+
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import sqlalchemy as sa
+from alembic import op
+
+from python.orm import DataScienceDevBase
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+# revision identifiers, used by Alembic.
+revision: str = "5cd7eee3549d"
+down_revision: str | None = "83bfc8af92d8"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+schema = DataScienceDevBase.schema_name
+
+
+def upgrade() -> None:
+    """Upgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table(
+        "legislator_social_media",
+        sa.Column("legislator_id", sa.Integer(), nullable=False),
+        sa.Column("platform", sa.String(), nullable=False),
+        sa.Column("account_name", sa.String(), nullable=False),
+        sa.Column("url", sa.String(), nullable=True),
+        sa.Column("source", sa.String(), nullable=False),
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("created", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.Column("updated", sa.DateTime(timezone=True), server_default=sa.text("now()"), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["legislator_id"],
+            [f"{schema}.legislator.id"],
+            name=op.f("fk_legislator_social_media_legislator_id_legislator"),
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_legislator_social_media")),
+        schema=schema,
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    """Downgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_table("legislator_social_media", schema=schema)
+    # ### end Alembic commands ###