feat: persist cleaned question insights

ai_knowledge_base_v2/apps/backend/alembic/versions/0020_question_insight_persistence.py

@@ -0,0 +1,104 @@
+"""persist cleaned question insights
+
+Revision ID: 0020_question_insight_persistence
+Revises: 0019_periodic_reports
+Create Date: 2026-07-31 00:00:00.000000
+"""
+
+from __future__ import annotations
+
+from alembic import op
+import sqlalchemy as sa
+
+
+revision = "0020_question_insight_persistence"
+down_revision = "0019_periodic_reports"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "sys_question_insight_cleaned_question",
+        sa.Column(
+            "id",
+            sa.BigInteger().with_variant(sa.Integer(), "sqlite"),
+            primary_key=True,
+            autoincrement=True,
+        ),
+        sa.Column("message_id", sa.BigInteger(), nullable=False),
+        sa.Column("session_id", sa.BigInteger(), nullable=False),
+        sa.Column("user_id", sa.BigInteger(), nullable=False),
+        sa.Column("part_index", sa.Integer(), nullable=False),
+        sa.Column("cleaner_version", sa.String(length=20), nullable=False),
+        sa.Column("source_hash", sa.String(length=64), nullable=False),
+        sa.Column("cleaned_text", sa.Text(), nullable=False),
+        sa.Column("normalized_text", sa.Text(), nullable=False),
+        sa.Column("category", sa.String(length=50), nullable=False, server_default="other"),
+        sa.Column("tokens_json", sa.Text(), nullable=False),
+        sa.Column("accepted", sa.Integer(), nullable=False, server_default="1"),
+        sa.Column("filtered_reason", sa.String(length=100), nullable=True),
+        sa.Column("source_created_at", sa.DateTime(), nullable=False),
+        sa.Column("created_at", sa.DateTime(), nullable=False, server_default=sa.func.now()),
+        sa.Column("updated_at", sa.DateTime(), nullable=False, server_default=sa.func.now()),
+        sa.ForeignKeyConstraint(
+            ["message_id"],
+            ["sys_chat_message.id"],
+            ondelete="CASCADE",
+        ),
+        sa.UniqueConstraint(
+            "message_id",
+            "cleaner_version",
+            "part_index",
+            name="uq_question_insight_message_version_part",
+        ),
+    )
+    op.create_index(
+        "ix_sys_question_insight_cleaned_question_session_id",
+        "sys_question_insight_cleaned_question",
+        ["session_id"],
+    )
+    op.create_index(
+        "ix_sys_question_insight_cleaned_question_user_id",
+        "sys_question_insight_cleaned_question",
+        ["user_id"],
+    )
+    op.create_index(
+        "ix_question_insight_version_accepted_created",
+        "sys_question_insight_cleaned_question",
+        ["cleaner_version", "accepted", "source_created_at"],
+    )
+    op.create_index(
+        "ix_question_insight_category_created",
+        "sys_question_insight_cleaned_question",
+        ["category", "source_created_at"],
+    )
+    op.create_index(
+        "ix_question_insight_session_created",
+        "sys_question_insight_cleaned_question",
+        ["session_id", "source_created_at"],
+    )
+
+
+def downgrade() -> None:
+    op.drop_index(
+        "ix_question_insight_session_created",
+        table_name="sys_question_insight_cleaned_question",
+    )
+    op.drop_index(
+        "ix_question_insight_category_created",
+        table_name="sys_question_insight_cleaned_question",
+    )
+    op.drop_index(
+        "ix_question_insight_version_accepted_created",
+        table_name="sys_question_insight_cleaned_question",
+    )
+    op.drop_index(
+        "ix_sys_question_insight_cleaned_question_user_id",
+        table_name="sys_question_insight_cleaned_question",
+    )
+    op.drop_index(
+        "ix_sys_question_insight_cleaned_question_session_id",
+        table_name="sys_question_insight_cleaned_question",
+    )
+    op.drop_table("sys_question_insight_cleaned_question")