Files
RAGcore-Public/alembic/versions/0009_content_publication.py
T

158 lines
5.6 KiB
Python

"""Create content documents, versions, artifacts and chunks."""
from __future__ import annotations
import sqlalchemy as sa
from alembic import op
revision = "0009_content_publication"
down_revision = "0008_jobs"
branch_labels = None
depends_on = None
def upgrade() -> None:
op.execute(sa.text("CREATE SCHEMA IF NOT EXISTS content"))
op.create_table(
"documents",
sa.Column("id", sa.Uuid(), primary_key=True),
sa.Column("workspace_id", sa.Uuid(), nullable=False),
sa.Column("source_key", sa.String(256), nullable=False),
sa.Column("external_id", sa.String(512), nullable=False),
sa.Column("title", sa.String(512), nullable=False),
sa.Column(
"source_deleted", sa.Boolean(), nullable=False, server_default=sa.false()
),
sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
sa.Column("updated_at", sa.DateTime(timezone=True), nullable=False),
sa.UniqueConstraint(
"workspace_id",
"source_key",
"external_id",
name="uq_content_document_source",
),
schema="content",
)
op.create_index(
"ix_content_documents_searchable",
"documents",
["workspace_id", "source_deleted"],
schema="content",
)
op.create_table(
"document_versions",
sa.Column("id", sa.Uuid(), primary_key=True),
sa.Column("document_id", sa.Uuid(), nullable=False),
sa.Column("source_fingerprint", sa.String(64), nullable=False),
sa.Column("content_hash", sa.String(64), nullable=False),
sa.Column("parser_version", sa.String(128), nullable=False),
sa.Column("chunk_profile_version", sa.String(128), nullable=False),
sa.Column("state", sa.String(32), nullable=False),
sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
sa.Column("updated_at", sa.DateTime(timezone=True), nullable=False),
sa.ForeignKeyConstraint(
["document_id"], ["content.documents.id"], ondelete="CASCADE"
),
sa.UniqueConstraint(
"document_id",
"source_fingerprint",
"content_hash",
"parser_version",
"chunk_profile_version",
name="uq_content_version_identity",
),
schema="content",
)
op.create_index(
"uq_content_one_published_version",
"document_versions",
["document_id"],
unique=True,
schema="content",
postgresql_where=sa.text("state = 'published'"),
)
op.create_index(
"ix_content_versions_document_state",
"document_versions",
["document_id", "state"],
schema="content",
)
op.create_table(
"artifacts",
sa.Column("id", sa.Uuid(), primary_key=True),
sa.Column("document_version_id", sa.Uuid(), nullable=False),
sa.Column("kind", sa.String(64), nullable=False),
sa.Column("content_hash", sa.String(64), nullable=False),
sa.Column("storage_key", sa.String(512), nullable=False),
sa.Column("size_bytes", sa.Integer(), nullable=False),
sa.Column("metadata", sa.JSON(), nullable=False),
sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
sa.ForeignKeyConstraint(
["document_version_id"],
["content.document_versions.id"],
ondelete="CASCADE",
),
sa.UniqueConstraint(
"document_version_id", "kind", name="uq_content_artifact_kind"
),
schema="content",
)
op.create_table(
"chunks",
sa.Column("id", sa.Uuid(), primary_key=True),
sa.Column("document_version_id", sa.Uuid(), nullable=False),
sa.Column("parent_id", sa.Uuid(), nullable=True),
sa.Column("structure_path", sa.String(512), nullable=False),
sa.Column("content_hash", sa.String(64), nullable=False),
sa.Column("chunk_profile_version", sa.String(128), nullable=False),
sa.Column("content", sa.Text(), nullable=False),
sa.Column("token_count", sa.Integer(), nullable=False),
sa.Column("metadata", sa.JSON(), nullable=False),
sa.ForeignKeyConstraint(
["document_version_id"],
["content.document_versions.id"],
ondelete="CASCADE",
),
sa.ForeignKeyConstraint(
["parent_id"], ["content.chunks.id"], ondelete="RESTRICT"
),
sa.UniqueConstraint(
"document_version_id",
"structure_path",
"content_hash",
"chunk_profile_version",
name="uq_content_chunk_identity",
),
schema="content",
)
op.create_index(
"ix_content_chunks_version_profile",
"chunks",
["document_version_id", "chunk_profile_version"],
schema="content",
)
def downgrade() -> None:
op.drop_index(
"ix_content_chunks_version_profile", table_name="chunks", schema="content"
)
op.drop_table("chunks", schema="content")
op.drop_table("artifacts", schema="content")
op.drop_index(
"ix_content_versions_document_state",
table_name="document_versions",
schema="content",
)
op.drop_index(
"uq_content_one_published_version",
table_name="document_versions",
schema="content",
)
op.drop_table("document_versions", schema="content")
op.drop_index(
"ix_content_documents_searchable", table_name="documents", schema="content"
)
op.drop_table("documents", schema="content")
op.execute(sa.text("DROP SCHEMA IF EXISTS content"))