added ebook embedding to orm

2026-06-11 18:08:31 -04:00
parent 57e98dc667
commit eee1b8f8e0
3 changed files with 346 additions and 0 deletions
@@ -0,0 +1,200 @@
+"""add ebook search tables.
+
+Revision ID: 2db132cace1a
+Revises: b3c60cc5beb5
+Create Date: 2026-06-10 22:10:54.379159
+
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import pgvector
+import sqlalchemy as sa
+from alembic import op
+
+from python.orm import RichieBase
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+# revision identifiers, used by Alembic.
+revision: str = "2db132cace1a"
+down_revision: str | None = "b3c60cc5beb5"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+schema = RichieBase.schema_name
+
+
+def upgrade() -> None:
+    """Upgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table(
+        "ebook_embedding_model",
+        sa.Column("name", sa.String(), nullable=False),
+        sa.Column("dimension", sa.Integer(), nullable=False),
+        sa.Column("is_default", sa.Boolean(), 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_ebook_embedding_model")),
+        sa.UniqueConstraint("name", name=op.f("uq_ebook_embedding_model_name")),
+        schema=schema,
+    )
+    op.create_table(
+        "ebook_source",
+        sa.Column("title", sa.String(), nullable=False),
+        sa.Column("author", sa.String(), nullable=True),
+        sa.Column("language", sa.String(), nullable=True),
+        sa.Column("publisher", sa.String(), nullable=True),
+        sa.Column("identifier", sa.String(), nullable=True),
+        sa.Column("file_path", sa.String(), nullable=False),
+        sa.Column("file_sha256", sa.String(length=64), nullable=False),
+        sa.Column("file_mtime", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("file_size", sa.BigInteger(), 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_ebook_source")),
+        sa.UniqueConstraint("file_path", name=op.f("uq_ebook_source_file_path")),
+        sa.UniqueConstraint("file_sha256", name=op.f("uq_ebook_source_file_sha256")),
+        schema=schema,
+    )
+    op.create_table(
+        "ebook_chapter",
+        sa.Column("source_id", sa.Integer(), nullable=False),
+        sa.Column("spine_index", sa.Integer(), nullable=False),
+        sa.Column("title", sa.String(), nullable=True),
+        sa.Column("href", 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.ForeignKeyConstraint(
+            ["source_id"],
+            [f"{schema}.ebook_source.id"],
+            name=op.f("fk_ebook_chapter_source_id_ebook_source"),
+            ondelete="CASCADE",
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_ebook_chapter")),
+        sa.UniqueConstraint("source_id", "spine_index", name=op.f("uq_ebook_chapter_source_id")),
+        schema=schema,
+    )
+    op.create_table(
+        "ebook_chunk",
+        sa.Column("source_id", sa.Integer(), nullable=False),
+        sa.Column("chapter_id", sa.Integer(), nullable=True),
+        sa.Column("chunk_index", sa.Integer(), nullable=False),
+        sa.Column("text", sa.String(), nullable=False),
+        sa.Column("token_start", sa.Integer(), nullable=False),
+        sa.Column("token_count", sa.Integer(), nullable=False),
+        sa.Column("page_label", sa.String(), nullable=True),
+        sa.Column("content_sha256", sa.String(length=64), nullable=False),
+        sa.Column("search_text", sa.String(), nullable=False),
+        sa.Column("id", sa.BigInteger(), 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(
+            ["chapter_id"],
+            [f"{schema}.ebook_chapter.id"],
+            name=op.f("fk_ebook_chunk_chapter_id_ebook_chapter"),
+            ondelete="SET NULL",
+        ),
+        sa.ForeignKeyConstraint(
+            ["source_id"],
+            [f"{schema}.ebook_source.id"],
+            name=op.f("fk_ebook_chunk_source_id_ebook_source"),
+            ondelete="CASCADE",
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_ebook_chunk")),
+        sa.UniqueConstraint("source_id", "chunk_index", name="uq_ebook_chunk_source_id_chunk_index"),
+        sa.UniqueConstraint("source_id", "content_sha256", name="uq_ebook_chunk_source_id_content_sha256"),
+        schema=schema,
+    )
+    op.create_table(
+        "ebook_chunk_embedding_1024",
+        sa.Column("chunk_id", sa.BigInteger(), nullable=False),
+        sa.Column("model_id", sa.Integer(), nullable=False),
+        sa.Column("embedding", pgvector.sqlalchemy.vector.VECTOR(dim=1024), nullable=False),
+        sa.Column("id", sa.BigInteger(), 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(
+            ["chunk_id"],
+            [f"{schema}.ebook_chunk.id"],
+            name=op.f("fk_ebook_chunk_embedding_1024_chunk_id_ebook_chunk"),
+            ondelete="CASCADE",
+        ),
+        sa.ForeignKeyConstraint(
+            ["model_id"],
+            [f"{schema}.ebook_embedding_model.id"],
+            name=op.f("fk_ebook_chunk_embedding_1024_model_id_ebook_embedding_model"),
+            ondelete="CASCADE",
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_ebook_chunk_embedding_1024")),
+        sa.UniqueConstraint("chunk_id", "model_id", name=op.f("uq_ebook_chunk_embedding_1024_chunk_id")),
+        schema=schema,
+    )
+    op.create_table(
+        "ebook_chunk_embedding_2560",
+        sa.Column("chunk_id", sa.BigInteger(), nullable=False),
+        sa.Column("model_id", sa.Integer(), nullable=False),
+        sa.Column("embedding", pgvector.sqlalchemy.vector.VECTOR(dim=2560), nullable=False),
+        sa.Column("id", sa.BigInteger(), 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(
+            ["chunk_id"],
+            [f"{schema}.ebook_chunk.id"],
+            name=op.f("fk_ebook_chunk_embedding_2560_chunk_id_ebook_chunk"),
+            ondelete="CASCADE",
+        ),
+        sa.ForeignKeyConstraint(
+            ["model_id"],
+            [f"{schema}.ebook_embedding_model.id"],
+            name=op.f("fk_ebook_chunk_embedding_2560_model_id_ebook_embedding_model"),
+            ondelete="CASCADE",
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_ebook_chunk_embedding_2560")),
+        sa.UniqueConstraint("chunk_id", "model_id", name=op.f("uq_ebook_chunk_embedding_2560_chunk_id")),
+        schema=schema,
+    )
+    op.create_table(
+        "ebook_chunk_embedding_4096",
+        sa.Column("chunk_id", sa.BigInteger(), nullable=False),
+        sa.Column("model_id", sa.Integer(), nullable=False),
+        sa.Column("embedding", pgvector.sqlalchemy.vector.VECTOR(dim=4096), nullable=False),
+        sa.Column("id", sa.BigInteger(), 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(
+            ["chunk_id"],
+            [f"{schema}.ebook_chunk.id"],
+            name=op.f("fk_ebook_chunk_embedding_4096_chunk_id_ebook_chunk"),
+            ondelete="CASCADE",
+        ),
+        sa.ForeignKeyConstraint(
+            ["model_id"],
+            [f"{schema}.ebook_embedding_model.id"],
+            name=op.f("fk_ebook_chunk_embedding_4096_model_id_ebook_embedding_model"),
+            ondelete="CASCADE",
+        ),
+        sa.PrimaryKeyConstraint("id", name=op.f("pk_ebook_chunk_embedding_4096")),
+        sa.UniqueConstraint("chunk_id", "model_id", name=op.f("uq_ebook_chunk_embedding_4096_chunk_id")),
+        schema=schema,
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    """Downgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_table("ebook_chunk_embedding_4096", schema=schema)
+    op.drop_table("ebook_chunk_embedding_2560", schema=schema)
+    op.drop_table("ebook_chunk_embedding_1024", schema=schema)
+    op.drop_table("ebook_chunk", schema=schema)
+    op.drop_table("ebook_chapter", schema=schema)
+    op.drop_table("ebook_source", schema=schema)
+    op.drop_table("ebook_embedding_model", schema=schema)
+    # ### end Alembic commands ###
@@ -11,6 +11,15 @@ from python.orm.richie.contact import (
    Need,
    RelationshipType,
 )
+from python.orm.richie.ebook import (
+    EbookChapter,
+    EbookChunk,
+    EbookChunkEmbedding1024,
+    EbookChunkEmbedding2560,
+    EbookChunkEmbedding4096,
+    EbookEmbeddingModel,
+    EbookSource,
+)

 __all__ = [
    "Audiobook",
@@ -19,6 +28,13 @@ __all__ = [
    "Contact",
    "ContactNeed",
    "ContactRelationship",
+    "EbookChapter",
+    "EbookChunk",
+    "EbookChunkEmbedding1024",
+    "EbookChunkEmbedding2560",
+    "EbookChunkEmbedding4096",
+    "EbookEmbeddingModel",
+    "EbookSource",
    "Need",
    "RelationshipType",
    "RichieBase",
@@ -0,0 +1,130 @@
+"""EPUB search models."""
+
+from __future__ import annotations
+
+from datetime import datetime
+
+from pgvector.sqlalchemy import Vector
+from sqlalchemy import BigInteger, Boolean, DateTime, ForeignKey, String, UniqueConstraint
+from sqlalchemy.orm import Mapped, mapped_column, relationship
+
+from python.orm.richie.base import TableBase, TableBaseBig
+
+
+class EbookSource(TableBase):
+    """One indexed EPUB file."""
+
+    __tablename__ = "ebook_source"
+    __table_args__ = (
+        UniqueConstraint("file_path"),
+        UniqueConstraint("file_sha256"),
+    )
+
+    title: Mapped[str]
+    author: Mapped[str | None]
+    language: Mapped[str | None]
+    publisher: Mapped[str | None]
+    identifier: Mapped[str | None]
+    file_path: Mapped[str]
+    file_sha256: Mapped[str] = mapped_column(String(64))
+    file_mtime: Mapped[datetime] = mapped_column(DateTime(timezone=True))
+    file_size: Mapped[int] = mapped_column(BigInteger)
+
+    chapters: Mapped[list[EbookChapter]] = relationship(
+        "EbookChapter",
+        back_populates="source",
+        cascade="all, delete-orphan",
+        passive_deletes=True,
+    )
+    chunks: Mapped[list[EbookChunk]] = relationship(
+        "EbookChunk",
+        back_populates="source",
+        cascade="all, delete-orphan",
+        passive_deletes=True,
+    )
+
+
+class EbookChapter(TableBase):
+    """A chapter or spine document inside an EPUB."""
+
+    __tablename__ = "ebook_chapter"
+    __table_args__ = (UniqueConstraint("source_id", "spine_index"),)
+
+    source_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_source.id", ondelete="CASCADE"))
+    spine_index: Mapped[int]
+    title: Mapped[str | None]
+    href: Mapped[str | None]
+
+    source: Mapped[EbookSource] = relationship("EbookSource", back_populates="chapters")
+    chunks: Mapped[list[EbookChunk]] = relationship(
+        "EbookChunk",
+        back_populates="chapter",
+        cascade="all, delete-orphan",
+        passive_deletes=True,
+    )
+
+
+class EbookChunk(TableBaseBig):
+    """A searchable text chunk."""
+
+    __tablename__ = "ebook_chunk"
+    __table_args__ = (
+        UniqueConstraint("source_id", "chunk_index", name="uq_ebook_chunk_source_id_chunk_index"),
+        UniqueConstraint("source_id", "content_sha256", name="uq_ebook_chunk_source_id_content_sha256"),
+    )
+
+    source_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_source.id", ondelete="CASCADE"))
+    chapter_id: Mapped[int | None] = mapped_column(ForeignKey("main.ebook_chapter.id", ondelete="SET NULL"))
+    chunk_index: Mapped[int]
+    text: Mapped[str]
+    token_start: Mapped[int]
+    token_count: Mapped[int]
+    page_label: Mapped[str | None]
+    content_sha256: Mapped[str] = mapped_column(String(64))
+    search_text: Mapped[str]
+
+    source: Mapped[EbookSource] = relationship("EbookSource", back_populates="chunks")
+    chapter: Mapped[EbookChapter | None] = relationship("EbookChapter", back_populates="chunks")
+
+
+class EbookEmbeddingModel(TableBase):
+    """A supported embedding model."""
+
+    __tablename__ = "ebook_embedding_model"
+
+    name: Mapped[str] = mapped_column(String, unique=True)
+    dimension: Mapped[int]
+    is_default: Mapped[bool] = mapped_column(Boolean, default=False)
+
+
+class EbookChunkEmbedding1024(TableBaseBig):
+    """1024-dimensional chunk embedding."""
+
+    __tablename__ = "ebook_chunk_embedding_1024"
+    __table_args__ = (UniqueConstraint("chunk_id", "model_id"),)
+
+    chunk_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_chunk.id", ondelete="CASCADE"))
+    model_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_embedding_model.id", ondelete="CASCADE"))
+    embedding: Mapped[list[float]] = mapped_column(Vector(1024))
+
+
+class EbookChunkEmbedding2560(TableBaseBig):
+    """2560-dimensional chunk embedding."""
+
+    __tablename__ = "ebook_chunk_embedding_2560"
+    __table_args__ = (UniqueConstraint("chunk_id", "model_id"),)
+
+    chunk_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_chunk.id", ondelete="CASCADE"))
+    model_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_embedding_model.id", ondelete="CASCADE"))
+    embedding: Mapped[list[float]] = mapped_column(Vector(2560))
+
+
+class EbookChunkEmbedding4096(TableBaseBig):
+    """4096-dimensional chunk embedding."""
+
+    __tablename__ = "ebook_chunk_embedding_4096"
+    __table_args__ = (UniqueConstraint("chunk_id", "model_id"),)
+
+    chunk_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_chunk.id", ondelete="CASCADE"))
+    model_id: Mapped[int] = mapped_column(ForeignKey("main.ebook_embedding_model.id", ondelete="CASCADE"))
+    embedding: Mapped[list[float]] = mapped_column(Vector(4096))