added a index for the VEctor DB

python/alembic/richie/versions/2026_06_13-add_1024_ebook_embedding_cosine_index_c460105682d2.py

@@ -0,0 +1,54 @@
+"""add 1024 ebook embedding cosine index.
+
+Revision ID: c460105682d2
+Revises: 2db132cace1a
+Create Date: 2026-06-13 19:53:45.680289
+
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+from alembic import op
+
+from python.orm import RichieBase
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+# revision identifiers, used by Alembic.
+revision: str = "c460105682d2"
+down_revision: str | None = "2db132cace1a"
+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_index(
+        "ix_ebook_chunk_embedding_1024_embedding_cosine",
+        "ebook_chunk_embedding_1024",
+        ["embedding"],
+        unique=False,
+        schema=schema,
+        postgresql_using="hnsw",
+        postgresql_ops={"embedding": "vector_cosine_ops"},
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    """Downgrade."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_index(
+        "ix_ebook_chunk_embedding_1024_embedding_cosine",
+        table_name="ebook_chunk_embedding_1024",
+        schema=schema,
+        postgresql_using="hnsw",
+        postgresql_ops={"embedding": "vector_cosine_ops"},
+    )
+    # ### end Alembic commands ###

python/orm/richie/ebook.py

@@ -5,7 +5,7 @@ from __future__ import annotations
 from datetime import datetime
 
 from pgvector.sqlalchemy import Vector
-from sqlalchemy import BigInteger, Boolean, DateTime, ForeignKey, String, UniqueConstraint
+from sqlalchemy import BigInteger, Boolean, DateTime, ForeignKey, Index, String, UniqueConstraint
 from sqlalchemy.orm import Mapped, mapped_column, relationship
 
 from python.orm.richie.base import TableBase, TableBaseBig
@@ -101,7 +101,15 @@ class EbookChunkEmbedding1024(TableBaseBig):
     """1024-dimensional chunk embedding."""
 
     __tablename__ = "ebook_chunk_embedding_1024"
-    __table_args__ = (UniqueConstraint("chunk_id", "model_id"),)
+    __table_args__ = (
+        UniqueConstraint("chunk_id", "model_id"),
+        Index(
+            "ix_ebook_chunk_embedding_1024_embedding_cosine",
+            "embedding",
+            postgresql_using="hnsw",
+            postgresql_ops={"embedding": "vector_cosine_ops"},
+        ),
+    )
 
     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"))

tests/test_ebook_search_core.py

@@ -38,7 +38,14 @@ from python.ebook_search.search import (
     search_ebooks,
 )
 from python.ebook_search.timing import RuntimeStep
-from python.orm.richie import EbookChapter, EbookChunk, EbookEmbeddingModel, EbookSource, RichieBase
+from python.orm.richie import (
+    EbookChapter,
+    EbookChunk,
+    EbookChunkEmbedding1024,
+    EbookEmbeddingModel,
+    EbookSource,
+    RichieBase,
+)
 
 
 def test_chunk_text_uses_overlap() -> None:
@@ -464,6 +471,15 @@ def test_ensure_embedding_models_registers_service_names() -> None:
     ]
 
 
+def test_1024_embedding_table_has_cosine_hnsw_index() -> None:
+    indexes = {index.name: index for index in EbookChunkEmbedding1024.__table__.indexes}
+    index = indexes["ix_ebook_chunk_embedding_1024_embedding_cosine"]
+
+    assert [column.name for column in index.columns] == ["embedding"]
+    assert index.dialect_options["postgresql"]["using"] == "hnsw"
+    assert index.dialect_options["postgresql"]["ops"] == {"embedding": "vector_cosine_ops"}
+
+
 def test_embedding_model_aliases_normalize_to_provider_names() -> None:
     assert normalize_embedding_model() == "qwen3-embedding-0.6b"