mirror of
https://github.com/basicmachines-co/basic-memory
synced 2026-06-21 13:47:35 +00:00
phernandez
89 lines
3.3 KiB
Python
89 lines
3.3 KiB
Python
"""Search models and tables."""
|
|
|
|
from sqlalchemy import DDL, Column, Integer, String, DateTime, Text, ForeignKey
|
|
from sqlalchemy.dialects.postgresql import JSONB
|
|
from sqlalchemy.types import JSON
|
|
|
|
from basic_memory.models.base import Base
|
|
|
|
|
|
class SearchIndex(Base):
|
|
"""Search index table for Postgres only.
|
|
|
|
For SQLite: This model is skipped; FTS5 virtual table is created via DDL instead.
|
|
For Postgres: This is the actual table structure with tsvector support.
|
|
"""
|
|
|
|
__tablename__ = "search_index"
|
|
|
|
# Primary key (rowid in SQLite FTS5, explicit id in Postgres)
|
|
id = Column(Integer, primary_key=True, autoincrement=True)
|
|
|
|
# Core searchable fields
|
|
title = Column(Text, nullable=True)
|
|
content_stems = Column(Text, nullable=True)
|
|
content_snippet = Column(Text, nullable=True)
|
|
permalink = Column(String(255), nullable=True, index=True)
|
|
file_path = Column(Text, nullable=True)
|
|
type = Column(String(50), nullable=True)
|
|
|
|
# Project context
|
|
project_id = Column(Integer, nullable=True, index=True)
|
|
|
|
# Relation fields
|
|
from_id = Column(Integer, nullable=True)
|
|
to_id = Column(Integer, nullable=True)
|
|
relation_type = Column(String(100), nullable=True)
|
|
|
|
# Observation fields
|
|
# Note: FK with CASCADE only applies to Postgres. SQLite uses FTS5 virtual tables
|
|
# which don't support foreign keys, so cascade delete is handled explicitly there.
|
|
entity_id = Column(Integer, ForeignKey("entity.id", ondelete="CASCADE"), nullable=True)
|
|
category = Column(String(100), nullable=True)
|
|
|
|
# Common fields
|
|
# Use JSONB for Postgres, JSON for SQLite
|
|
# Note: 'metadata' is a reserved name in SQLAlchemy, so we use 'metadata_' and map to 'metadata'
|
|
metadata_ = Column("metadata", JSON().with_variant(JSONB(), "postgresql"), nullable=True)
|
|
created_at = Column(DateTime(timezone=True), nullable=True)
|
|
updated_at = Column(DateTime(timezone=True), nullable=True)
|
|
|
|
# Note: textsearchable_index_col (tsvector) will be added by migration for Postgres only
|
|
|
|
|
|
# Define FTS5 virtual table creation for SQLite only
|
|
# This DDL is executed separately for SQLite databases
|
|
CREATE_SEARCH_INDEX = DDL("""
|
|
CREATE VIRTUAL TABLE IF NOT EXISTS search_index USING fts5(
|
|
-- Core entity fields
|
|
id UNINDEXED, -- Row ID
|
|
title, -- Title for searching
|
|
content_stems, -- Main searchable content split into stems
|
|
content_snippet, -- File content snippet for display
|
|
permalink, -- Stable identifier (now indexed for path search)
|
|
file_path UNINDEXED, -- Physical location
|
|
type UNINDEXED, -- entity/relation/observation
|
|
|
|
-- Project context
|
|
project_id UNINDEXED, -- Project identifier
|
|
|
|
-- Relation fields
|
|
from_id UNINDEXED, -- Source entity
|
|
to_id UNINDEXED, -- Target entity
|
|
relation_type UNINDEXED, -- Type of relation
|
|
|
|
-- Observation fields
|
|
entity_id UNINDEXED, -- Parent entity
|
|
category UNINDEXED, -- Observation category
|
|
|
|
-- Common fields
|
|
metadata UNINDEXED, -- JSON metadata
|
|
created_at UNINDEXED, -- Creation timestamp
|
|
updated_at UNINDEXED, -- Last update
|
|
|
|
-- Configuration
|
|
tokenize='unicode61 tokenchars 0x2F', -- Hex code for /
|
|
prefix='1,2,3,4' -- Support longer prefixes for paths
|
|
);
|
|
""")
|