revert search service changes

src/basic_memory/models/search.py

@@ -5,11 +5,12 @@ from sqlalchemy import DDL
 # Define FTS5 virtual table creation
 CREATE_SEARCH_INDEX = DDL("""
 CREATE VIRTUAL TABLE IF NOT EXISTS search_index USING fts5(
-    content,           -- Searchable text content
-    permalink UNINDEXED, -- Link to entity/document (must be unique)
+    title,            -- Title for exact/fuzzy matching
+    content,          -- Additional searchable content
+    permalink UNINDEXED, -- Link to entity/document
     file_path UNINDEXED, -- Filesystem path
-    type UNINDEXED,    -- 'entity' or 'document'
-    metadata UNINDEXED, -- JSON with timestamps, types, etc.
+    type UNINDEXED,    -- entity type
+    metadata UNINDEXED, -- Additional metadata
     tokenize='porter unicode61' -- Enable stemming + unicode
 );
 """)

src/basic_memory/repository/search_repository.py

@@ -8,8 +8,8 @@ from sqlalchemy import text
 from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker
 
 from basic_memory import db
-from basic_memory.schemas.search import SearchQuery, SearchResult, SearchItemType
 from basic_memory.models.search import CREATE_SEARCH_INDEX
+from basic_memory.schemas.search import SearchQuery, SearchResult, SearchItemType
 
 
 class SearchRepository:
@@ -27,46 +27,13 @@ class SearchRepository:
     async def search(
         self, query: SearchQuery, context: Optional[List[str]] = None
     ) -> List[SearchResult]:
-        """Search across all indexed content using FTS5.
-        
-        Uses a three-tier matching strategy:
-        1. title:term matches (highest priority)
-        2. Exact term matches (medium priority)
-        3. Prefix matches (lowest priority)
-        """
+        """Search across all indexed content."""
         conditions = []
         params = {}
 
-        # Handle text search with fuzzy matching
+        # Handle text search
         if query.text:
-            # Prepare search terms
-            search_terms = query.text.lower().split()
-            params["search_terms"] = search_terms
-            
-            # Build match conditions in priority order
-            matches = []
-            
-            # 1. Title field matches (highest weight)
-            title_matches = [f'title:"{term}"' for term in search_terms]
-            if len(search_terms) > 1:
-                # Multi-word title match
-                title_phrase = " ".join(f'title:"{term}"' for term in search_terms)
-                title_matches.append(f'NEAR({title_phrase}, {len(search_terms)})')
-            matches.extend(title_matches)
-            
-            # 2. Exact word matches
-            matches.extend([f'"{term}"' for term in search_terms])
-            if len(search_terms) > 1:
-                # Multi-word proximity match
-                phrase = " ".join(f'"{term}"' for term in search_terms)
-                matches.append(f'NEAR({phrase}, {len(search_terms) * 2})')
-            
-            # 3. Prefix matches
-            matches.extend([f'{term}*' for term in search_terms])
-            
-            # Complete match expression
-            match_expr = " OR ".join(matches)
-            conditions.append(f"content MATCH '{match_expr}'")
+            conditions.append(f"content MATCH '{query.text}'")
 
         # Handle type filter
         if query.types:
@@ -87,26 +54,16 @@ class SearchRepository:
         # Build WHERE clause
         where_clause = " AND ".join(conditions) if conditions else "1=1"
 
-        # Build SQL query
         sql = f"""
-            WITH search_results AS (
-                SELECT 
-                    permalink,
-                    file_path,
-                    type,
-                    metadata,
-                    rank
-                FROM search_index 
-                WHERE {where_clause}
-            )
             SELECT 
                 permalink,
                 file_path,
                 type,
                 metadata,
-                rank as score
-            FROM search_results
-            ORDER BY rank ASC
+                bm25(search_index) as score
+            FROM search_index 
+            WHERE {where_clause}
+            ORDER BY score DESC
         """
 
         logger.debug(f"Search query: {sql}")
@@ -129,6 +86,7 @@ class SearchRepository:
 
     async def index_item(
         self,
+        title: str,
         content: str,
         permalink: str,
         file_path: str,
@@ -170,4 +128,4 @@ class SearchRepository:
                 text("DELETE FROM search_index WHERE permalink = :permalink"),
                 {"permalink": permalink},
             )
-            await session.commit()
+            await session.commit()

src/basic_memory/services/search_service.py

@@ -50,12 +50,11 @@ class SearchService:
     async def index_entity(
         self, entity: Entity, background_tasks: Optional[BackgroundTasks] = None
     ) -> None:
-        """Index an entity's core identity for fuzzy matching.
+        """Index an entity's content for search.
         
-        Core content:
-        - Original title (lowercase)
-        - Title with field marker (for weighted matching)
-        - Path components as discrete terms
+        Indexes:
+        - Title in dedicated field for better match control
+        - Path components in content field for findability
         """
         # Build searchable content
         content_parts = []
@@ -64,6 +63,7 @@ class SearchService:
         title = entity.title.lower()
         content_parts.extend([
             title,  # Base title
+            entity.summary or "",
             f"title:{title}",  # Field marker for targeted matching
             *title.split(),  # Individual words
         ])
@@ -88,6 +88,7 @@ class SearchService:
         if background_tasks:
             background_tasks.add_task(
                 self._do_index,
+                title=title,
                 content=content,
                 permalink=entity.permalink,
                 file_path=entity.file_path,
@@ -96,6 +97,7 @@ class SearchService:
             )
         else:
             await self._do_index(
+                title=title,
                 content=content,
                 permalink=entity.permalink,
                 file_path=entity.file_path,
@@ -104,10 +106,17 @@ class SearchService:
             )
 
     async def _do_index(
-        self, content: str, permalink: str, file_path: str, type: SearchItemType, metadata: dict
+        self,
+        title: str,
+        content: str,
+        permalink: str,
+        file_path: str,
+        type: SearchItemType,
+        metadata: dict,
     ) -> None:
         """Actually perform the indexing."""
         await self.repository.index_item(
+            title=title,
             content=content,
             permalink=permalink,
             file_path=file_path,