mcp-code-indexer 1.1.0__py3-none-any.whl → 1.1.2__py3-none-any.whl

mcp_code_indexer/database/database.py

@@ -234,6 +234,21 @@ class DatabaseManager:
             
             return projects
     
+    async def get_branch_file_counts(self, project_id: str) -> Dict[str, int]:
+        """Get file counts per branch for a project."""
+        async with self.get_connection() as db:
+            cursor = await db.execute(
+                """
+                SELECT branch, COUNT(*) as file_count 
+                FROM file_descriptions 
+                WHERE project_id = ? 
+                GROUP BY branch
+                """,
+                (project_id,)
+            )
+            rows = await cursor.fetchall()
+            return {row[0]: row[1] for row in rows}
+    
     # File description operations
     
     async def create_file_description(self, file_desc: FileDescription) -> None:
@@ -369,13 +384,13 @@ class DatabaseManager:
                     fd.branch,
                     fd.file_path,
                     fd.description,
-                    fts.rank
-                FROM file_descriptions_fts fts
-                JOIN file_descriptions fd ON fd.rowid = fts.rowid
-                WHERE fts MATCH ? 
+                    bm25(file_descriptions_fts) as rank
+                FROM file_descriptions_fts
+                JOIN file_descriptions fd ON fd.rowid = file_descriptions_fts.rowid
+                WHERE file_descriptions_fts MATCH ? 
                   AND fd.project_id = ? 
                   AND fd.branch = ?
-                ORDER BY fts.rank
+                ORDER BY bm25(file_descriptions_fts)
                 LIMIT ?
                 """,
                 (query, project_id, branch, max_results)

mcp_code_indexer/server/mcp_server.py

@@ -473,13 +473,65 @@ class MCPCodeIndexServer:
             await self.db_manager.update_project(project)
             logger.debug(f"Updated project metadata for {project.name}")
     
+    async def _resolve_branch(self, project_id: str, requested_branch: str) -> str:
+        """
+        Resolve the actual branch to use for operations.
+        
+        For new projects or branches with no existing files, uses the requested branch.
+        For existing projects, tries to find a consistent branch to avoid data fragmentation.
+        
+        Returns the resolved branch name.
+        """
+        try:
+            # Get all branches and their file counts for this project
+            branch_counts = await self.db_manager.get_branch_file_counts(project_id)
+            
+            # If no existing data, use the requested branch
+            if not branch_counts:
+                return requested_branch
+            
+            # If requested branch has files, use it
+            if requested_branch in branch_counts and branch_counts[requested_branch] > 0:
+                return requested_branch
+            
+            # Try common branch name variations to find existing data
+            common_variations = {
+                'main': ['master', 'develop', 'development', 'dev'],
+                'master': ['main', 'develop', 'development', 'dev'], 
+                'develop': ['development', 'main', 'master', 'dev'],
+                'development': ['develop', 'main', 'master', 'dev'],
+                'dev': ['develop', 'development', 'main', 'master']
+            }
+            
+            # Try variations of the requested branch
+            if requested_branch.lower() in common_variations:
+                for variation in common_variations[requested_branch.lower()]:
+                    if variation in branch_counts and branch_counts[variation] > 0:
+                        logger.info(f"Resolved branch '{requested_branch}' to existing branch '{variation}' with {branch_counts[variation]} files")
+                        return variation
+            
+            # If no variations found, check if we should use the main data branch
+            # (to avoid fragmenting data across multiple branches)
+            best_branch = max(branch_counts.items(), key=lambda x: x[1])
+            if best_branch[1] > 10:  # Only if there's substantial existing data
+                logger.info(f"Using primary branch '{best_branch[0]}' instead of '{requested_branch}' to avoid data fragmentation")
+                return best_branch[0]
+            
+            # Fall back to requested branch for new/small projects
+            return requested_branch
+            
+        except Exception as e:
+            logger.warning(f"Error resolving branch: {e}")
+            return requested_branch
+    
     async def _handle_get_file_description(self, arguments: Dict[str, Any]) -> Dict[str, Any]:
         """Handle get_file_description tool calls."""
         project_id = await self._get_or_create_project_id(arguments)
+        resolved_branch = await self._resolve_branch(project_id, arguments["branch"])
         
         file_desc = await self.db_manager.get_file_description(
             project_id=project_id,
-            branch=arguments["branch"],
+            branch=resolved_branch,
             file_path=arguments["filePath"]
         )
         
@@ -500,10 +552,11 @@ class MCPCodeIndexServer:
     async def _handle_update_file_description(self, arguments: Dict[str, Any]) -> Dict[str, Any]:
         """Handle update_file_description tool calls."""
         project_id = await self._get_or_create_project_id(arguments)
+        resolved_branch = await self._resolve_branch(project_id, arguments["branch"])
         
         file_desc = FileDescription(
             project_id=project_id,
-            branch=arguments["branch"],
+            branch=resolved_branch,
             file_path=arguments["filePath"],
             description=arguments["description"],
             file_hash=arguments.get("fileHash"),
@@ -523,11 +576,12 @@ class MCPCodeIndexServer:
     async def _handle_check_codebase_size(self, arguments: Dict[str, Any]) -> Dict[str, Any]:
         """Handle check_codebase_size tool calls."""
         project_id = await self._get_or_create_project_id(arguments)
+        resolved_branch = await self._resolve_branch(project_id, arguments["branch"])
         
-        # Get all file descriptions for this project/branch
+        # Get file descriptions for this project/branch
         file_descriptions = await self.db_manager.get_all_file_descriptions(
             project_id=project_id,
-            branch=arguments["branch"]
+            branch=resolved_branch
         )
         
         # Calculate total tokens
@@ -546,12 +600,13 @@ class MCPCodeIndexServer:
     async def _handle_find_missing_descriptions(self, arguments: Dict[str, Any]) -> Dict[str, Any]:
         """Handle find_missing_descriptions tool calls."""
         project_id = await self._get_or_create_project_id(arguments)
+        resolved_branch = await self._resolve_branch(project_id, arguments["branch"])
         folder_path = Path(arguments["folderPath"])
         
         # Get existing file descriptions
         existing_descriptions = await self.db_manager.get_all_file_descriptions(
             project_id=project_id,
-            branch=arguments["branch"]
+            branch=resolved_branch
         )
         existing_paths = {desc.file_path for desc in existing_descriptions}
         
@@ -585,12 +640,13 @@ class MCPCodeIndexServer:
     async def _handle_search_descriptions(self, arguments: Dict[str, Any]) -> Dict[str, Any]:
         """Handle search_descriptions tool calls."""
         project_id = await self._get_or_create_project_id(arguments)
+        resolved_branch = await self._resolve_branch(project_id, arguments["branch"])
         max_results = arguments.get("maxResults", 20)
         
         # Perform search
         search_results = await self.db_manager.search_file_descriptions(
             project_id=project_id,
-            branch=arguments["branch"],
+            branch=resolved_branch,
             query=arguments["query"],
             max_results=max_results
         )
@@ -614,11 +670,12 @@ class MCPCodeIndexServer:
     async def _handle_get_codebase_overview(self, arguments: Dict[str, Any]) -> Dict[str, Any]:
         """Handle get_codebase_overview tool calls."""
         project_id = await self._get_or_create_project_id(arguments)
+        resolved_branch = await self._resolve_branch(project_id, arguments["branch"])
         
         # Get all file descriptions
         file_descriptions = await self.db_manager.get_all_file_descriptions(
             project_id=project_id,
-            branch=arguments["branch"]
+            branch=resolved_branch
         )
         
         # Calculate total tokens
@@ -641,7 +698,8 @@ class MCPCodeIndexServer:
         
         return {
             "projectName": arguments["projectName"],
-            "branch": arguments["branch"],
+            "branch": resolved_branch,
+            "requestedBranch": arguments["branch"] if arguments["branch"] != resolved_branch else None,
             "totalFiles": len(file_descriptions),
             "totalTokens": total_tokens,
             "isLarge": is_large,

{mcp_code_indexer-1.1.0.dist-info → mcp_code_indexer-1.1.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mcp-code-indexer
-Version: 1.1.0
+Version: 1.1.2
 Summary: MCP server that tracks file descriptions across codebases, enabling AI agents to efficiently navigate and understand code through searchable summaries and token-aware overviews.
 Author: MCP Code Indexer Contributors
 Maintainer: MCP Code Indexer Contributors

{mcp_code_indexer-1.1.0.dist-info → mcp_code_indexer-1.1.2.dist-info}/RECORD

@@ -6,17 +6,17 @@ mcp_code_indexer/main.py,sha256=Rou-mAN9-12PPP8jC7dIs2_UNambJuC2F8BF--j-0m8,3715
 mcp_code_indexer/merge_handler.py,sha256=lJR8eVq2qSrF6MW9mR3Fy8UzrNAaQ7RsI2FMNXne3vQ,14692
 mcp_code_indexer/token_counter.py,sha256=WrifOkbF99nWWHlRlhCHAB2KN7qr83GOHl7apE-hJcE,8460
 mcp_code_indexer/database/__init__.py,sha256=aPq_aaRp0aSwOBIq9GkuMNjmLxA411zg2vhdrAuHm-w,38
-mcp_code_indexer/database/database.py,sha256=CqkGXkJLANr2qfxA2FLX1oYiNj0lTm1LyxD97dzJHSs,19650
+mcp_code_indexer/database/database.py,sha256=eG2xY5cd-oxRZ6mgGkqqBiJJfGCPqJgzoFq6kR99WfA,20300
 mcp_code_indexer/database/models.py,sha256=3wOxHKb6j3zKPWFSwB5g1TLpI507vLNZcqsxZR4VuRs,5528
 mcp_code_indexer/middleware/__init__.py,sha256=p-mP0pMsfiU2yajCPvokCUxUEkh_lu4XJP1LyyMW2ug,220
 mcp_code_indexer/middleware/error_middleware.py,sha256=v6jaHmPxf3qerYdb85X1tHIXLxgcbybpitKVakFLQTA,10109
 mcp_code_indexer/server/__init__.py,sha256=16xMcuriUOBlawRqWNBk6niwrvtv_JD5xvI36X1Vsmk,41
-mcp_code_indexer/server/mcp_server.py,sha256=RNM-0tB_1Gm5P5Vz4rNTXQKcjOyIWa4s6nirZ1BNi-g,43208
+mcp_code_indexer/server/mcp_server.py,sha256=pa7CIwokah5T8Wi28FlxzzgiozOjabY-u51VoevOHy0,46414
 mcp_code_indexer/tiktoken_cache/9b5ad71b2ce5302211f9c61530b329a4922fc6a4,sha256=Ijkht27pm96ZW3_3OFE-7xAPtR0YyTWXoRO8_-hlsqc,1681126
 mcp_code_indexer/tools/__init__.py,sha256=m01mxML2UdD7y5rih_XNhNSCMzQTz7WQ_T1TeOcYlnE,49
-mcp_code_indexer-1.1.0.dist-info/licenses/LICENSE,sha256=JN9dyPPgYwH9C-UjYM7FLNZjQ6BF7kAzpF3_4PwY4rY,1086
-mcp_code_indexer-1.1.0.dist-info/METADATA,sha256=n6fIk9JrhEmPG2LVpMCFI7XkP3Aey4dM1V_zKuhAkoA,11930
-mcp_code_indexer-1.1.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-mcp_code_indexer-1.1.0.dist-info/entry_points.txt,sha256=8HqWOw1Is7jOP1bvIgaSwouvT9z_Boe-9hd4NzyJOhY,68
-mcp_code_indexer-1.1.0.dist-info/top_level.txt,sha256=yKYCM-gMGt-cnupGfAhnZaoEsROLB6DQ1KFUuyKx4rw,17
-mcp_code_indexer-1.1.0.dist-info/RECORD,,
+mcp_code_indexer-1.1.2.dist-info/licenses/LICENSE,sha256=JN9dyPPgYwH9C-UjYM7FLNZjQ6BF7kAzpF3_4PwY4rY,1086
+mcp_code_indexer-1.1.2.dist-info/METADATA,sha256=GKi-NMcH93vJanUrHiABIojml8pQRdD2jTx4umQulMw,11930
+mcp_code_indexer-1.1.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+mcp_code_indexer-1.1.2.dist-info/entry_points.txt,sha256=8HqWOw1Is7jOP1bvIgaSwouvT9z_Boe-9hd4NzyJOhY,68
+mcp_code_indexer-1.1.2.dist-info/top_level.txt,sha256=yKYCM-gMGt-cnupGfAhnZaoEsROLB6DQ1KFUuyKx4rw,17
+mcp_code_indexer-1.1.2.dist-info/RECORD,,