mcp-vector-search 0.15.7__py3-none-any.whl

mcp_vector_search/mcp/server.py

@@ -0,0 +1,752 @@
+"""MCP server implementation for MCP Vector Search."""
+
+import asyncio
+import os
+import sys
+from pathlib import Path
+from typing import Any
+
+from loguru import logger
+from mcp.server import Server
+from mcp.server.models import InitializationOptions
+from mcp.server.stdio import stdio_server
+from mcp.types import (
+    CallToolRequest,
+    CallToolResult,
+    ServerCapabilities,
+    TextContent,
+    Tool,
+)
+
+from ..core.database import ChromaVectorDatabase
+from ..core.embeddings import create_embedding_function
+from ..core.exceptions import ProjectNotFoundError
+from ..core.indexer import SemanticIndexer
+from ..core.project import ProjectManager
+from ..core.search import SemanticSearchEngine
+from ..core.watcher import FileWatcher
+
+
+class MCPVectorSearchServer:
+    """MCP server for vector search functionality."""
+
+    def __init__(
+        self,
+        project_root: Path | None = None,
+        enable_file_watching: bool | None = None,
+    ):
+        """Initialize the MCP server.
+
+        Args:
+            project_root: Project root directory. If None, will auto-detect from:
+                         1. PROJECT_ROOT or MCP_PROJECT_ROOT environment variable
+                         2. Current working directory
+            enable_file_watching: Enable file watching for automatic reindexing.
+                                  If None, checks MCP_ENABLE_FILE_WATCHING env var (default: True).
+        """
+        # Auto-detect project root from environment or current directory
+        if project_root is None:
+            # Priority 1: MCP_PROJECT_ROOT (new standard)
+            # Priority 2: PROJECT_ROOT (legacy)
+            # Priority 3: Current working directory
+            env_project_root = os.getenv("MCP_PROJECT_ROOT") or os.getenv(
+                "PROJECT_ROOT"
+            )
+            if env_project_root:
+                project_root = Path(env_project_root).resolve()
+                logger.info(f"Using project root from environment: {project_root}")
+            else:
+                project_root = Path.cwd()
+                logger.info(f"Using current directory as project root: {project_root}")
+
+        self.project_root = project_root
+        self.project_manager = ProjectManager(self.project_root)
+        self.search_engine: SemanticSearchEngine | None = None
+        self.file_watcher: FileWatcher | None = None
+        self.indexer: SemanticIndexer | None = None
+        self.database: ChromaVectorDatabase | None = None
+        self._initialized = False
+
+        # Determine if file watching should be enabled
+        if enable_file_watching is None:
+            # Check environment variable, default to True
+            env_value = os.getenv("MCP_ENABLE_FILE_WATCHING", "true").lower()
+            self.enable_file_watching = env_value in ("true", "1", "yes", "on")
+        else:
+            self.enable_file_watching = enable_file_watching
+
+    async def initialize(self) -> None:
+        """Initialize the search engine and database."""
+        if self._initialized:
+            return
+
+        try:
+            # Load project configuration
+            config = self.project_manager.load_config()
+
+            # Setup embedding function
+            embedding_function, _ = create_embedding_function(
+                model_name=config.embedding_model
+            )
+
+            # Setup database
+            self.database = ChromaVectorDatabase(
+                persist_directory=config.index_path,
+                embedding_function=embedding_function,
+            )
+
+            # Initialize database
+            await self.database.__aenter__()
+
+            # Setup search engine
+            self.search_engine = SemanticSearchEngine(
+                database=self.database, project_root=self.project_root
+            )
+
+            # Setup indexer for file watching
+            if self.enable_file_watching:
+                self.indexer = SemanticIndexer(
+                    database=self.database,
+                    project_root=self.project_root,
+                    config=config,
+                )
+
+                # Setup file watcher
+                self.file_watcher = FileWatcher(
+                    project_root=self.project_root,
+                    config=config,
+                    indexer=self.indexer,
+                    database=self.database,
+                )
+
+                # Start file watching
+                await self.file_watcher.start()
+                logger.info("File watching enabled for automatic reindexing")
+            else:
+                logger.info("File watching disabled")
+
+            self._initialized = True
+            logger.info(f"MCP server initialized for project: {self.project_root}")
+
+        except ProjectNotFoundError:
+            logger.error(f"Project not initialized at {self.project_root}")
+            raise
+        except Exception as e:
+            logger.error(f"Failed to initialize MCP server: {e}")
+            raise
+
+    async def cleanup(self) -> None:
+        """Cleanup resources."""
+        # Stop file watcher if running
+        if self.file_watcher and self.file_watcher.is_running:
+            logger.info("Stopping file watcher...")
+            await self.file_watcher.stop()
+            self.file_watcher = None
+
+        # Cleanup database connection
+        if self.database and hasattr(self.database, "__aexit__"):
+            await self.database.__aexit__(None, None, None)
+            self.database = None
+
+        # Clear references
+        self.search_engine = None
+        self.indexer = None
+        self._initialized = False
+        logger.info("MCP server cleanup completed")
+
+    def get_tools(self) -> list[Tool]:
+        """Get available MCP tools."""
+        tools = [
+            Tool(
+                name="search_code",
+                description="Search for code using semantic similarity",
+                inputSchema={
+                    "type": "object",
+                    "properties": {
+                        "query": {
+                            "type": "string",
+                            "description": "The search query to find relevant code",
+                        },
+                        "limit": {
+                            "type": "integer",
+                            "description": "Maximum number of results to return",
+                            "default": 10,
+                            "minimum": 1,
+                            "maximum": 50,
+                        },
+                        "similarity_threshold": {
+                            "type": "number",
+                            "description": "Minimum similarity threshold (0.0-1.0)",
+                            "default": 0.3,
+                            "minimum": 0.0,
+                            "maximum": 1.0,
+                        },
+                        "file_extensions": {
+                            "type": "array",
+                            "items": {"type": "string"},
+                            "description": "Filter by file extensions (e.g., ['.py', '.js'])",
+                        },
+                        "language": {
+                            "type": "string",
+                            "description": "Filter by programming language",
+                        },
+                        "function_name": {
+                            "type": "string",
+                            "description": "Filter by function name",
+                        },
+                        "class_name": {
+                            "type": "string",
+                            "description": "Filter by class name",
+                        },
+                        "files": {
+                            "type": "string",
+                            "description": "Filter by file patterns (e.g., '*.py' or 'src/*.js')",
+                        },
+                    },
+                    "required": ["query"],
+                },
+            ),
+            Tool(
+                name="search_similar",
+                description="Find code similar to a specific file or function",
+                inputSchema={
+                    "type": "object",
+                    "properties": {
+                        "file_path": {
+                            "type": "string",
+                            "description": "Path to the file to find similar code for",
+                        },
+                        "function_name": {
+                            "type": "string",
+                            "description": "Optional function name within the file",
+                        },
+                        "limit": {
+                            "type": "integer",
+                            "description": "Maximum number of results to return",
+                            "default": 10,
+                            "minimum": 1,
+                            "maximum": 50,
+                        },
+                        "similarity_threshold": {
+                            "type": "number",
+                            "description": "Minimum similarity threshold (0.0-1.0)",
+                            "default": 0.3,
+                            "minimum": 0.0,
+                            "maximum": 1.0,
+                        },
+                    },
+                    "required": ["file_path"],
+                },
+            ),
+            Tool(
+                name="search_context",
+                description="Search for code based on contextual description",
+                inputSchema={
+                    "type": "object",
+                    "properties": {
+                        "description": {
+                            "type": "string",
+                            "description": "Contextual description of what you're looking for",
+                        },
+                        "focus_areas": {
+                            "type": "array",
+                            "items": {"type": "string"},
+                            "description": "Areas to focus on (e.g., ['security', 'authentication'])",
+                        },
+                        "limit": {
+                            "type": "integer",
+                            "description": "Maximum number of results to return",
+                            "default": 10,
+                            "minimum": 1,
+                            "maximum": 50,
+                        },
+                    },
+                    "required": ["description"],
+                },
+            ),
+            Tool(
+                name="get_project_status",
+                description="Get project indexing status and statistics",
+                inputSchema={"type": "object", "properties": {}, "required": []},
+            ),
+            Tool(
+                name="index_project",
+                description="Index or reindex the project codebase",
+                inputSchema={
+                    "type": "object",
+                    "properties": {
+                        "force": {
+                            "type": "boolean",
+                            "description": "Force reindexing even if index exists",
+                            "default": False,
+                        },
+                        "file_extensions": {
+                            "type": "array",
+                            "items": {"type": "string"},
+                            "description": "File extensions to index (e.g., ['.py', '.js'])",
+                        },
+                    },
+                    "required": [],
+                },
+            ),
+        ]
+
+        return tools
+
+    def get_capabilities(self) -> ServerCapabilities:
+        """Get server capabilities."""
+        return ServerCapabilities(tools={"listChanged": True}, logging={})
+
+    async def call_tool(self, request: CallToolRequest) -> CallToolResult:
+        """Handle tool calls."""
+        if not self._initialized:
+            await self.initialize()
+
+        try:
+            if request.params.name == "search_code":
+                return await self._search_code(request.params.arguments)
+            elif request.params.name == "search_similar":
+                return await self._search_similar(request.params.arguments)
+            elif request.params.name == "search_context":
+                return await self._search_context(request.params.arguments)
+            elif request.params.name == "get_project_status":
+                return await self._get_project_status(request.params.arguments)
+            elif request.params.name == "index_project":
+                return await self._index_project(request.params.arguments)
+            else:
+                return CallToolResult(
+                    content=[
+                        TextContent(
+                            type="text", text=f"Unknown tool: {request.params.name}"
+                        )
+                    ],
+                    isError=True,
+                )
+        except Exception as e:
+            logger.error(f"Tool call failed: {e}")
+            return CallToolResult(
+                content=[
+                    TextContent(type="text", text=f"Tool execution failed: {str(e)}")
+                ],
+                isError=True,
+            )
+
+    async def _search_code(self, args: dict[str, Any]) -> CallToolResult:
+        """Handle search_code tool call."""
+        query = args.get("query", "")
+        limit = args.get("limit", 10)
+        similarity_threshold = args.get("similarity_threshold", 0.3)
+        file_extensions = args.get("file_extensions")
+        language = args.get("language")
+        function_name = args.get("function_name")
+        class_name = args.get("class_name")
+        files = args.get("files")
+
+        if not query:
+            return CallToolResult(
+                content=[TextContent(type="text", text="Query parameter is required")],
+                isError=True,
+            )
+
+        # Build filters
+        filters = {}
+        if file_extensions:
+            filters["file_extension"] = {"$in": file_extensions}
+        if language:
+            filters["language"] = language
+        if function_name:
+            filters["function_name"] = function_name
+        if class_name:
+            filters["class_name"] = class_name
+        if files:
+            # Convert file pattern to filter (simplified)
+            filters["file_pattern"] = files
+
+        # Perform search
+        results = await self.search_engine.search(
+            query=query,
+            limit=limit,
+            similarity_threshold=similarity_threshold,
+            filters=filters,
+        )
+
+        # Format results
+        if not results:
+            response_text = f"No results found for query: '{query}'"
+        else:
+            response_lines = [f"Found {len(results)} results for query: '{query}'\n"]
+
+            for i, result in enumerate(results, 1):
+                response_lines.append(
+                    f"## Result {i} (Score: {result.similarity_score:.3f})"
+                )
+                response_lines.append(f"**File:** {result.file_path}")
+                if result.function_name:
+                    response_lines.append(f"**Function:** {result.function_name}")
+                if result.class_name:
+                    response_lines.append(f"**Class:** {result.class_name}")
+                response_lines.append(
+                    f"**Lines:** {result.start_line}-{result.end_line}"
+                )
+                response_lines.append("**Code:**")
+                response_lines.append("```" + (result.language or ""))
+                response_lines.append(result.content)
+                response_lines.append("```\n")
+
+            response_text = "\n".join(response_lines)
+
+        return CallToolResult(content=[TextContent(type="text", text=response_text)])
+
+    async def _get_project_status(self, args: dict[str, Any]) -> CallToolResult:
+        """Handle get_project_status tool call."""
+        try:
+            config = self.project_manager.load_config()
+
+            # Get database stats
+            if self.search_engine:
+                stats = await self.search_engine.database.get_stats()
+
+                status_info = {
+                    "project_root": str(config.project_root),
+                    "index_path": str(config.index_path),
+                    "file_extensions": config.file_extensions,
+                    "embedding_model": config.embedding_model,
+                    "languages": config.languages,
+                    "total_chunks": stats.total_chunks,
+                    "total_files": stats.total_files,
+                    "index_size": (
+                        f"{stats.index_size_mb:.2f} MB"
+                        if hasattr(stats, "index_size_mb")
+                        else "Unknown"
+                    ),
+                }
+            else:
+                status_info = {
+                    "project_root": str(config.project_root),
+                    "index_path": str(config.index_path),
+                    "file_extensions": config.file_extensions,
+                    "embedding_model": config.embedding_model,
+                    "languages": config.languages,
+                    "status": "Not indexed",
+                }
+
+            response_text = "# Project Status\n\n"
+            response_text += f"**Project Root:** {status_info['project_root']}\n"
+            response_text += f"**Index Path:** {status_info['index_path']}\n"
+            response_text += (
+                f"**File Extensions:** {', '.join(status_info['file_extensions'])}\n"
+            )
+            response_text += f"**Embedding Model:** {status_info['embedding_model']}\n"
+            response_text += f"**Languages:** {', '.join(status_info['languages'])}\n"
+
+            if "total_chunks" in status_info:
+                response_text += f"**Total Chunks:** {status_info['total_chunks']}\n"
+                response_text += f"**Total Files:** {status_info['total_files']}\n"
+                response_text += f"**Index Size:** {status_info['index_size']}\n"
+            else:
+                response_text += f"**Status:** {status_info['status']}\n"
+
+            return CallToolResult(
+                content=[TextContent(type="text", text=response_text)]
+            )
+
+        except ProjectNotFoundError:
+            return CallToolResult(
+                content=[
+                    TextContent(
+                        type="text",
+                        text=f"Project not initialized at {self.project_root}. Run 'mcp-vector-search init' first.",
+                    )
+                ],
+                isError=True,
+            )
+
+    async def _index_project(self, args: dict[str, Any]) -> CallToolResult:
+        """Handle index_project tool call."""
+        force = args.get("force", False)
+        file_extensions = args.get("file_extensions")
+
+        try:
+            # Import indexing functionality
+            from ..cli.commands.index import run_indexing
+
+            # Run indexing
+            await run_indexing(
+                project_root=self.project_root,
+                force_reindex=force,
+                extensions=file_extensions,
+                show_progress=False,  # Disable progress for MCP
+            )
+
+            # Reinitialize search engine after indexing
+            await self.cleanup()
+            await self.initialize()
+
+            return CallToolResult(
+                content=[
+                    TextContent(
+                        type="text", text="Project indexing completed successfully!"
+                    )
+                ]
+            )
+
+        except Exception as e:
+            return CallToolResult(
+                content=[TextContent(type="text", text=f"Indexing failed: {str(e)}")],
+                isError=True,
+            )
+
+    async def _search_similar(self, args: dict[str, Any]) -> CallToolResult:
+        """Handle search_similar tool call."""
+        file_path = args.get("file_path", "")
+        function_name = args.get("function_name")
+        limit = args.get("limit", 10)
+        similarity_threshold = args.get("similarity_threshold", 0.3)
+
+        if not file_path:
+            return CallToolResult(
+                content=[
+                    TextContent(type="text", text="file_path parameter is required")
+                ],
+                isError=True,
+            )
+
+        try:
+            from pathlib import Path
+
+            # Convert to Path object
+            file_path_obj = Path(file_path)
+            if not file_path_obj.is_absolute():
+                file_path_obj = self.project_root / file_path_obj
+
+            if not file_path_obj.exists():
+                return CallToolResult(
+                    content=[
+                        TextContent(type="text", text=f"File not found: {file_path}")
+                    ],
+                    isError=True,
+                )
+
+            # Run similar search
+            results = await self.search_engine.search_similar(
+                file_path=file_path_obj,
+                function_name=function_name,
+                limit=limit,
+                similarity_threshold=similarity_threshold,
+            )
+
+            # Format results
+            if not results:
+                return CallToolResult(
+                    content=[
+                        TextContent(
+                            type="text", text=f"No similar code found for {file_path}"
+                        )
+                    ]
+                )
+
+            response_lines = [
+                f"Found {len(results)} similar code snippets for {file_path}\n"
+            ]
+
+            for i, result in enumerate(results, 1):
+                response_lines.append(
+                    f"## Result {i} (Score: {result.similarity_score:.3f})"
+                )
+                response_lines.append(f"**File:** {result.file_path}")
+                if result.function_name:
+                    response_lines.append(f"**Function:** {result.function_name}")
+                if result.class_name:
+                    response_lines.append(f"**Class:** {result.class_name}")
+                response_lines.append(
+                    f"**Lines:** {result.start_line}-{result.end_line}"
+                )
+                response_lines.append("**Code:**")
+                response_lines.append("```" + (result.language or ""))
+                # Show more of the content for similar search
+                content_preview = (
+                    result.content[:500]
+                    if len(result.content) > 500
+                    else result.content
+                )
+                response_lines.append(
+                    content_preview + ("..." if len(result.content) > 500 else "")
+                )
+                response_lines.append("```\n")
+
+            result_text = "\n".join(response_lines)
+
+            return CallToolResult(content=[TextContent(type="text", text=result_text)])
+
+        except Exception as e:
+            return CallToolResult(
+                content=[
+                    TextContent(type="text", text=f"Similar search failed: {str(e)}")
+                ],
+                isError=True,
+            )
+
+    async def _search_context(self, args: dict[str, Any]) -> CallToolResult:
+        """Handle search_context tool call."""
+        description = args.get("description", "")
+        focus_areas = args.get("focus_areas")
+        limit = args.get("limit", 10)
+
+        if not description:
+            return CallToolResult(
+                content=[
+                    TextContent(type="text", text="description parameter is required")
+                ],
+                isError=True,
+            )
+
+        try:
+            # Perform context search
+            results = await self.search_engine.search_by_context(
+                context_description=description, focus_areas=focus_areas, limit=limit
+            )
+
+            # Format results
+            if not results:
+                return CallToolResult(
+                    content=[
+                        TextContent(
+                            type="text",
+                            text=f"No contextually relevant code found for: {description}",
+                        )
+                    ]
+                )
+
+            response_lines = [
+                f"Found {len(results)} contextually relevant code snippets"
+            ]
+            if focus_areas:
+                response_lines[0] += f" (focus: {', '.join(focus_areas)})"
+            response_lines[0] += f" for: {description}\n"
+
+            for i, result in enumerate(results, 1):
+                response_lines.append(
+                    f"## Result {i} (Score: {result.similarity_score:.3f})"
+                )
+                response_lines.append(f"**File:** {result.file_path}")
+                if result.function_name:
+                    response_lines.append(f"**Function:** {result.function_name}")
+                if result.class_name:
+                    response_lines.append(f"**Class:** {result.class_name}")
+                response_lines.append(
+                    f"**Lines:** {result.start_line}-{result.end_line}"
+                )
+                response_lines.append("**Code:**")
+                response_lines.append("```" + (result.language or ""))
+                # Show more of the content for context search
+                content_preview = (
+                    result.content[:500]
+                    if len(result.content) > 500
+                    else result.content
+                )
+                response_lines.append(
+                    content_preview + ("..." if len(result.content) > 500 else "")
+                )
+                response_lines.append("```\n")
+
+            result_text = "\n".join(response_lines)
+
+            return CallToolResult(content=[TextContent(type="text", text=result_text)])
+
+        except Exception as e:
+            return CallToolResult(
+                content=[
+                    TextContent(type="text", text=f"Context search failed: {str(e)}")
+                ],
+                isError=True,
+            )
+
+
+def create_mcp_server(
+    project_root: Path | None = None, enable_file_watching: bool | None = None
+) -> Server:
+    """Create and configure the MCP server.
+
+    Args:
+        project_root: Project root directory. If None, will auto-detect.
+        enable_file_watching: Enable file watching for automatic reindexing.
+                              If None, checks MCP_ENABLE_FILE_WATCHING env var (default: True).
+    """
+    server = Server("mcp-vector-search")
+    mcp_server = MCPVectorSearchServer(project_root, enable_file_watching)
+
+    @server.list_tools()
+    async def handle_list_tools() -> list[Tool]:
+        """List available tools."""
+        return mcp_server.get_tools()
+
+    @server.call_tool()
+    async def handle_call_tool(name: str, arguments: dict | None):
+        """Handle tool calls."""
+        # Create a mock request object for compatibility
+        from types import SimpleNamespace
+
+        mock_request = SimpleNamespace()
+        mock_request.params = SimpleNamespace()
+        mock_request.params.name = name
+        mock_request.params.arguments = arguments or {}
+
+        result = await mcp_server.call_tool(mock_request)
+
+        # Return the content from the result
+        return result.content
+
+    # Store reference for cleanup
+    server._mcp_server = mcp_server
+
+    return server
+
+
+async def run_mcp_server(
+    project_root: Path | None = None, enable_file_watching: bool | None = None
+) -> None:
+    """Run the MCP server using stdio transport.
+
+    Args:
+        project_root: Project root directory. If None, will auto-detect.
+        enable_file_watching: Enable file watching for automatic reindexing.
+                              If None, checks MCP_ENABLE_FILE_WATCHING env var (default: True).
+    """
+    server = create_mcp_server(project_root, enable_file_watching)
+
+    # Create initialization options with proper capabilities
+    init_options = InitializationOptions(
+        server_name="mcp-vector-search",
+        server_version="0.4.0",
+        capabilities=ServerCapabilities(tools={"listChanged": True}, logging={}),
+    )
+
+    try:
+        async with stdio_server() as (read_stream, write_stream):
+            await server.run(read_stream, write_stream, init_options)
+    except KeyboardInterrupt:
+        logger.info("Received interrupt signal, shutting down...")
+    except Exception as e:
+        logger.error(f"MCP server error: {e}")
+        raise
+    finally:
+        # Cleanup
+        if hasattr(server, "_mcp_server"):
+            logger.info("Performing server cleanup...")
+            await server._mcp_server.cleanup()
+
+
+if __name__ == "__main__":
+    # Allow specifying project root as command line argument
+    project_root = Path(sys.argv[1]) if len(sys.argv) > 1 else None
+
+    # Check for file watching flag in command line args
+    enable_file_watching = None
+    if "--no-watch" in sys.argv:
+        enable_file_watching = False
+        sys.argv.remove("--no-watch")
+    elif "--watch" in sys.argv:
+        enable_file_watching = True
+        sys.argv.remove("--watch")
+
+    asyncio.run(run_mcp_server(project_root, enable_file_watching))