Qdrant MCP Server

#!/usr/bin/env python3 """ Example of programmatic usage of qdrant-mcp-server for semantic code search """ import os import sys from pathlib import Path # Add parent directory to path for imports sys.path.insert(0, str(Path(__file__).parent.parent / "src")) from qdrant_client import QdrantClient from qdrant_client.models import Distance, VectorParams, PointStruct import openai from typing import List, Dict import json class CodeSearcher: """Simple semantic code searcher using Qdrant""" def __init__(self, qdrant_url: str = "localhost", collection_name: str = "codebase"): self.client = QdrantClient(url=qdrant_url) self.collection_name = collection_name self.openai_client = openai.OpenAI() def search(self, query: str, limit: int = 10) -> List[Dict]: """Search for code semantically similar to the query""" # Generate embedding for query response = self.openai_client.embeddings.create( model="text-embedding-3-small", input=query ) query_embedding = response.data[0].embedding # Search in Qdrant results = self.client.search( collection_name=self.collection_name, query_vector=query_embedding, limit=limit, with_payload=True ) # Format results formatted_results = [] for result in results: formatted_results.append({ "file_path": result.payload.get("file_path", "Unknown"), "content": result.payload.get("content", "")[:200] + "...", "score": result.score, "language": result.payload.get("language", "Unknown"), "functions": result.payload.get("functions", []), "classes": result.payload.get("classes", []) }) return formatted_results def search_by_function(self, function_name: str) -> List[Dict]: """Search for files containing a specific function""" # Search with function-specific query query = f"function named {function_name} implementation code" results = self.search(query, limit=20) # Filter by function name in metadata filtered = [] for result in results: if function_name.lower() in [f.lower() for f in result.get("functions", [])]: filtered.append(result) return filtered def find_similar_code(self, file_path: str) -> List[Dict]: """Find code similar to a given file""" # Read the file content try: with open(file_path, 'r', encoding='utf-8') as f: content = f.read() except Exception as e: print(f"Error reading file: {e}") return [] # Search for similar code return self.search(content[:1000], limit=10) # Use first 1000 chars def main(): """Example usage""" # Check for API key if not os.getenv("OPENAI_API_KEY"): print("Error: OPENAI_API_KEY environment variable is required") sys.exit(1) # Initialize searcher searcher = CodeSearcher( qdrant_url=os.getenv("QDRANT_URL", "localhost"), collection_name="codebase" ) print("Qdrant Code Search Example\n") # Example 1: Basic semantic search print("Example 1: Semantic search for 'authentication'") print("-" * 50) results = searcher.search("user authentication and login", limit=5) for i, result in enumerate(results, 1): print(f"{i}. {result['file_path']} (score: {result['score']:.3f})") print(f" Language: {result['language']}") print(f" Preview: {result['content'][:100]}...") print() # Example 2: Search by function name print("\nExample 2: Search for function 'validateUser'") print("-" * 50) results = searcher.search_by_function("validateUser") for result in results: print(f"- {result['file_path']}") print(f" Functions: {', '.join(result['functions'][:3])}") # Example 3: Find similar code if len(sys.argv) > 1: file_path = sys.argv[1] print(f"\nExample 3: Find code similar to {file_path}") print("-" * 50) results = searcher.find_similar_code(file_path) for i, result in enumerate(results[:5], 1): print(f"{i}. {result['file_path']} (similarity: {result['score']:.3f})") # Example 4: Pattern search print("\nExample 4: Search for error handling patterns") print("-" * 50) results = searcher.search("try catch error handling exception logging", limit=5) for result in results: print(f"- {result['file_path']}") if result['classes']: print(f" Classes: {', '.join(result['classes'][:3])}") print("\nSearch complete!") if __name__ == "__main__": main()