memory-service-client 0.2.0

Python client library for Memory Service gRPC API

Memory Service Client

A Python client library for the Memory Service gRPC API.

Installation

pip install memory-service-client

Usage

Basic Usage

from memory_service_client import MemoryServiceClient

# Connect to the service
with MemoryServiceClient("localhost:50051") as client:
    # Process a document
    response = client.process_document(
        document_id="doc1",
        content="This is a sample document content.",
        title="Sample Document"
    )
    print(f"Document processed: {response.status}")
    
    # Query knowledge
    response = client.query_knowledge(
        query="What is this document about?",
        top_k=3
    )
    print(f"Answer: {response.answer}")

Using gRPC Stubs Directly

import grpc
from memory_service_client import RagServiceStub, ProcessDocumentRequest, ChunkingConfig

# Create channel and stub
channel = grpc.insecure_channel("localhost:50051")
stub = RagServiceStub(channel)

# Create request
request = ProcessDocumentRequest(
    document_id="doc1",
    content="Document content",
    chunking_config=ChunkingConfig(
        chunk_size=1000,
        chunk_overlap=100,
        strategy="sentence"
    )
)

# Make the call
response = stub.ProcessDocument(request)
print(response)

API Reference

MemoryServiceClient

High-level client class that provides convenient methods for interacting with the Memory Service.

Methods

• process_document(document_id, content, title="", source_url="", metadata=None, chunk_size=1000, chunk_overlap=100, chunking_strategy="sentence"): Process a document for RAG
• query_knowledge(query, top_k=5, similarity_threshold=0.7, include_metadata=True, use_reranking=False, llm_model="gpt-3.5-turbo", max_tokens=1000): Query the knowledge base
• get_document_status(document_id): Get document processing status
• hello_world(name="World"): Test connection

gRPC Messages

The library exposes all protobuf message types:

• ProcessDocumentRequest, ProcessDocumentResponse
• QueryKnowledgeRequest, QueryKnowledgeResponse
• DocumentStatusRequest, DocumentStatusResponse
• HelloWorldRequest, HelloWorldResponse
• ChunkingConfig, QueryConfig
• ProcessingStats, QueryStats, RetrievedChunk

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Format code
black .

# Lint code
ruff check .

License

MIT License