barq-sdk-python 0.1.0

Python client for Barq DB

Barq SDK Python

Official Python client for Barq DB - a high-performance vector database.

Installation

pip install barq-sdk-python

Quick Start

HTTP Client

from barq import BarqClient

# Initialize client
client = BarqClient("http://localhost:8080", api_key="your-api-key")

# Check health
if client.health():
    print("Connected!")

# Create a collection
client.create_collection(
    name="documents",
    dimension=384,
    metric="L2"
)

# Insert a document
client.insert_document(
    collection="documents",
    id=1,
    vector=[0.1] * 384,
    payload={"title": "Hello World"}
)

# Search
results = client.search(
    collection="documents",
    vector=[0.1] * 384,
    top_k=5
)

for result in results:
    print(f"ID: {result['id']}, Score: {result['score']}")

client.close()

gRPC Client

from barq import GrpcClient

# Initialize gRPC client
client = GrpcClient("localhost:50051")

# Check health
if client.health():
    print("Connected via gRPC!")

# Create collection
client.create_collection(
    name="documents",
    dimension=384,
    metric="L2"
)

# Insert document
client.insert_document(
    collection="documents",
    id="doc1",
    vector=[0.1] * 384,
    payload={"title": "Hello World"}
)

# Search
results = client.search(
    collection="documents",
    vector=[0.1] * 384,
    top_k=5
)

for result in results:
    print(f"ID: {result['id']}, Score: {result['score']}")

API Reference

BarqClient (HTTP)

Method	Description
health()	Check server health
create_collection(name, dimension, metric, index, text_fields)	Create a new collection
insert_document(collection, id, vector, payload)	Insert a document
search(collection, vector, query, top_k, filter)	Search for similar vectors
close()	Close the client connection

GrpcClient

Method	Description
health()	Check server health
create_collection(name, dimension, metric)	Create a new collection
insert_document(collection, id, vector, payload)	Insert a document
search(collection, vector, top_k)	Search for similar vectors

Requirements

• Python >= 3.8
• httpx >= 0.23
• grpcio >= 1.50.0
• protobuf >= 4.21.0

License

MIT License - see LICENSE for details.