chquery 0.0.2

CHQuery is an async Python client and query builder for ClickHouse, built on top of asynch using native TCP connection for speed.

chquery: Async ClickHouse Python Client

chquery is an async Python client and query builder for ClickHouse, built on top of asynch. It provides a simple, high-level API for querying ClickHouse asynchronously, with a convenient query builder. Used by myself for most of my data analytics since the beginning of 2025.

Features

• Async connection pool management
• High-level query builder (select, filters, order, limit, insert, etc.)
• Raw SQL execution
• Custom error handling
• Batch insert and async iteration for large result sets
• Type hints and docstrings for easy development

Installation

Install chquery:

pip install git+https://github.com/filipnyquist/chquery.git

Usage Example

import asyncio
from chquery import CHDriver, DbError

async def main():
    driver = CHDriver(dsn="clickhouse://user:pass@localhost/default")
    await driver.startup()
    try:
        # Simple query
        rows = await driver.fetch_all("SELECT * FROM my_table LIMIT 10")
        print(rows)

        # Using the query builder
        qb = driver.qb("my_table").fields(["id", "name"]).filters({"id[gt]": 10}).order_by(["-id"]).limit(5)
        results = await qb.all()
        print(results)

        # Insert data
        await driver.qb("my_table").insert({"id": 123, "name": "Alice"})

        # Check if a row exists
        exists = await driver.qb("my_table").filters({"id[eq]": 123}).exists()
        print("Exists:", exists)

        # Count rows
        count = await driver.qb("my_table").count()
        print("Count:", count)

        # Async iteration for large results
        async for row in driver.qb("my_table").all_iter(batch_size=100):
            print(row)
    except DbError as e:
        print("Database error:", e)
    finally:
        await driver.shutdown()

asyncio.run(main())

API Overview

CHDriver

• startup() / shutdown() / close()
• fetch_all(query, *args) / fetch_one(query, *args)
• insert_many(query, data)
• raw(query, *args, fetch='all')
• ping()
• qb(table) — returns a CHQueryBuilder

CHQueryBuilder

• fields(list)
• filters(dict) — supports [eq], [ne], [lt], [lte], [gt], [gte], [in], [like], [isnull]
• order_by(list)
• limit(n) / offset(n)
• all() / one() / count() / exists()
• all_iter(batch_size=1000) — async generator
• insert(data)
• reset()

License

MIT