ClickHaskell

Haskell implementation of ClickHouse DBMS Native protocol and client

Why ClickHouse+Haskell?

ClickHouse is a well known open source DBMS for building data intensive apps

Its design philosophy is close to functional programming due to append-only, support of lambdas and higher-order functions

It’s a best-in-class database for storing data in event-driven architecture

Why ClickHaskell?

Pros:

• unique type-safe and low boilerplate API
• low dependency footprint
• thread-safe and well documented network code
• extensible CI/CD pipeline
• partially formalized Native protocol based implementation

Cons:

• single active maintainer
• limited support for protocol features
• API is too static

Project structure

├contribution/     Development related code
├library/          ClickHaskell source code
├ecosystem/        Integrations with external libraries
└documentation/    clickhaskell.dev/ source code

General state of ClickHouse+Haskell

There are no other actively maintained open source alternatives to ClickHaskell at this moment.

All of the abandoned projects have hard-to-maintain codebases and poorer QA:

• clickhouse-haskell Provides much more dynamic API- hdbc-clickhouse Based on RDBMS engine HDBC Has never been released on Hackage- persistent-clickhouse RDBMS backend of Persistent library Has never been released on Hackage