<img width="24px" height="24px" src="./documentation/app/public/assets/logo.svg"> 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:

<ul type="square"> <li>unique type-safe and low boilerplate API</li> <li>low dependency footprint</li> <li>thread-safe and well documented network code</li> <li>extensible CI/CD pipeline</li> <li>partially formalized <a href="https://clickhouse.com/docs/native-protocol/basics">Native protocol</a> based implementation</li> </ul>

Cons:

<ul type="square"> <li>single active maintainer</li> <li>limited support for protocol features</li> <li>API is too static</li> </ul>

Project structure

<pre> ├contribution/ Development related code ├library/ ClickHaskell source code ├ecosystem/ Integrations with external libraries └documentation/ <a href="https://clickhaskell.dev/">clickhaskell.dev/</a> source code </pre>

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:

<ul type=square> <li><a href="https://github.com/2049foundation/clickhouse-haskell/">clickhouse-haskell</a></li> Provides much more dynamic API <li><a href="https://github.com/zaneli/hdbc-clickhouse/tree/master">hdbc-clickhouse</a></li> Based on RDBMS engine HDBC<br> Has never been released on Hackage <li><a href="https://github.com/Diamondy4/persistent-clickhouse">persistent-clickhouse</a></li> RDBMS backend of Persistent library<br> Has never been released on Hackage </ul>