Writing Parquet Files Using Haskell

13 Likes

Also now listed under the Apache parquet implementations.

4 Likes

That’s awesome. Is it production-ready?

As a matter of testing yes. It’s tested against the official testing corpus: GitHub - apache/parquet-testing: Apache Parquet Testing · GitHub

So anything that’s covered is pretty thoroughly done.

As a matter of feature completeness that depends.

As a matter of speed we haven’t profiled it publicly but its performance is similar to the golang implementation.

1 Like

How much work would it be to make it available as a standalone package?

@iand675 is working on GitHub - iand675/wireform-: The fast, correct, Haskell package ecosystem. · GitHub which might be better for non-DS parquet stuff.

2 Likes

As always with dataframe (and me working in research), I’m happy to see standards being adopted into the Haskell ecosystem. Nice job!