Brief overview

AS A user of DuckDB (and many other tools)

I WANT csvbase-client to support Parquet as well as csv

SO THAT I can avoid any type inference

Additional details

Brief overview

AS A querier of various datasets

I WANT to be able to query non-csvbase files and tables

SO THAT I can benefit from the caching features without having to use csvbase

Additional details

It should work like:

import pandas as pd

df = pd.read_csv("csvbase://www.example.com/something/something-else.csv")

And it should use Last-Modified/ETags/etc.