Datui is a terminal UI for exploring tabular data. Point it at a file or URL (S3, GCS, or HTTP) and you get a keyboard-driven view: scroll, chart, query, filter, sort, pivot, export, and run stats without leaving the terminal.

It's powered by Polars streaming API under the hood, so evaluation is lazy. Supports Parquet, CSV, JSON, NDJSON, Avro, Arrow, ORC, Excel.

Hive-partitioned directories work too!

There's also a Python binding: `pip install datui` then `datui.view(your_lazyframe)`.

I got tired firing up a notebook just to peek at, or run a quick analysis on, a potentially large partitioned dataset. I wanted something in between: stay in the terminal, but with real query power and a bit of analysis.

Try it:

  curl -fsSL https://raw.githubusercontent.com/derekwisong/datui/main/scripts/install/install.sh | sh
  datui yourfile.parquet

There are more installtion options in the GitHub README.md if you dont want to pipe to shell.

GitHub: https://github.com/derekwisong/datui

Docs: https://derekwisong.github.io/datui/

First open-source project I'm sharing. UI is still rough in places. Would love feedback on what to improve next.