datui 0.2.21

View Polars LazyFrames and DataFrames in the terminal with datui

datui Python package

View Polars LazyFrames or DataFrames in the terminal with the datui TUI. The package provides datui.view(lf) / datui.view(df) and the datui CLI command (when the bundled binary is present or datui is on PATH). Data is passed via LazyFrame.serialize(format="json") for compatibility across Polars versions. Binary serialization is only used as a fallback when JSON is not available.

Install

From PyPI (stable):

pip install datui "polars>=0.20"

From the repo (development):

From the python/ directory, with a venv activated:

# Requires maturin: pip install maturin
cd python && maturin develop
pip install "polars>=0.20"

Usage

View a LazyFrame or DataFrame in the TUI:

import polars as pl
import datui

# From a LazyFrame (e.g. scan)
lf = pl.scan_csv("data.csv")
datui.view(lf)

# From a DataFrame (converted to LazyFrame internally)
df = pl.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
datui.view(df)

Press q to exit the TUI and return to Python.

Run the CLI from Python (e.g. datui file.csv):

If you installed via pip, the datui command is available and runs the bundled CLI. From a development install, it uses the datui binary on your PATH (e.g. from cargo build at the repo root).

Options (for datui.view)

• debug=False: Enable debug overlay.
• row_numbers=False: Show row numbers in the table.

Configuration (theme, etc.) uses the same config as the CLI: ~/.config/datui/config.toml.

Low-level API

• datui.view_from_json(json_str, *, debug=False, row_numbers=False) accepts a JSON string from LazyFrame.serialize(format="json"). This is the path used by view() and works across Polars versions.
• datui.view_from_bytes(data, *, debug=False, row_numbers=False) accepts bytes from LazyFrame.serialize() (binary). Use only when you already have binary data; format compatibility with the extension is not guaranteed across Polars versions.

Run tests

From the repo root (with the package installed, e.g. after cd python && maturin develop):

pytest python/tests/ -v