dv - a fast and powerful CLI for viewing and editing tabular data

[need47]

My first ever TUI. Would be honored if it can be listed under apps made with Textual.

dv is a powerful, interactive terminal-based viewer/editor for CSV/TSV/Excel/Parquet/JSON/NDJSON built with Python, Polars, and Textual. Inspired by VisiData, this tool provides smooth keyboard navigation, data manipulation, and a clean interface for exploring tabular data directly in terminal with multi-tab support for multiple files!

Here is the repo:
https://github.com/need47/dataframe-textual

screenshot:

[nnako]

Hi @need47 , nice implementation :-) .

But one question: As VisiData is mainly implemented in Python as well, covers loads of use cases and features (statistic analysis, ...) and has a Plugin-System for new features,... what motivated you to create dv in the first place? Have you considered contributing to VisiData? Maybe they would have liked to integrate textual into VisiData? Or was it just to pick a (already big and established project) to play with TUI applications and textual?

[need47]

Hi @nnako Thanks for the great questions—they’re ones I’ve asked myself as well. I work extensively with tabular data and was a heavy VisiData user for quite some time. While I didn’t contribute code directly, I did propose several features that were accepted—for example, displaying a single data row in a record view (column name vs. value). I also recall some discussion about integrating Textual into VisiData, though I’m not sure about the current status.

One of the initial motivations for creating dv was to make the application more visually appealing by leveraging colors and styling from Textual (or more precisely, Rich). Another major reason is performance. I often work with very large data files, and VisiData can feel slow in those scenarios. In dv, I leverage Polars as the backend, which remains fast even with millions of rows. Additionally, Polars infers data types (e.g., integers, floats) by default. As a dataframe library, it opens up many possibilities for data analysis—such as efficient searching, filtering, statistics, and more advanced features (SQL).

Again, thank you for your comment. dv started as a toy project, but it has since become practical—at least for my daily use cases. I hope it will be useful to others as well.

[dafeder]

Hi just wanted to let you know I have used this a bit and am really impressed with the UI and performance. Visidata is amazing but it stops being usable, especially on a lower-specced machine, for files over a couple gigs. Everything being "prettier" is a nice bonus.