Show HN: A macOS app for data refinement

2026-03-0518:5730github.com

A macOS app for data refinement. Contribute to jshchnz/refiner development by creating an account on GitHub.

NameName

A macOS app that auto-detects and renders JSON, XML, CSV, Markdown, code, and more.

Or in other words, an app for data refinement.

Download for macOS

Screenshot

You deal with messy text constantly: minified JSON from an API response, raw XML payloads, CSVs from a database export, Markdown you want to make quick edits to. Instead of hunting for an online formatter, pasting into some random website, or writing a quick script, just hit a hotkey. Refiner detects the format automatically and shows you a clean, syntax-highlighted, interactive view instantly.

The long-term goal is to prettify any text format. Contributions for new format renderers are very welcome.

  • Auto-detection — identifies JSON, XML, CSV, Markdown, Code, and plain text automatically
  • Interactive tree views — collapsible JSON and XML trees with expand/collapse all
  • CSV tables — clean grid layout with alternating row highlighting
  • Markdown rendering — headings, lists, blockquotes, code blocks, inline formatting
  • Code highlighting — syntax coloring for keywords, strings, numbers, comments
  • Three view modes — Raw, Formatted, and Side-by-Side (Cmd+1 / Cmd+2 / Cmd+3)
  • Floating panel — always-on-top, follows you across spaces
  • Global hotkey — summon with Cmd+Opt+R (customizable)
  • Copy formatted output — one-click copy of prettified content
  • JSON auto-fix with revert — automatically repairs malformed JSON, with one-click revert to the original

No dependencies. Just clone and run.

git clone https://github.com/jshchnz/refiner.git
cd refiner/Refiner
open Refiner.xcodeproj

Build and run with Cmd+R. Requires macOS 14+ and Xcode 15+.

Contributions are welcome — especially new format renderers. If there's a text format Refiner doesn't handle yet, open an issue or submit a PR.

MIT

You can’t perform that action at this time.


Read the original article

HackerNews