// the find
sxzz/ast-explorer
A web tool to explore the ASTs generated by parsers.
A modern rewrite of the classic astexplorer.net, covering JS/TS, CSS, HTML, Markdown, SQL, Rust, Python, C#, and more — each with multiple parser choices. State lives entirely in the URL hash, so there's no backend to maintain and links are shareable forever. Aimed at tooling authors, linter writers, and anyone who needs to understand what a parser actually produces.
The parser breadth is genuinely impressive — 12+ JS parsers alone, which lets you diff Babel vs Oxc vs SWC output side-by-side in different tabs. URL-encoded state via hash fragment is the right call: no server, no login, no expiry. CDN-selectable parser versions mean you can test against an alpha release without cloning anything. Monaco gives you real completions and formatting rather than a plain textarea.
AGPL-3.0 is a meaningful license choice that most teams won't notice until legal asks — the original astexplorer.net is MIT. The README mentions a diff/compare feature nowhere, which is the obvious next ask when you're evaluating two parsers. Adding a new parser requires touching several files across app/parser/, app/state/, and templates with no codegen or scaffolding to guide you, which slows community contributions. No offline/PWA support, so it's useless in an airplane or a conference with bad wifi despite being entirely stateless.