Julia → WebAssembly GC · no Julia runtime

Julia's parser, running in your browser

The JuliaSyntax parser that ships with Julia 1.13 (Base.JuliaSyntax), compiled from Julia to WasmGC with WasmCodegen. Every edit is lexed and parsed in WebAssembly, error recovery included; the page reads back the parser's raw output, tokens and tree ranges, and builds the green tree from it as build_tree does.

Source

Parse–
Bytes–
Nodes–
Tokens–

As the parser sees it hover the tree to find a node


      
Diagnostics

      
    
      
    Loading the parser…
    
    
    

    
    
      
    Green tree kind, byte range; trivia dimmed
    
      
      
    
      
  
      

  
      
    
      
      
      The whole parser
      
      
      JuliaSyntax's lexer and recursive-descent parser, unchanged: 425 Julia functions in
        718 KB of WebAssembly. Its tables (kind names, operator sets) were built when Julia loaded
        and are serialized into the module: 1215 objects.
      
    
      
    
      
      
      What WebAssembly lacks
      
      
      The lexer asks libutf8proc about Unicode characters, and numbers go through C's
        strtod. Here the Unicode tables come from the native libraries at build time and
        floats are parsed by Parsers.jl, compiled into the module with the parser.
      
    
      
    
      
      
      Checked against Julia
      
      
      On 308 KB of Julia source, Base files included, the module's output matches native
        JuliaSyntax node for node: kinds, spans, flags and diagnostics.