node package manager


r/w stream of glsl tokens


Maps GLSL string data into GLSL tokens, either synchronously or using a streaming API.

var tokenString = require('glsl-tokenizer/string')
var tokenStream = require('glsl-tokenizer/stream')
var fs = require('fs')
// Synchronously: 
var tokens = tokenString(fs.readFileSync('some.glsl'))
// Streaming API: 
  .on('data', function(token) {
    console.log(, token.position, token.type)


tokens = require('glsl-tokenizer/string')(src, [opt])

Returns an array of tokens given the GLSL source string src

You can specify opt.version string to use different keywords/builtins, such as '300 es' for WebGL2. Otherwise, will assume GLSL 100 (WebGL1).

var tokens = tokenizer(src, {
  version: '300 es'

stream = require('glsl-tokenizer/stream')([opt])

Emits 'data' events whenever a token is parsed with a token object as output.

As above, you can specify opt.version.


{ 'type': TOKEN_TYPE
, 'data': "string of constituent data"
, 'position': integer position within the GLSL source
, 'line': line number within the GLSL source
, 'column': column number within the GLSL source }

The available token types are:

  • block-comment: /* ... */
  • line-comment: // ... \n
  • preprocessor: # ... \n
  • operator: Any operator. If it looks like punctuation, it's an operator.
  • float: Optionally suffixed with f
  • ident: User defined identifier.
  • builtin: Builtin function.
  • eof: Emitted on end; data will === '(eof)'.
  • integer
  • whitespace
  • keyword


MIT, see for further information.