tokenizer-array

general purpose regex tokenizer that returns an array of tokens

This module is based on Floby's node-tokenizer, but returns an array instead of a stream.

example

var tokenizer = require('tokenizer-array')
var rules = [
  { regex: /^\/\*([^*]|\*(?!\/))*\*\/$/, type: 'area comment' },
  { regex: /^\/\*([^*]|\*(?!\/))*\*?$/, type: 'area comment continue' },
  { regex: /^\/\/[^\n]*$/, type: 'line comment' },
  { regex: /^"([^"\n]|\\")*"?$/, type: 'quote' },
  { regex: /^'(\\?[^'\n]|\\')'?$/, type: 'char' },
  { regex: /^'[^']*$/, type: 'char continue' },
  { regex: /^#(\S*)$/, type: 'directive' },
  { regex: /^\($/, type: 'open paren' },
  { regex: /^\)$/, type: 'close paren' },
  { regex: /^\[$/, type: 'open square' },
  { regex: /^\]$/, type: 'close square' },
  { regex: /^{$/, type: 'open curly' },
  { regex: /^}$/, type: 'close curly' },
  { regex: /^([-<>~!%^&*\/+=?|.,:;]|->|<<|>>|\*\*|\|\||&&|--|\+\+|[-+*|&%\/=]=)$/,
      type: 'operator' },
  { regex: /^([_A-Za-z]\w*)$/, type: 'identifier' },
  { regex: /^[0-9]*\.?[0-9]+([eE][-+]?[0-9]+)?$/, type: 'number' },
  { regex: /^(\s+)$/, type: 'whitespace' },
  { regex: /^\\\n?$/, type: 'line continue' }
]
var src = process.argv[2]
var tokens = tokenizer(src, rules)

tokens.forEach(function (token) {
  console.log(JSON.stringify(token))
})

output:

$ node c.js 'float b=c+2;'
{"type":"identifier","source":"float"}
{"type":"whitespace","source":" "}
{"type":"identifier","source":"b"}
{"type":"operator","source":"="}
{"type":"identifier","source":"c"}
{"type":"operator","source":"+"}
{"type":"number","source":"2"}
{"type":"operator","source":";"}

You can have nested pattern matching on capture groups:

var tokenizer = require('tokenizer-array')
var rules = [
  { type: 'string', regex: /^"([^"]*)"?$/, children: [ [
    { type: 'word', regex: /^\w+$/ },
    { type: 'whitespace', regex: /^\s+$/ },
    { type: 'variable', regex: /^\$\w*$/ }
  ] ] },
  { type: 'identifier', regex: /^\w+$/ },
  { type: 'whitespace', regex: /^\s+$/ },
  { type: 'operator', regex: /^[-+=*\/.();]$/ },
]
var src = process.argv[2]
var tokens = tokenizer(src, rules)

tokens.forEach(function (token) {
  console.log(JSON.stringify(token))
})

output:

$ node children.js 'x="hello $xyz world"'
{"type":"identifier","source":"x"}
{"type":"operator","source":"="}
{"type":"string","source":"\"hello $xyz world\"","children":[[{"type":"word","source":"hello"},{"type":"whitespace","source":" "},{"type":"variable","source":"$xyz"},{"type":"whitespace","source":" "},{"type":"word","source":"world"}]]}

api

var tokenizer = require('tokenizer-array')

var tokens = tokenizer(src, rules)

Return an array of tokens by parsing a string src with an array of rules.

For each rule in rules:

• rule.regex - a pattern to match (required)
• rule.type - a string label (required)
• rule.children - an array of arrays of rules to apply for each capture group

install

npm install tokenizer-array

license

BSD