node package manager
Easy collaboration. Discover, share, and reuse code in your team. Create a free org »


First Mate

Linux Build Status Windows Build Status Dependency Status

TextMate helpers


npm install first-mate



{ScopeSelector} = require 'first-mate'
selector = new ScopeSelector('a | b')
selector.matches(['c']) # false 
selector.matches(['a']) # true 


{GrammarRegistry} = require 'first-mate'
registry = new GrammarRegistry()
grammar = registry.loadGrammarSync('./spec/fixtures/javascript.json')
{linetags} = grammar.tokenizeLine('var offset = 3;')
# convert compact tags representation into convenient, space-inefficient tokens 
tokens = registry.decodeTokens(linetags)
for {valuescopes} in tokens
  console.log("Token text: '#{value}' with scopes: #{scopes}")

loadGrammar(grammarPath, callback)

Asynchronously load a grammar and add it to the registry.

grammarPath - A string path to the grammar file.

callback - A function to call after the grammar is read and added to the registry. The callback receives (error, grammar) arguments.


Synchronously load a grammar and add it to the registry.

grammarPath - A string path to the grammar file.

Returns a Grammar instance.


Translate an integer representing an open scope tag from a tags array to a scope name.

id - A negative, odd integer.

Returns a scope String.

decodeTokens(line, tags)

Convert a line and a corresponding tags array returned from Grammar::tokenizeLine into an array of token objects.

line - A String representing a line of text.

tags - An Array of integers returned from Grammar::tokenizeLine.

Returns an Array of token objects, each with a value field containing a string of the token's text and a scopes field pointing to an array of every scope name containing the token.


tokenizeLine(line, [ruleStack], [firstLine])

Generate the tokenize for the given line of text.

line - The string text of the line.

ruleStack - An array of Rule objects that was returned from a previous call to this method.

firstLine - true to indicate that the very first line is being tokenized.

Returns an object with a tags key pointing to an array of integers encoding the scope structure of the line, a line key returning the line provided for convenience, and a ruleStack key pointing to an array of rules to pass to this method on future calls for lines proceeding the line that was just tokenized.

The tags array encodes the structure of the line as integers for efficient storage. This can be converted to a more convenient representation if storage is not an issue by passing the line string and tags array to GrammarRegistry::decodeTokens.

Otherwise, the integers can be interpreted as follows:

  • Positive integers represent tokens, with the number indicating the length of the token. All positive integers in the array should total to the length of the line passed to this method.

  • Negative integers represent scope start/stop tags. Odd integers are scope starts, and even integers are scope stops. An odd scope tag can be converted to a string via GrammarRegistry::scopeForId. If you want to convert an even scope tag, representing a scope end, add 1 to it to determine the corresponding scope start tag before calling ::scopeForId.


text - The string text possibly containing newlines.

Returns an object containing a lines key, pointing to an array of tokenized lines and a tags key, pointing to an array of tags arrays described above.


  • Clone the repository
  • Run npm install
  • Run npm test to run the specs
  • Run npm run benchmark to benchmark fully tokenizing jQuery 2.0.3 and the CSS for Twitter Bootstrap 3.1.1