#Css lexing for the rest of us

Look here sonny, back in my day we didn't have no fancy regular expressions or tools that generated lexers/parsers from grammar definitions.

What we DID have was a can-do attitude, a spring in our step, and first order functions. And we made due, dammit.

Ok, but what is this?

This is a (bad) css lexer with some search functionality bolted on the front of it

Getting stuff out in node

Lexer.js provides a Lexer constructor. The lexer object inherits from EventEmitter

var Lexer = require('./lexer').Lexer;

//instantiate lexer (you need a new one for each string you lex)
var lexy = new Lexer();

// Lexer emits 2 kinds of events:

//'lexerToken' events are emitted when lexer has lexes another token and it's ready.
lexy.on('lexerToken', function (token) {
  //do something with your token.
});

// The 'finished' event is emitted when the lexer runs out of string to lex or encounters and error. 
// (Unfortunately, no error events yet)
lexy.on('finished', function () {
  if (process.argv[3] === "-s") {
    searchtime();
  }
});

//To get the lexer rolling, simply call 
lexy.begin(data); // data is a string

Command Line Usage

Run captain.js on a file to get all tokens in stdout:

./captain.js sample.css

Run captain.js with the search flag like this:

./captain.js sample.css -s ".one .two"

you will get an output of how many times this selector was found in the code (possibly as a subset of another selector)

Why not save yourself a lot of time and use a regex?

This, basically.

Credits

Ideas for writing a lexer come from this excellent presentation. Ideas for what makes lexing useful outside of a parser/interpreter come from pfff.