JSON parser in kpeg

escaper.rb

module StringEscaper
  def process_escapes (str)
    str.gsub! /\\(u\d{4}|\D)/m do
      seq = $1
      case seq
        when '/' then "\/"
        when 'b' then "\b"
        when 'f' then "\f"
        when 'n' then "\n"
        when 'r' then "\r"
        when 't' then "\t"
        when '"' then '"'
        when '\\' then '\\'
        when /u\d{4}/ then seq[1..-1].to_i.chr
        else seq
      end
    end
    
    str
  end
end

json.kpeg

%% name = JsonParser
%% {
  require "./escaper"
  include StringEscaper
  attr_reader :result
}


value = object
      | array
      | string
      | number
      | "true" { true }
      | "false" { false }
      | "null" { nil }

object = "{" - "}" { {} }
       | "{" - object-body:obj - "}" { obj }
object-body = object-pair:x (- "," - object-pair)*:xs { Hash[[x].concat(xs)] }
object-pair = string:k - ":" - value:v { [k, v] }

array = "[" - "]" { [] }
      | "[" - array-body:arr - "]" { arr }
array-body = value:x (- "," - value)*:xs { [x].concat(xs) }

string = "\"" <string-char+> "\"" { process_escapes(text) }
string-char = !/["\\]/ .
            | "\\" string-char-escape
string-char-escape = /[\/\"bfnrt]/
                   | "u" /\d{4}/

number = number-base:b number-exponent:e { b * (10 ** e) }
       | number-base:b { b }
number-base = <number-base-whole number-base-frac> { text.to_f }
            | <number-base-whole> { text.to_i }
number-base-whole = "0" | /-?[1-9]\d*/ 
number-base-frac = /\.\d+/
number-exponent = ("E"|"e") </[+-]?\d+/> { text.to_i }

- = /[ \t]*/

root = value:v { @result = v }

This version supports any value as the "root" rather than just objects and arrays, but it's common/useful enough that I did it anyways. To be strict about it, change the root= line to:

root = (object | array):v { @result = v }