Parse output of var_dump and return as ruby data

README.markdown

Simple Treetop based parser for converting a subset of data output by the php var_dump() function to ruby.

Based on the code from http://thingsaaronmade.com/blog/a-quick-intro-to-writing-a-parser-using-treetop.html

phpdump.treetop

grammar PhpDump
	rule value
		(php_array / php_object / php_string / php_number / php_null)
	end

	rule php_array
		"array(" [0-9]+ ")" os "{" os key_value_array_list "}" <PhpArray>
	end

	rule key_value_array_list
		(key_value_pair os)* <PhpArrayList>
	end

	rule php_object
		"object(" [a-zA-Z_0-9]+ ")" os "#" [0-9]+ os "(" integer ")" os "{" os
			key_value_object_list
         "}" <PhpObject>
	end

	rule key_value_object_list
		(key_value_pair os)* <PhpObjectList>
	end

	rule key_value_pair
		"[" keyvalue "]" os "=>" os value <PhpKeyValuePair>
	end

	rule keyvalue
		string / integer
	end
 	
	rule php_string
		"string(" [0-9]+ ")" os string <PhpValueString>
	end

	rule php_number
		php_int / php_float / php_float_nan
	end

	rule php_null
		"NULL" <NullValue>
	end
	
	rule php_int
		"int(" integer ")" <PhpValue>
	end

	rule php_float
		"float(" float ")" <PhpValue>
	end

	rule php_float_nan
		"float(NAN)" <PhpNaN>
	end

	rule integer
		('+' / '-')? [0-9]+ <IntegerLiteral>
	end

	rule float
		('+' / '-')? [0-9]+ (('.' [0-9]+) / ('e' [0-9]+))? <FloatLiteral>
	end

	rule string
		'"' ([^"\\] / "\\" . )* '"' <StringLiteral>
	end

	rule ws
		[\s]+
	end

	rule os
		[\s]*
	end
end

phpdump_extensions.rb

module PhpDump
	class PhpArray < Treetop::Runtime::SyntaxNode
		def to_data
			self.elements.select { |e| e.is_a?(PhpArrayList) }.inject([]) { |arr, x| arr + x.to_data }
		end
	end

	class PhpArrayList < Treetop::Runtime::SyntaxNode
		def to_data
			self.elements.map{|a|a.elements[0]}.select { |e| e.is_a?(PhpKeyValuePair) }.map { |x| x.to_data.values[0] }
		end
	end

	class PhpObject < Treetop::Runtime::SyntaxNode
		def to_data
			self.elements.select { |e| e.is_a?(PhpObjectList) }.inject({}) { |o, x| o.merge(x.to_data) }
		end
	end

	class PhpObjectList < Treetop::Runtime::SyntaxNode
		def to_data
			self.elements.map{|a|a.elements[0]}.select { |e| e.is_a?(PhpKeyValuePair) }.inject({}) { |o, x| o.merge(x.to_data) }
		end
	end

	class PhpValue < Treetop::Runtime::SyntaxNode
		def to_data
			self.elements[1].to_data
		end
	end

	class PhpValueString < Treetop::Runtime::SyntaxNode
		def to_data
			self.elements[4].to_data
		end
	end

	class IntegerLiteral < Treetop::Runtime::SyntaxNode
		def to_data
			self.text_value.to_i
		end
	end

	class StringLiteral < Treetop::Runtime::SyntaxNode
		def to_data
			chars = self.text_value[1...-1].scan(/./)
			buf = ''
			until chars.empty?
				c = chars.shift
				if c == '\\'
					c = chars.shift
					case c
					when "n"
						buf << "\n"
					when "r"
						buf << "\r"
					when "\\"
						buf << "\\"
					when "t"
						buf << "\t"
					else
						buf << c
					end
				else
					buf << c
				end
			end
			buf
		end
	end

	class FloatLiteral < Treetop::Runtime::SyntaxNode
		def to_data
			self.text_value.to_f
		end
	end

	class NullValue < Treetop::Runtime::SyntaxNode
		def to_data
			nil
		end
	end

	class PhpNaN < Treetop::Runtime::SyntaxNode
		def to_data
			0.0/0.0
		end
	end

	class PhpInf < Treetop::Runtime::SyntaxNode
		def to_data
			1.0/0.0
		end
	end

	class PhpIntKey < Treetop::Runtime::SyntaxNode
		def to_data
			self.text_value.to_i
		end
	end

	class PhpStringKey < Treetop::Runtime::SyntaxNode
		def to_data
			self.text_value
		end
	end	

	class PhpKeyValuePair < Treetop::Runtime::SyntaxNode
		def to_data
			return { self.elements[1].to_data => self.elements[6].to_data }
		end
	end
end