* ext/psych/lib/psych.rb: Adding Psych.safe_load for loading a user

  defined, restricted subset of Ruby object types.
* ext/psych/lib/psych/class_loader.rb: A class loader for
  encapsulating the logic for which objects are allowed to be
  deserialized.
* ext/psych/lib/psych/deprecated.rb: Changes to use the class loader
* ext/psych/lib/psych/exception.rb: ditto
* ext/psych/lib/psych/json/stream.rb: ditto
* ext/psych/lib/psych/nodes/node.rb: ditto
* ext/psych/lib/psych/scalar_scanner.rb: ditto
* ext/psych/lib/psych/stream.rb: ditto
* ext/psych/lib/psych/streaming.rb: ditto
* ext/psych/lib/psych/visitors/json_tree.rb: ditto
* ext/psych/lib/psych/visitors/to_ruby.rb: ditto
* ext/psych/lib/psych/visitors/yaml_tree.rb: ditto
* ext/psych/psych_to_ruby.c: ditto
* test/psych/helper.rb: ditto
* test/psych/test_safe_load.rb: tests for restricted subset.
* test/psych/test_scalar_scanner.rb: ditto
* test/psych/visitors/test_to_ruby.rb: ditto
* test/psych/visitors/test_yaml_tree.rb: ditto

CHANGELOG.rdoc

@@ -1,3 +1,27 @@
+Wed May 15 02:22:16 2013  Aaron Patterson <aaron@tenderlovemaking.com>
+
+	* ext/psych/lib/psych.rb: Adding Psych.safe_load for loading a user
+	  defined, restricted subset of Ruby object types.
+	* ext/psych/lib/psych/class_loader.rb: A class loader for
+	  encapsulating the logic for which objects are allowed to be
+	  deserialized.
+	* ext/psych/lib/psych/deprecated.rb: Changes to use the class loader
+	* ext/psych/lib/psych/exception.rb: ditto
+	* ext/psych/lib/psych/json/stream.rb: ditto
+	* ext/psych/lib/psych/nodes/node.rb: ditto
+	* ext/psych/lib/psych/scalar_scanner.rb: ditto
+	* ext/psych/lib/psych/stream.rb: ditto
+	* ext/psych/lib/psych/streaming.rb: ditto
+	* ext/psych/lib/psych/visitors/json_tree.rb: ditto
+	* ext/psych/lib/psych/visitors/to_ruby.rb: ditto
+	* ext/psych/lib/psych/visitors/yaml_tree.rb: ditto
+	* ext/psych/psych_to_ruby.c: ditto
+	* test/psych/helper.rb: ditto
+	* test/psych/test_safe_load.rb: tests for restricted subset.
+	* test/psych/test_scalar_scanner.rb: ditto
+	* test/psych/visitors/test_to_ruby.rb: ditto
+	* test/psych/visitors/test_yaml_tree.rb: ditto
+
 Sat Apr  6 02:54:08 2013  Aaron Patterson <aaron@tenderlovemaking.com>
 
 	* ext/psych/lib/psych/exception.rb: there should be only one exception

ext/psych/psych_to_ruby.c

@@ -31,11 +31,13 @@ static VALUE path2class(VALUE self, VALUE path)
 void Init_psych_to_ruby(void)
 {
     VALUE psych     = rb_define_module("Psych");
+    VALUE class_loader  = rb_define_class_under(psych, "ClassLoader", rb_cObject);
+
     VALUE visitors  = rb_define_module_under(psych, "Visitors");
     VALUE visitor   = rb_define_class_under(visitors, "Visitor", rb_cObject);
     cPsychVisitorsToRuby = rb_define_class_under(visitors, "ToRuby", visitor);
 
     rb_define_private_method(cPsychVisitorsToRuby, "build_exception", build_exception, 2);
-    rb_define_private_method(cPsychVisitorsToRuby, "path2class", path2class, 1);
+    rb_define_private_method(class_loader, "path2class", path2class, 1);
 }
 /* vim: set noet sws=4 sw=4: */

lib/psych.rb

@@ -245,6 +245,55 @@ def self.load yaml, filename = nil
     result ? result.to_ruby : result
   end
 
+  ###
+  # Safely load the yaml string in +yaml+.  By default, only the following
+  # classes are allowed to be deserialized:
+  #
+  # * TrueClass
+  # * FalseClass
+  # * NilClass
+  # * Numeric
+  # * String
+  # * Array
+  # * Hash
+  #
+  # Recursive data structures are not allowed by default.  Arbitrary classes
+  # can be allowed by adding those classes to the +whitelist+.  They are
+  # additive.  For example, to allow Date deserialization:
+  #
+  #   Psych.safe_load(yaml, [Date])
+  #
+  # Now the Date class can be loaded in addition to the classes listed above.
+  #
+  # Aliases can be explicitly allowed by changing the +aliases+ parameter.
+  # For example:
+  #
+  #   x = []
+  #   x << x
+  #   yaml = Psych.dump x
+  #   Psych.safe_load yaml               # => raises an exception
+  #   Psych.safe_load yaml, [], [], true # => loads the aliases
+  #
+  # A Psych::DisallowedClass exception will be raised if the yaml contains a
+  # class that isn't in the whitelist.
+  #
+  # A Psych::BadAlias exception will be raised if the yaml contains aliases
+  # but the +aliases+ parameter is set to false.
+  def self.safe_load yaml, whitelist_classes = [], whitelist_symbols = [], aliases = false, filename = nil
+    result = parse(yaml, filename)
+    return unless result
+
+    class_loader = ClassLoader::Restricted.new(whitelist_classes.map(&:to_s),
+                                               whitelist_symbols.map(&:to_s))
+    scanner      = ScalarScanner.new class_loader
+    if aliases
+      visitor = Visitors::ToRuby.new scanner, class_loader
+    else
+      visitor = Visitors::NoAliasRuby.new scanner, class_loader
+    end
+    visitor.accept result
+  end
+
   ###
   # Parse a YAML string in +yaml+.  Returns the Psych::Nodes::Document.
   # +filename+ is used in the exception message if a Psych::SyntaxError is
@@ -355,7 +404,7 @@ def self.dump o, io = nil, options = {}
       io      = nil
     end
 
-    visitor = Psych::Visitors::YAMLTree.new options
+    visitor = Psych::Visitors::YAMLTree.create options
     visitor << o
     visitor.tree.yaml io, options
   end
@@ -367,7 +416,7 @@ def self.dump o, io = nil, options = {}
   #
   #   Psych.dump_stream("foo\n  ", {}) # => "--- ! \"foo\\n  \"\n--- {}\n"
   def self.dump_stream *objects
-    visitor = Psych::Visitors::YAMLTree.new {}
+    visitor = Psych::Visitors::YAMLTree.create({})
     objects.each do |o|
       visitor << o
     end
@@ -377,7 +426,7 @@ def self.dump_stream *objects
   ###
   # Dump Ruby +object+ to a JSON string.
   def self.to_json object
-    visitor = Psych::Visitors::JSONTree.new
+    visitor = Psych::Visitors::JSONTree.create
     visitor << object
     visitor.tree.yaml
   end
@@ -435,7 +484,7 @@ def self.remove_type type_tag
   @load_tags = {}
   @dump_tags = {}
   def self.add_tag tag, klass
-    @load_tags[tag] = klass
+    @load_tags[tag] = klass.name
     @dump_tags[klass] = tag
   end
 

lib/psych/class_loader.rb

@@ -0,0 +1,101 @@
+require 'psych/omap'
+require 'psych/set'
+
+module Psych
+  class ClassLoader # :nodoc:
+    BIG_DECIMAL = 'BigDecimal'
+    COMPLEX     = 'Complex'
+    DATE        = 'Date'
+    DATE_TIME   = 'DateTime'
+    EXCEPTION   = 'Exception'
+    OBJECT      = 'Object'
+    PSYCH_OMAP  = 'Psych::Omap'
+    PSYCH_SET   = 'Psych::Set'
+    RANGE       = 'Range'
+    RATIONAL    = 'Rational'
+    REGEXP      = 'Regexp'
+    STRUCT      = 'Struct'
+    SYMBOL      = 'Symbol'
+
+    def initialize
+      @cache = CACHE.dup
+    end
+
+    def load klassname
+      return nil if !klassname || klassname.empty?
+
+      find klassname
+    end
+
+    def symbolize sym
+      symbol
+      sym.to_sym
+    end
+
+    constants.each do |const|
+      konst = const_get const
+      define_method(const.to_s.downcase) do
+        load konst
+      end
+    end
+
+    private
+
+    def find klassname
+      @cache[klassname] ||= resolve(klassname)
+    end
+
+    def resolve klassname
+      name    = klassname
+      retried = false
+
+      begin
+        path2class(name)
+      rescue ArgumentError, NameError => ex
+        unless retried
+          name    = "Struct::#{name}"
+          retried = ex
+          retry
+        end
+        raise retried
+      end
+    end
+
+    CACHE = Hash[constants.map { |const|
+      val = const_get const
+      begin
+        [val, ::Object.const_get(val)]
+      rescue
+        nil
+      end
+    }.compact]
+
+    class Restricted < ClassLoader
+      def initialize classes, symbols
+        @classes = classes
+        @symbols = symbols
+        super()
+      end
+
+      def symbolize sym
+        return super if @symbols.empty?
+
+        if @symbols.include? sym
+          super
+        else
+          raise DisallowedClass, 'Symbol'
+        end
+      end
+
+      private
+
+      def find klassname
+        if @classes.include? klassname
+          super
+        else
+          raise DisallowedClass, klassname
+        end
+      end
+    end
+  end
+end

lib/psych/deprecated.rb

@@ -35,7 +35,8 @@ def self.detect_implicit thing
     warn "#{caller[0]}: detect_implicit is deprecated" if $VERBOSE
     return '' unless String === thing
     return 'null' if '' == thing
-    ScalarScanner.new.tokenize(thing).class.name.downcase
+    ss = ScalarScanner.new(ClassLoader.new)
+    ss.tokenize(thing).class.name.downcase
   end
 
   def self.add_ruby_type type_tag, &block

lib/psych/exception.rb

@@ -4,4 +4,10 @@ class Exception < RuntimeError
 
   class BadAlias < Exception
   end
+
+  class DisallowedClass < Exception
+    def initialize klass_name
+      super "Tried to load unspecified class: #{klass_name}"
+    end
+  end
 end

lib/psych/json/stream.rb

@@ -6,6 +6,7 @@ module JSON
     class Stream < Psych::Visitors::JSONTree
       include Psych::JSON::RubyEvents
       include Psych::Streaming
+      extend Psych::Streaming::ClassMethods
 
       class Emitter < Psych::Stream::Emitter # :nodoc:
         include Psych::JSON::YAMLEvents

lib/psych/nodes/node.rb

@@ -1,4 +1,6 @@
 require 'stringio'
+require 'psych/class_loader'
+require 'psych/scalar_scanner'
 
 module Psych
   module Nodes
@@ -32,7 +34,7 @@ def each &block
       #
       # See also Psych::Visitors::ToRuby
       def to_ruby
-        Visitors::ToRuby.new.accept self
+        Visitors::ToRuby.create.accept(self)
       end
       alias :transform :to_ruby
 

lib/psych/scalar_scanner.rb

@@ -19,10 +19,13 @@ class ScalarScanner
                   |[-+]?(?:0|[1-9][0-9_]*) (?# base 10)
                   |[-+]?0x[0-9a-fA-F_]+    (?# base 16))$/x
 
+    attr_reader :class_loader
+
     # Create a new scanner
-    def initialize
+    def initialize class_loader
       @string_cache = {}
       @symbol_cache = {}
+      @class_loader = class_loader
     end
 
     # Tokenize +string+ returning the ruby object
@@ -63,7 +66,7 @@ def tokenize string
       when /^\d{4}-(?:1[012]|0\d|\d)-(?:[12]\d|3[01]|0\d|\d)$/
         require 'date'
         begin
-          Date.strptime(string, '%Y-%m-%d')
+          class_loader.date.strptime(string, '%Y-%m-%d')
         rescue ArgumentError
           string
         end
@@ -75,9 +78,9 @@ def tokenize string
         Float::NAN
       when /^:./
         if string =~ /^:(["'])(.*)\1/
-          @symbol_cache[string] = $2.sub(/^:/, '').to_sym
+          @symbol_cache[string] = class_loader.symbolize($2.sub(/^:/, ''))
         else
-          @symbol_cache[string] = string.sub(/^:/, '').to_sym
+          @symbol_cache[string] = class_loader.symbolize(string.sub(/^:/, ''))
         end
       when /^[-+]?[0-9][0-9_]*(:[0-5]?[0-9])+$/
         i = 0
@@ -117,17 +120,19 @@ def parse_int string
     ###
     # Parse and return a Time from +string+
     def parse_time string
+      klass = class_loader.load 'Time'
+
       date, time = *(string.split(/[ tT]/, 2))
       (yy, m, dd) = date.split('-').map { |x| x.to_i }
       md = time.match(/(\d+:\d+:\d+)(?:\.(\d*))?\s*(Z|[-+]\d+(:\d\d)?)?/)
 
       (hh, mm, ss) = md[1].split(':').map { |x| x.to_i }
       us = (md[2] ? Rational("0.#{md[2]}") : 0) * 1000000
 
-      time = Time.utc(yy, m, dd, hh, mm, ss, us)
+      time = klass.utc(yy, m, dd, hh, mm, ss, us)
 
       return time if 'Z' == md[3]
-      return Time.at(time.to_i, us) unless md[3]
+      return klass.at(time.to_i, us) unless md[3]
 
       tz = md[3].match(/^([+\-]?\d{1,2})\:?(\d{1,2})?$/)[1..-1].compact.map { |digit| Integer(digit, 10) }
       offset = tz.first * 3600
@@ -138,7 +143,7 @@ def parse_time string
         offset += ((tz[1] || 0) * 60)
       end
 
-      Time.at((time - offset).to_i, us)
+      klass.at((time - offset).to_i, us)
     end
   end
 end

lib/psych/stream.rb

@@ -32,5 +32,6 @@ def streaming?
     end
 
     include Psych::Streaming
+    extend Psych::Streaming::ClassMethods
   end
 end

lib/psych/streaming.rb

@@ -1,10 +1,15 @@
 module Psych
   module Streaming
-    ###
-    # Create a new streaming emitter.  Emitter will print to +io+.  See
-    # Psych::Stream for an example.
-    def initialize io
-      super({}, self.class.const_get(:Emitter).new(io))
+    module ClassMethods
+      ###
+      # Create a new streaming emitter.  Emitter will print to +io+.  See
+      # Psych::Stream for an example.
+      def new io
+        emitter      = const_get(:Emitter).new(io)
+        class_loader = ClassLoader.new
+        ss           = ScalarScanner.new class_loader
+        super(emitter, ss, {})
+      end
     end
 
     ###

lib/psych/visitors/json_tree.rb

@@ -5,8 +5,11 @@ module Visitors
     class JSONTree < YAMLTree
       include Psych::JSON::RubyEvents
 
-      def initialize options = {}, emitter = Psych::JSON::TreeBuilder.new
-        super
+      def self.create options = {}
+        emitter = Psych::JSON::TreeBuilder.new
+        class_loader = ClassLoader.new
+        ss           = ScalarScanner.new class_loader
+        new(emitter, ss, options)
       end
 
       def accept target