JSON: encode objects that don't have a native JSON representation using to_hash, if available, instead of instance_values (the old fallback) or to_s (other encoders' default). Encode BigDecimal and Regexp encode as strings to conform with other encoders. Try to transcode non-UTF-8 strings.

3 files changed, 43 insertions, 7 deletions

diff --git a/activesupport/CHANGELOG b/activesupport/CHANGELOG
index 5ada5a1e9b..c47839b001 100644
--- a/activesupport/CHANGELOG
+++ b/activesupport/CHANGELOG
@@ -1,3 +1,8 @@
+*Rails 3.0.0 [beta 4/release candidate] (unreleased)*
+
+* JSON: encode objects that don't have a native JSON representation using to_hash, if available, instead of instance_values (the old fallback) or to_s (other encoders' default). Encode BigDecimal and Regexp encode as strings to conform with other encoders. Try to transcode non-UTF-8 strings.  [Jeremy Kemper]
+
+
 *Rails 3.0.0 [beta 3] (April 13th, 2010)*
 
 * HashWithIndifferentAccess: remove inherited symbolize_keys! since its keys are always strings.  [Santiago Pastorino]
diff --git a/activesupport/lib/active_support/json/encoding.rb b/activesupport/lib/active_support/json/encoding.rb
index 8ba45f7ea2..0f38fd0e89 100644
--- a/activesupport/lib/active_support/json/encoding.rb
+++ b/activesupport/lib/active_support/json/encoding.rb
@@ -1,4 +1,5 @@
 # encoding: utf-8
+require 'bigdecimal'
 require 'active_support/core_ext/array/wrap'
 require 'active_support/core_ext/hash/except'
 require 'active_support/core_ext/hash/slice'
@@ -102,7 +103,9 @@ module ActiveSupport
         end
 
         def escape(string)
-          string = string.dup.force_encoding(::Encoding::BINARY) if string.respond_to?(:force_encoding)
+          if string.respond_to?(:force_encoding)
+            string = string.encode(::Encoding::UTF_8, undef: :replace).force_encoding(::Encoding::BINARY)
+          end
           json = string.
             gsub(escape_regex) { |s| ESCAPED_CHARS[s] }.
             gsub(/([\xC0-\xDF][\x80-\xBF]|
@@ -110,7 +113,9 @@ module ActiveSupport
                    [\xF0-\xF7][\x80-\xBF]{3})+/nx) { |s|
             s.unpack("U*").pack("n*").unpack("H*")[0].gsub(/.{4}/n, '\\\\u\&')
           }
-          %("#{json}")
+          json = %("#{json}")
+          json.force_encoding(::Encoding::UTF_8) if json.respond_to?(:force_encoding)
+          json
         end
       end
 
@@ -128,7 +133,13 @@ class Object
     ActiveSupport::JSON.encode(self, options)
   end
 
-  def as_json(options = nil) instance_values end #:nodoc:
+  def as_json(options = nil) #:nodoc:
+    if respond_to?(:to_hash)
+      to_hash
+    else
+      instance_values
+    end
+  end
 end
 
 # A string that returns itself as its JSON-encoded form.
@@ -166,9 +177,12 @@ class Numeric
   def encode_json(encoder) to_s end #:nodoc:
 end
 
+class BigDecimal
+  def as_json(options = nil) to_s end #:nodoc:
+end
+
 class Regexp
-  def as_json(options = nil) self end #:nodoc:
-  def encode_json(encoder) inspect end #:nodoc:
+  def as_json(options = nil) to_s end #:nodoc:
 end
 
 module Enumerable
diff --git a/activesupport/test/json/encoding_test.rb b/activesupport/test/json/encoding_test.rb
index 188b799f3f..ff95c0ca18 100644
--- a/activesupport/test/json/encoding_test.rb
+++ b/activesupport/test/json/encoding_test.rb
@@ -9,6 +9,12 @@ class TestJSONEncoding < Test::Unit::TestCase
     end
   end
 
+  class Hashlike
+    def to_hash
+      { :a => 1 }
+    end
+  end
+
   class Custom
     def as_json(options)
       'custom'
@@ -19,7 +25,8 @@ class TestJSONEncoding < Test::Unit::TestCase
   FalseTests    = [[ false, %(false) ]]
   NilTests      = [[ nil,   %(null)  ]]
   NumericTests  = [[ 1,     %(1)     ],
-                   [ 2.5,   %(2.5)   ]]
+                   [ 2.5,   %(2.5)   ],
+                   [ BigDecimal('2.5'), %("#{BigDecimal('2.5').to_s}") ]]
 
   StringTests   = [[ 'this is the <string>',     %("this is the \\u003Cstring\\u003E")],
                    [ 'a "string" with quotes & an ampersand', %("a \\"string\\" with quotes \\u0026 an ampersand") ],
@@ -35,11 +42,12 @@ class TestJSONEncoding < Test::Unit::TestCase
                    [ :"a b", %("a b")  ]]
 
   ObjectTests   = [[ Foo.new(1, 2), %({\"a\":1,\"b\":2}) ]]
+  HashlikeTests = [[ Hashlike.new, %({\"a\":1}) ]]
   CustomTests   = [[ Custom.new, '"custom"' ]]
 
   VariableTests = [[ ActiveSupport::JSON::Variable.new('foo'), 'foo'],
                    [ ActiveSupport::JSON::Variable.new('alert("foo")'), 'alert("foo")']]
-  RegexpTests   = [[ /^a/, '/^a/' ], [/^\w{1,2}[a-z]+/ix, '/^\\w{1,2}[a-z]+/ix']]
+  RegexpTests   = [[ /^a/, '"(?-mix:^a)"' ], [/^\w{1,2}[a-z]+/ix, '"(?ix-m:^\\\\w{1,2}[a-z]+)"']]
 
   DateTests     = [[ Date.new(2005,2,1), %("2005/02/01") ]]
   TimeTests     = [[ Time.utc(2005,2,1,15,15,10), %("2005/02/01 15:15:10 +0000") ]]
@@ -91,6 +99,15 @@ class TestJSONEncoding < Test::Unit::TestCase
     end
   end
 
+  if '1.9'.respond_to?(:force_encoding)
+    def test_non_utf8_string_transcodes
+      s = '二'.encode('Shift_JIS')
+      result = ActiveSupport::JSON.encode(s)
+      assert_equal '"\\u4e8c"', result
+      assert_equal Encoding::UTF_8, result.encoding
+    end
+  end
+
   def test_exception_raised_when_encoding_circular_reference
     a = [1]
     a << a