Fixed validates_length_of to work on UTF-8 strings by using characters instead of bytes (closes #3699) [Masao Mutoh]

git-svn-id: http://svn-commit.rubyonrails.org/rails/trunk@3654 5ecf4fe2-1ee6-0310-87b1-e25e094e27de

4 files changed, 163 insertions, 21 deletions

diff --git a/activerecord/CHANGELOG b/activerecord/CHANGELOG
index d894dd9ee3..1bc3af1987 100644
--- a/activerecord/CHANGELOG
+++ b/activerecord/CHANGELOG
@@ -1,5 +1,7 @@
 *SVN*
 
+* Fixed validates_length_of to work on UTF-8 strings by using characters instead of bytes #3699 [Masao Mutoh]
+
 * Fixed that reflections would bleed across class boundaries in single-table inheritance setups #3796 [lars@pind.com]
 
 * Added calculations: Base.count, Base.average, Base.sum, Base.minimum, Base.maxmium, and the generic Base.calculate. All can be used with :group and :having. Calculations and statitics need no longer require custom SQL. #3958 [Rick Olson]. Examples:
diff --git a/activerecord/lib/active_record/validations.rb b/activerecord/lib/active_record/validations.rb
index f733620298..950cf54b12 100755
--- a/activerecord/lib/active_record/validations.rb
+++ b/activerecord/lib/active_record/validations.rb
@@ -439,31 +439,35 @@ module ActiveRecord
         option_value = options[range_options.first]
 
         case option
-        when :within, :in
-          raise ArgumentError, ":#{option} must be a Range" unless option_value.is_a?(Range)
-
-          too_short = options[:too_short] % option_value.begin
-          too_long  = options[:too_long]  % option_value.end
-
-          validates_each(attrs, options) do |record, attr, value|
-            if value.nil? or value.size < option_value.begin
-              record.errors.add(attr, too_short)
-            elsif value.size > option_value.end
-              record.errors.add(attr, too_long)
+          when :within, :in
+            raise ArgumentError, ":#{option} must be a Range" unless option_value.is_a?(Range)
+
+            too_short = options[:too_short] % option_value.begin
+            too_long  = options[:too_long]  % option_value.end
+
+            validates_each(attrs, options) do |record, attr, value|
+              if value.nil? or value.split(//).size < option_value.begin
+                record.errors.add(attr, too_short)
+              elsif value.split(//).size > option_value.end
+                record.errors.add(attr, too_long)
+              end
             end
-          end
-        when :is, :minimum, :maximum
-          raise ArgumentError, ":#{option} must be a nonnegative Integer" unless option_value.is_a?(Integer) and option_value >= 0
+          when :is, :minimum, :maximum
+            raise ArgumentError, ":#{option} must be a nonnegative Integer" unless option_value.is_a?(Integer) and option_value >= 0
 
-          # Declare different validations per option.
-          validity_checks = { :is => "==", :minimum => ">=", :maximum => "<=" }
-          message_options = { :is => :wrong_length, :minimum => :too_short, :maximum => :too_long }
+            # Declare different validations per option.
+            validity_checks = { :is => "==", :minimum => ">=", :maximum => "<=" }
+            message_options = { :is => :wrong_length, :minimum => :too_short, :maximum => :too_long }
 
-          message = (options[:message] || options[message_options[option]]) % option_value
+            message = (options[:message] || options[message_options[option]]) % option_value
 
-          validates_each(attrs, options) do |record, attr, value|
-            record.errors.add(attr, message) unless !value.nil? and value.size.method(validity_checks[option])[option_value]
-          end
+            validates_each(attrs, options) do |record, attr, value|
+              if value.kind_of?(String)
+                record.errors.add(attr, message) unless !value.nil? and value.split(//).size.method(validity_checks[option])[option_value]
+              else
+                record.errors.add(attr, message) unless !value.nil? and value.size.method(validity_checks[option])[option_value]
+              end
+            end
         end
       end
 
diff --git a/activerecord/test/connections/native_mysql/connection.rb b/activerecord/test/connections/native_mysql/connection.rb
index dea337567b..b665a6b449 100644
--- a/activerecord/test/connections/native_mysql/connection.rb
+++ b/activerecord/test/connections/native_mysql/connection.rb
@@ -10,6 +10,7 @@ db2 = 'activerecord_unittest2'
 ActiveRecord::Base.establish_connection(
   :adapter  => "mysql",
   :username => "rails",
+  :encoding => "utf8",
   :database => db1
 )
 
diff --git a/activerecord/test/validations_test.rb b/activerecord/test/validations_test.rb
index 811585b203..4356ea7428 100755
--- a/activerecord/test/validations_test.rb
+++ b/activerecord/test/validations_test.rb
@@ -614,6 +614,141 @@ class ValidationsTest < Test::Unit::TestCase
     assert_equal "hoo 5", t.errors["title"]
   end
 
+  def kcode_scope(kcode)
+    orig_kcode = $KCODE
+    $KCODE = kcode
+    begin
+      yield
+    ensure
+      $KCODE = orig_kcode
+    end
+  end
+
+  def test_validates_length_of_using_minimum_utf8
+    kcode_scope('UTF8') do
+      Topic.validates_length_of :title, :minimum => 5
+
+      t = Topic.create("title" => "一二三四五", "content" => "whatever")
+      assert t.valid?
+
+      t.title = "一二三四"
+      assert !t.valid?
+      assert t.errors.on(:title)
+      assert_equal "is too short (min is 5 characters)", t.errors["title"]
+    end
+  end
+
+  def test_validates_length_of_using_maximum_utf8
+    kcode_scope('UTF8') do
+      Topic.validates_length_of :title, :maximum => 5
+
+      t = Topic.create("title" => "一二三四五", "content" => "whatever")
+      assert t.valid?
+      
+      t.title = "一二34五六"
+      assert !t.valid?
+      assert t.errors.on(:title)
+      assert_equal "is too long (max is 5 characters)", t.errors["title"]
+    end
+  end
+
+  def test_validates_length_of_using_within_utf8
+    kcode_scope('UTF8') do
+      Topic.validates_length_of(:title, :content, :within => 3..5)
+
+      t = Topic.new("title" => "一二", "content" => "12三四五六七")
+      assert !t.valid?
+      assert_equal "is too short (min is 3 characters)", t.errors.on(:title)
+      assert_equal "is too long (max is 5 characters)", t.errors.on(:content)
+      t.title = "一二三"
+      t.content  = "12三"
+      assert t.valid?
+    end
+  end
+
+  def test_optionally_validates_length_of_using_within_utf8
+    kcode_scope('UTF8') do
+      Topic.validates_length_of :title, :content, :within => 3..5, :allow_nil => true
+
+      t = Topic.create('title' => '一二三', 'content' => '一二三四五')
+      assert t.valid?
+
+      t.title = nil
+      assert t.valid?
+    end
+  end
+
+  def test_optionally_validates_length_of_using_within_on_create_utf8
+    kcode_scope('UTF8') do
+      Topic.validates_length_of :title, :content, :within => 5..10, :on => :create, :too_long => "長すぎます: %d"
+
+      t = Topic.create("title" => "一二三四五六七八九十A", "content" => "whatever")
+      assert !t.save
+      assert t.errors.on(:title)
+      assert_equal "長すぎます: 10", t.errors[:title]
+      
+      t.title = "一二三四五六七八九"
+      assert t.save
+      
+      t.title = "一二3"
+      assert t.save
+      
+      t.content = "一二三四五六七八九十"
+      assert t.save
+      
+      t.content = t.title = "一二三四五六"
+      assert t.save
+    end
+  end
+
+  def test_optionally_validates_length_of_using_within_on_update_utf8
+    kcode_scope('UTF8') do    
+      Topic.validates_length_of :title, :content, :within => 5..10, :on => :update, :too_short => "短すぎます: %d"
+
+      t = Topic.create("title" => "一二三4", "content" => "whatever")
+      assert !t.save
+      assert t.errors.on(:title)
+      
+      t.title = "1二三4"
+      assert !t.save
+      assert t.errors.on(:title)
+      assert_equal "短すぎます: 5", t.errors[:title]
+      
+      t.title = "valid"
+      t.content = "一二三四五六七八九十A"
+      assert !t.save
+      assert t.errors.on(:content)
+      
+      t.content = "一二345"
+      assert t.save
+    end
+  end
+
+  def test_validates_length_of_using_is_utf8
+    kcode_scope('UTF8') do
+      Topic.validates_length_of :title, :is => 5
+
+      t = Topic.create("title" => "一二345", "content" => "whatever")
+      assert t.valid?
+
+      t.title = "一二345六"
+      assert !t.valid?
+      assert t.errors.on(:title)
+      assert_equal "is the wrong length (should be 5 characters)", t.errors["title"]
+    end
+  end
+
+  def test_validates_size_of_association_utf8
+    kcode_scope('UTF8') do
+      assert_nothing_raised { Topic.validates_size_of :replies, :minimum => 1 }
+      t = Topic.new('title' => 'あいうえお', 'content' => 'かきくけこ')
+      assert !t.save
+      assert t.errors.on(:replies)
+      t.replies.create('title' => 'あいうえお', 'content' => 'かきくけこ')
+      assert t.valid?
+    end
+  end
+
   def test_validates_associated_many
     Topic.validates_associated( :replies )
     t = Topic.create("title" => "uhohuhoh", "content" => "whatever")