Using padding zeros for low unicode code points

Java's native2ascii tool also uses zero padding for low code points.

Refs #5

lib/java-properties/encoding/unicode.rb

@@ -3,7 +3,7 @@ module Encoding
     # Module to encode and decode unicode chars
     # @see JavaProperties::Encoding
     module Unicode
-      
+
       # Marker for encoded unicode chars
       # @return [Regexp]
       UNICODE_MARKER  = /\\[uU]([0-9a-fA-F]{4,5}|10[0-9a-fA-F]{4})/
@@ -42,20 +42,16 @@ def self.encode!(text)
       private
 
       def self.unicode(code)
-        [code].pack("U")
+        code.chr(::Encoding::UTF_8)
       end
 
       def self.hex(codepoint)
         hex  = codepoint.to_s(16)
-        size = hex.size
-        # padding the hex value for uneven digest
-        if (size % 2) == 1
-          "0#{hex}"
-        else
-          hex
-        end
+        size = [4, hex.size].max
+        target_size = size.even? ? size : size+1
+        hex.rjust(target_size, '0')
       end
 
     end
   end
-end
+end

spec/fixtures/test_out.properties

@@ -8,4 +8,4 @@ it\:em6=item6
 item7=line 1 line 2 line 3
 item8=line 1 line 2 line 3
 item9=line 1 line 2 line 3
-item10=test\n\ttest\u05d4 test\n\ttest test\n\ttest = test
+item10=test\n\ttest\u05d4 test\n\ttest test\n\ttest\u00fc = test

spec/fixtures/test_out_skip_separators.properties

@@ -8,4 +8,4 @@ it:em6=item6
 item7=line 1 line 2 line 3
 item8=line 1 line 2 line 3
 item9=line 1 line 2 line 3
-item10=test\n\ttest\u05d4 test\n\ttest test\n\ttest = test
+item10=test\n\ttest\u05d4 test\n\ttest test\n\ttest\u00fc = test

spec/fixtures/test_out_skip_special_chars.properties

@@ -11,4 +11,4 @@ item9=line 1 line 2 line 3
 item10=test
 	test\u05d4 test
 	test test
-	test = test
+	test\u00fc = test

spec/fixtures/test_out_skip_unicode.properties

@@ -8,4 +8,4 @@ it\:em6=item6
 item7=line 1 line 2 line 3
 item8=line 1 line 2 line 3
 item9=line 1 line 2 line 3
-item10=test\n\ttestה test\n\ttest test\n\ttest = test
+item10=test\n\ttestה test\n\ttest test\n\ttestü = test

spec/java-properties/encoding/unicode_spec.rb

@@ -4,9 +4,9 @@
 describe JavaProperties::Encoding::Unicode do
   subject{ JavaProperties::Encoding::Unicode }
 
-  let(:encoded) { 'this is some \u0024 text \U05D4 with unicode' }
-  let(:encoded_normalized) { 'this is some $ text \u05d4 with unicode' }
-  let(:decoded) { 'this is some $ text ה with unicode' }
+  let(:encoded) { 'this is some \u0024 text \U05D4 with unicode \u00fc' }
+  let(:encoded_normalized) { 'this is some $ text \u05d4 with unicode \u00fc' }
+  let(:decoded) { 'this is some $ text ה with unicode ü' }
 
   it "decodes unicode chars" do
     subject.decode!(encoded.dup).must_equal decoded
@@ -16,14 +16,15 @@
     subject.encode!(decoded.dup).must_equal encoded_normalized
   end
 
-  it "encodes unicode chars but has 2-based hex size" do
+  it "encodes unicode chars but has 2-based hex size, padded to at least 4" do
+    subject.encode!("ü").must_equal '\u00fc'
     subject.encode!("ה").must_equal '\u05d4'
     subject.encode!("ᘓ").must_equal '\u1613'
   end
 
   it "decodes and encodes" do
     encoded  = subject.encode!(decoded.dup)
-    deconded = subject.decode!(encoded.dup)
-    deconded.must_equal decoded
+    dec = subject.decode!(encoded.dup)
+    dec.must_equal decoded
   end
-end
+end

spec/java-properties/generating/generator_spec.rb

@@ -16,7 +16,7 @@
       :item7 => "line 1 line 2 line 3",
       :item8 => "line 1 line 2 line 3",
       :item9 => "line 1 line 2 line 3",
-      :item10 => "test\n\ttestה test\n\ttest test\n\ttest = test"
+      :item10 => "test\n\ttestה test\n\ttest test\n\ttestü = test"
     }
   end
 
@@ -44,4 +44,4 @@
     content.must_equal expected
   end
 
-end
+end