Why UTF-8 as native2ascii encoding option rather than ISO-8859-1?

Question

When create Properties class in Java, it requires the input stream must be encoded in ISO-8859-1. But in practice, I use UTF-8 to convert non-Latin properties files by native2ascii tool rather than ISO-8859-1.

• According to JDK doc, input stream should be encoded ISO-8859-1. That's to say, the source file is encoding in ISO-8859-1.
• According to decoding and encoding should using the same ISO-8859-1, Properties class in Java should decode using ISO-8859-1.
• According to testing, actually we use UTF-8 as encoding option in native2ascii rather than ISO-8859-1. Why?

Test as follow:

• Create test.properties file which contains : "key=Ü"

• Generated ISO-8859-1 property file: key=\u00c3\u009c

    native2ascii -encoding ISO-8859-1 test.properties iso88591.propertie: 
  

• Generated UTF-8 property file: key=\u00dc

    native2ascii -encoding UTF-8 test.properties utf8.properties 
  

• Create Properties to load the two generated property files:

  Properties p = new Properties();
  //InputStream inStream = new FileInputStream("src/test/java/com/active/translation/iso88591.properties");
  InputStream inStream = new FileInputStream("src/test/java/com/active/translation/utf8.properties");
  p.load(inStream);
  
  System.out.println(p.getProperty("key"));
  

• iso88591.propertie result is: Ã

• utf8.properties result is: Ü

Answer:

That -encoding needs to match the actual encoding used in your source file. From the looks of it, that is UTF-8. – Thilo Apr 3 at 2:52

Answer 1

That -encoding needs to match the actual encoding used in your source file. From the looks of it, that is UTF-8. – Thilo Apr 3 at 2:52