Java挪威小数点分隔符在Java 6和7中有所不同

Question

The following snippet behaves different in Java 6 than Java 7:

final Locale locale = new Locale("nb", "NO");
System.out.println(locale.getDisplayLanguage()); // Norwegian Bokmål

final DecimalFormatSymbols dfs = new DecimalFormatSymbols(locale);
System.out.println(dfs.getDecimalSeparator());   // Java 6: .
                                                 // Java 7: ,

Why is that? Is this change documented somewhere?

Answer 1

Java 6 had a number of issues regarding Locale s, and this may well be one of them. Certainly, the correct separator for the Norway locale is , .

The Oracle bug database does show quite a few bugs related to NO locale...

Answer 2

According to JDK 6 and JRE 6 Supported Locales and JDK 7 and JRE 7 Supported Locales , the correct/supported syntax for selecting Norwegian Bokmål is "no"/"NO".

new Locale("no", "NO") gives the correct result under both Java 6 and Java 7.

Answer 3

CLDR

In Java 9 and later, the OpenJDK -based implementations of Java switched to using locale data obtained from the Unicode Consortium 's Common Locale Data Repository (CLDR) (see Wikipedia ) by default.

This switch provides much richer locale data than previously seen in Java. And this switch may mean that you see a change in behavior for some aspects of some locales.

See: JEP 252: Use CLDR Locale Data by Default

COMMA in Java 10

When I run this code:

    System.out.println( "java.version: " + System.getProperty( "java.version" ) );

    final Locale locale = new Locale( "nb" , "NO" );
    System.out.println( locale.getDisplayLanguage() ); // Norwegian Bokmål

    final DecimalFormatSymbols dfs = new DecimalFormatSymbols( locale );
    System.out.println( dfs.getDecimalSeparator() );

I get the COMMA character.

java.version: 10.0.2

Norwegian Bokmål

,