When my runtime environment is a Chinese encoding environment, the list of error messages returned is garbled

[GodMeowIceSun]

input

critical code

result

individual opinion

There is a problem with different character sets, so I think the output content of the default character set should be provided.

[GodMeowIceSun]

This issue occurs when I use dependencies in Maven, but it no longer exists after I compile locally

[GodMeowIceSun]

No, it's not that there's no problem with local compilation, but rather that getErrorMessages has been called multiple times

[GodMeowIceSun]

After applying changes and compiling locally, There's no garbled code left.

result

[davewichers]

Per @kwwall - I suppose the correct thing to do to address this is for someone to correct all the AntiSamy properties files and XML files as per the Oracle JDK 9 document that @GodMeowIceSun referenced, and use UTF-8 instead of ISO-8859-1. (And if that doesn't work, use a system property like I suggested earlier in the thread for: #457

[davewichers]

@spassarop - Do you think you'd be able to work on this at some point? It's been dangling out there for a while.

[spassarop]

@GodMeowIceSun I made changes on PR #569. Could you test that project branch on your environment?

You can reproduce with your own code but you could also run this on a JUnit test case (in AntiSamyTest.java) to print all messages in the bundle:

    Enumeration<String> keys = messages.getKeys();
    while (keys.hasMoreElements()) {
      String key = keys.nextElement();
      String value = messages.getString(key);
      System.out.println(key + " = " + value);
    }