1 files changed, 24 insertions, 0 deletions
diff --git a/tests/bugs199/github_145/add_modules/UseJDKExtendedCharsets.java b/tests/bugs199/github_145/add_modules/UseJDKExtendedCharsets.java
new file mode 100644
index 000000000..62a351868
--- /dev/null
+++ b/tests/bugs199/github_145/add_modules/UseJDKExtendedCharsets.java
@@ -0,0 +1,24 @@
+import java.nio.charset.Charset;
+import sun.nio.cs.ext.ExtendedCharsets;
+
+/**
+ * Note that the Windows JDK knows many more extended charsets than the Linux one. Originally, this test was using
+ * "hebrew", but that yielded failing GitHub CI tests due to the unavailability of the character set there. I actually
+ * had to inspect a Linux 'lib/modules' file using 'jimage' in order to filter for available 'ISO*' classes in the
+ * extended character set package. So if this test ever breaks again on any OS platform, you know where to look for the
+ * root cause and have a clue how to fix the test.
+ */
+public class UseJDKExtendedCharsets {
+  static ExtendedCharsets charsets = new ExtendedCharsets();
+  static Charset iso2022jp = charsets.charsetForName("ISO-2022-JP");
+  static Charset jis = charsets.charsetForName("jis");
+  static Charset jis_encoding = charsets.charsetForName("jis_encoding");
+
+  public static void main(String[] args) {
+    // The 3 charsets are aliases of each other
+    assert iso2022jp != null;
+    System.out.println(iso2022jp);
+    assert iso2022jp.equals(jis);
+    assert iso2022jp.equals(jis_encoding);
+  }
+}