Name: dfC67450 Date: 01/26/2000



java.net.URLDecoder.decode(String s) works in different manner with some strings
which are not in "x-www-form-urlencoded" MIME format. Javadoc spec does not
specify behaviour for this case.

This method throws StringIndexOutOfBoundsException with "%" or "%A" and
IllegalArgumentException for "%xy" and does not throw any exception when passing
characters that should not appear in "x-www-form-urlencoded" MIME format.

Here is the test demonstrating the bug:

---------------------------------------------
import java.net.*;


public class Test {
    public static void main (String args[]){

        String s[] = {"%", "%A", "%xy", "#", "X\u00aaY"};
        for (int i = 0; i < s.length; i++) {
            try {
                String decoded = URLDecoder.decode(s[i]);
                System.out.println("\"" + s[i] + "\" --> \"" + decoded + "\"");
            } catch (Exception e) {
                System.out.println("\"" + s[i] + "\" --> " + e);
            }
        }
    }
}


------------- output ---------
"%" --> java.lang.StringIndexOutOfBoundsException: String index out of range: 3
"%A" --> java.lang.StringIndexOutOfBoundsException: String index out of range: 3
"%xy" --> java.lang.IllegalArgumentException
"#" --> "#"
"XªY" --> "XªY"
----------------------------------------------------------

======================================================================