FULL PRODUCT VERSION :
java version "1.6.0-beta"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.6.0-beta-b59g)
Java HotSpot(TM) Client VM (build 1.6.0-beta-b59g, mixed mode, sharing)

ADDITIONAL OS VERSION INFORMATION :
Windows XP home edition sp2 Korean

A DESCRIPTION OF THE PROBLEM :
If a xml file has a DOCTYPE and the SYSTEMID of DOCTYPE has a korean character relative url, MalformedException occurs during parsing. I guess that the same problem will occur if japanese or chinese character is used in SYSTEMID. This does not occur in JDK 1.4.2 with its internal crimson parser.

STEPS TO FOLLOW TO REPRODUCE THE PROBLEM :
Test case needs only 3 files.
1. Test.java
2. .dtd (file name is a Korean character)
3. test.xml
all of which are in the same directory.

in console, please prompt as follows
1. javac Test.java
2. java Test

EXPECTED VERSUS ACTUAL BEHAVIOR :
EXPECTED -
DocumentBuilder successfully parses the xml file according to the dtd file which is the Korean file name.
ACTUAL -
java.net.MalformedURLException exception thrown

ERROR MESSAGES/STACK TRACES THAT OCCUR :
Attached seperatly

REPRODUCIBILITY :
This bug can be reproduced always.

---------- BEGIN SOURCE ----------
Attached seperatly
---------- END SOURCE ----------

CUSTOMER SUBMITTED WORKAROUND :
We've suggested the bug patch at apache xerces project (XERCESJ-1156). (http://issues.apache.org/jira/browse/XERCESJ-1156)