NullPointerException in org.apache.xerces.utils.StringPool (example included!)

[Michael Bayer]

This bug applies to versions 1.2.* and 1.3.* of Xerces.  It does not exist
in 1.0.* .

It seems the XMLDTDScanner is relying upon the SystemId property of
org.xml.sax.InputSource when an XML document contains external unparsed
entities (example XML is below).  If the SystemId is not set, you get a
NullPointerException at
org.apache.xerces.utils.StringPool.addSymbol(StringPool.java:348).  

Constructing an InputSource from a Reader or InputStream will create this
behavior, i.e. getSystemId() returns null.  The method setSystemId(String)
can be called to manually set the systemid as a workaround.

This requirement should be either fixed, or properly handled with a
message (and documented).

Example XML:

<?xml version="1.0" encoding="UTF-8"?>

<!DOCTYPE test [
<!ENTITY hello SYSTEM "hello.gif" NDATA gif>
]>


<tag>hello world</tag>




Example program:

import org.xml.sax.*;
import org.xml.sax.helpers.*;
import java.io.*;

public class TestXerces {
        public static void main(String [] argv) {
                String uri = argv[0];

                try {
                        XMLReader reader =
XMLReaderFactory.createXMLReader("org.apache.xerces.parsers.SAXParser");

                        InputSource i = new InputSource(new
FileReader(uri));

                        // uncomment this line to make it work
                        //i.setSystemId(uri);

                        reader.parse(i);

                }
                catch (SAXException se) {
                        if (se.getException() !=
null) se.getException().printStackTrace();
                        se.printStackTrace();
                }
                catch (Exception e) {
                        e.printStackTrace();
                }

        }
}

usage: 

java TestXerces <filename>