Possible bug in exclusive canonicalization

[Elliotte Harold]

When processing xmltest/invalid/002.xml from the W3C XML conformance 
test suite using exclusive canonicalization with comments in 
XML-Security 1.2, the results are:

<doc></doc><doc></doc>
However, the input document is:
<!DOCTYPE doc SYSTEM "002.ent">
<doc></doc>
002.ent is
<!ENTITY % e "(#PCDATA">
<!ELEMENT doc %e;)>
The result should be
<doc></doc>
This bug is seen by my program which follows. It would be nice if there 
were a sample program bundled with XML_security that could test this 
standalone. is there any such program?

import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.OutputStream;
import java.net.URI;
import java.net.URISyntaxException;
import javax.xml.parsers.DocumentBuilder;
import javax.xml.parsers.DocumentBuilderFactory;
import javax.xml.transform.TransformerException;
import org.w3c.dom.Document;
import org.xml.sax.SAXException;
import nu.xom.Builder;
import nu.xom.Elements;
import org.apache.xml.security.c14n.*;
import org.apache.xml.security.utils.IgnoreAllErrorHandler;
public class ExclusiveC14N {
    public static void main(String args[]) throws Exception {
      org.apache.xml.security.Init.init();
      DocumentBuilderFactory dfactory = 
DocumentBuilderFactory.newInstance();

      dfactory.setNamespaceAware(true);
      dfactory.setValidating(true);
      DocumentBuilder domBuilder = dfactory.newDocumentBuilder();
      // throw away all validation warnings
      domBuilder.setErrorHandler(new IgnoreAllErrorHandler());
      File masterList = new File("data/canonical/xmlconf/xmlconf.xml");
      Builder builder = new Builder();
      nu.xom.Document xmlconf = builder.build(masterList);
      Elements testcases = 
xmlconf.getRootElement().getChildElements("TESTCASES");

      processTestCases(domBuilder, testcases);
    }
    private static void processTestCases(DocumentBuilder domBuilder, 
Elements testcases)
      throws URISyntaxException, SAXException, IOException, 
TransformerException, Exception {

        for (int i = 0; i < testcases.size(); i++) {
              nu.xom.Element testcase = testcases.get(i);
              Elements tests = testcase.getChildElements("TEST");
              processTests(domBuilder, tests);
              Elements level2 = testcase.getChildElements("TESTCASES");
              // need to be recursive to handle recursive IBM test cases
              processTestCases(domBuilder, level2);
          }
    }
    private static void processTests(DocumentBuilder domBuilder, 
Elements tests)
      throws URISyntaxException, SAXException, IOException, 
TransformerException,
             Exception {

        Canonicalizer canonicalizer = 
Canonicalizer.getInstance(Canonicalizer.ALGO_ID_C14N_EXCL_WITH_COMMENTS);
        for (int j = 0; j < tests.size(); j++) {
              nu.xom.Element test = tests.get(j);
              String namespace = test.getAttributeValue("NAMESPACE");
              if ("no".equals(namespace)) continue;
              String type = test.getAttributeValue("TYPE");
              if ("not-wf".equals(type)) continue;
              String uri = test.getAttributeValue("URI");
              String base = test.getBaseURI();
              URI baseURI= new URI(base);
              URI testURI = baseURI.resolve(uri);
              Document testdoc = domBuilder.parse(testURI.toString());
              byte[] result = canonicalizer.canonicalizeSubtree(testdoc);
              String outputFilename = testURI.toString().substring(5) + 
".exc";
              OutputStream out = new FileOutputStream(outputFilename);
              out.write(result);
              out.flush();
              out.close();
              System.out.println(outputFilename);
          }

    }
}

--
Elliotte Rusty Harold  [EMAIL PROTECTED]
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim