Author: tilman
Date: Fri Dec 19 11:29:57 2025
New Revision: 1930725
Log:
PDFBOX-5977: improve namespace handling
Modified:
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/impl/XMLUtil.java
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/xmp/XMPMetadata.java
pdfbox/branches/1.8/jempbox/src/test/java/org/apache/jempbox/xmp/XMPMetadataTest.java
Modified:
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/impl/XMLUtil.java
==============================================================================
---
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/impl/XMLUtil.java
Fri Dec 19 08:34:41 2025 (r1930724)
+++
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/impl/XMLUtil.java
Fri Dec 19 11:29:57 2025 (r1930725)
@@ -90,6 +90,7 @@ public class XMLUtil
builderFactory.setFeature(
"http://apache.org/xml/features/nonvalidating/load-external-dtd";, false);
builderFactory.setXIncludeAware(false);
+ builderFactory.setNamespaceAware(true); // PDFBOX-5977
DocumentBuilder builder = builderFactory.newDocumentBuilder();
// prevents validation messages polluting the console
builder.setErrorHandler(null);
Modified:
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/xmp/XMPMetadata.java
==============================================================================
---
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/xmp/XMPMetadata.java
Fri Dec 19 08:34:41 2025 (r1930724)
+++
pdfbox/branches/1.8/jempbox/src/main/java/org/apache/jempbox/xmp/XMPMetadata.java
Fri Dec 19 11:29:57 2025 (r1930725)
@@ -617,30 +617,18 @@ public class XMPMetadata
String value = attribute.getNodeValue();
if (name.startsWith("xmlns:") && nsMappings.containsKey(value))
{
- Class<?> schemaClass = nsMappings.get(value);
- try
- {
- Constructor<?> ctor = schemaClass
- .getDeclaredConstructor(new Class[] {
Element.class,
- String.class });
- retval.add((XMPSchema)ctor.newInstance(new Object[] {
schema,
- name.substring(6) }));
- found = true;
- }
- catch(NoSuchMethodException e)
- {
- throw new IOException(
- "Error: Class "
- + schemaClass.getName()
- + " must have a constructor with the
signature of "
- + schemaClass.getName()
- + "( org.w3c.dom.Element,
java.lang.String )");
- }
- catch(Exception e)
- {
- e.printStackTrace();
- throw new IOException(e.getMessage());
- }
+ String prefix = name.substring(6);
+ retval.add(createXMPSchema(value, schema, prefix));
+ found = true;
+ }
+ // PDFBOX-5977
+ else if (attribute.getNamespaceURI() != null &&
+ nsMappings.containsKey(attribute.getNamespaceURI()) &&
+ name.contains(":"))
+ {
+ String prefix = name.substring(0, name.indexOf(':'));
+ retval.add(createXMPSchema(attribute.getNamespaceURI(),
schema, prefix));
+ found = true;
}
}
if (!found)
@@ -651,6 +639,32 @@ public class XMPMetadata
return retval;
}
+ private XMPSchema createXMPSchema(String value, Element schemaElement,
String prefix) throws IOException
+ {
+ Class<?> schemaClass = nsMappings.get(value);
+ try
+ {
+ Constructor<?> ctor = schemaClass
+ .getDeclaredConstructor(new Class[] { Element.class,
+ String.class });
+ return (XMPSchema) ctor.newInstance(new Object[] { schemaElement,
prefix });
+ }
+ catch(NoSuchMethodException e)
+ {
+ throw new IOException(
+ "Error: Class "
+ + schemaClass.getName()
+ + " must have a constructor with the signature of "
+ + schemaClass.getName()
+ + "( org.w3c.dom.Element, java.lang.String )");
+ }
+ catch(Exception e)
+ {
+ e.printStackTrace();
+ throw new IOException(e.getMessage());
+ }
+ }
+
/**
* Will return all schemas that fit the given namespaceURI. Which is only
* done by using the namespace mapping (nsMapping) and not by actually
Modified:
pdfbox/branches/1.8/jempbox/src/test/java/org/apache/jempbox/xmp/XMPMetadataTest.java
==============================================================================
---
pdfbox/branches/1.8/jempbox/src/test/java/org/apache/jempbox/xmp/XMPMetadataTest.java
Fri Dec 19 08:34:41 2025 (r1930724)
+++
pdfbox/branches/1.8/jempbox/src/test/java/org/apache/jempbox/xmp/XMPMetadataTest.java
Fri Dec 19 11:29:57 2025 (r1930725)
@@ -16,12 +16,14 @@
*/
package org.apache.jempbox.xmp;
+import java.io.ByteArrayInputStream;
import java.io.IOException;
import java.io.StringReader;
import java.util.Calendar;
import java.util.GregorianCalendar;
import junit.framework.TestCase;
+import org.apache.jempbox.xmp.pdfa.XMPSchemaPDFAId;
import org.xml.sax.InputSource;
@@ -139,4 +141,18 @@ public class XMPMetadataTest extends Tes
assertEquals("uuid:2c7eb5da-9210-4666-8cef-e02ef6631c5e",
mediaManagementSchema.getInstanceID());
}
-}
+ public void testPDFBOX5977() throws IOException
+ {
+ String s = "<?xml version=\"1.0\" encoding=\"UTF-8\"
standalone=\"no\"?>\n" +
+ "<?xpacket begin=\"\"
id=\"W5M0MpCehiHzreSzNTczkc9d\"?><rdf:RDF
xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\"; " +
+ "xmlns:pdf=\"http://ns.adobe.com/pdf/1.3/\";
xmlns:pdfaid=\"http://www.aiim.org/pdfa/ns/id/\";>\n" +
+ " <rdf:Description pdfaid:conformance=\"B\"
pdfaid:part=\"3\" rdf:about=\"\"/>\n" +
+ " <rdf:Description pdf:Producer=\"WeasyPrint 64.1\"
rdf:about=\"\"/>\n" +
+ "</rdf:RDF><?xpacket end=\"r\"?>";
+ XMPMetadata xmp = XMPMetadata.load(new
ByteArrayInputStream(s.getBytes()));
+ xmp.addXMLNSMapping(XMPSchemaPDFAId.NAMESPACE, XMPSchemaPDFAId.class);
+ XMPSchemaPDFAId schema = (XMPSchemaPDFAId)
xmp.getSchemaByClass(XMPSchemaPDFAId.class);
+ assertEquals("B", schema.getConformance());
+ assertEquals((Integer) 3, schema.getPart());
+ }
+}
\ No newline at end of file
