Skip to content

Commit 794d6e1

Browse files
THausherrTHausherr
committed
PDFBOX-6126: reroute inner rdf:description
git-svn-id: https://svn.apache.org/repos/asf/pdfbox/trunk@1930635 13f79535-47bb-0310-9956-ffa450edef68
1 parent 6b00f96 commit 794d6e1

File tree

2 files changed

+78
-1
lines changed

2 files changed

+78
-1
lines changed

‎xmpbox/src/main/java/org/apache/xmpbox/xml/DomXmpParser.java‎

Lines changed: 6 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -687,8 +687,13 @@ private AbstractStructuredType parseLiDescription(XMPMetadata xmp, QName descrip
687687
// The list is empty
688688
return tryParseAttributesAsProperties(xmp, liDescriptionElement, tm, null, null, descriptor);
689689
}
690-
// Instantiate abstract structured type with hint from first element
691690
Element firstLiDescriptionElementChild = liDescriptionElementChildren.get(0);
691+
if ("rdf:Description".equals(firstLiDescriptionElementChild.getTagName()))
692+
{
693+
// PDFBOX-6126: "<rdf:Description" as child of "<rdf:li"
694+
return parseLiDescription(xmp, descriptor, firstLiDescriptionElementChild);
695+
}
696+
// Instantiate abstract structured type with hint from first element
692697
nsFinder.push(firstLiDescriptionElementChild);
693698
QName qName = DomHelper.getQName(firstLiDescriptionElementChild);
694699
PropertyType ctype = checkPropertyDefinition(xmp, qName);

‎xmpbox/src/test/java/org/apache/xmpbox/xml/DomXmpParserTest.java‎

Lines changed: 72 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -21,11 +21,13 @@
2121

2222
package org.apache.xmpbox.xml;
2323

24+
import java.io.ByteArrayOutputStream;
2425
import java.io.IOException;
2526
import java.io.InputStream;
2627
import java.nio.charset.StandardCharsets;
2728
import java.util.Calendar;
2829
import java.util.List;
30+
import javax.xml.transform.TransformerException;
2931

3032
import org.apache.xmpbox.XMPMetadata;
3133
import org.apache.xmpbox.schema.AdobePDFSchema;
@@ -788,4 +790,74 @@ void testBadSchema() throws XmpParsingException
788790
() -> xmpParser1.parse(s.getBytes(StandardCharsets.UTF_8)));
789791
assertEquals("This namespace is not from a schema: http://ns.adobe.com/xap/1.0/sType/Job#", ex.getMessage());
790792
}
793+
794+
@Test
795+
void testPDFBOX6126() throws XmpParsingException, BadFieldValueException, TransformerException
796+
{
797+
// XMP originally from PDFBOX-4325, had this exception:
798+
// Cannot find a definition for the namespace http://www.w3.org/1999/02/22-rdf-syntax-ns#, property: rdf:Description
799+
// Cause: "<rdf:Description" as child of <rdf:li .
800+
String s = "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\n" +
801+
"<?xpacket begin=\"\" id=\"W5M0MpCehiHzreSzNTczkc9d\"?>\n" +
802+
"<x:xmpmeta xmlns:x=\"adobe:ns:meta/\"\n" +
803+
" x:xmptk=\"Adobe XMP Core 5.1.0-jc003\">\n" +
804+
" <rdf:RDF xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\">\n" +
805+
" <rdf:Description xmlns:dc=\"http://purl.org/dc/elements/1.1/\"\n" +
806+
" xmlns:pdf=\"http://ns.adobe.com/pdf/1.3/\"\n" +
807+
" xmlns:pdfaExtension=\"http://www.aiim.org/pdfa/ns/extension/\"\n" +
808+
" xmlns:pdfaProperty=\"http://www.aiim.org/pdfa/ns/property#\"\n" +
809+
" xmlns:pdfaSchema=\"http://www.aiim.org/pdfa/ns/schema#\"\n" +
810+
" xmlns:pdfaid=\"http://www.aiim.org/pdfa/ns/id/\"\n" +
811+
" xmlns:pdfuaid=\"http://www.aiim.org/pdfua/ns/id/\"\n" +
812+
" xmlns:xmp=\"http://ns.adobe.com/xap/1.0/\"\n" +
813+
" dc:format=\"application/pdf\"\n" +
814+
" pdf:Producer=\"iText® 5.5.13 ©2000-2018 iText Group NV (AGPL-version)\"\n" +
815+
" pdfaid:conformance=\"B\"\n" +
816+
" pdfaid:part=\"1\"\n" +
817+
" rdf:about=\"\"\n" +
818+
" xmp:CreateDate=\"2018-09-24T09:00:57+02:00\"\n" +
819+
" xmp:ModifyDate=\"2018-09-24T09:00:57+02:00\">\n" +
820+
" <pdfaExtension:schemas>\n" +
821+
" <rdf:Bag>\n" +
822+
" <rdf:li rdf:parseType=\"Resource\">\n" +
823+
" <rdf:Description pdfaSchema:namespaceURI=\"http://www.aiim.org/pdfua/ns/id/\"\n" +
824+
" pdfaSchema:prefix=\"pdfuaid\"\n" +
825+
" pdfaSchema:schema=\"PDF/UA identification schema\">\n" +
826+
" <pdfaSchema:property>\n" +
827+
" <rdf:Seq>\n" +
828+
" <rdf:li pdfaProperty:category=\"internal\"\n" +
829+
" pdfaProperty:description=\"PDF/UA version identifier\"\n" +
830+
" pdfaProperty:name=\"part\"\n" +
831+
" pdfaProperty:valueType=\"Integer\"/>\n" +
832+
" <rdf:li pdfaProperty:category=\"internal\"\n" +
833+
" pdfaProperty:description=\"PDF/UA amendment identifier\"\n" +
834+
" pdfaProperty:name=\"amd\"\n" +
835+
" pdfaProperty:valueType=\"Text\"/>\n" +
836+
" <rdf:li pdfaProperty:category=\"internal\"\n" +
837+
" pdfaProperty:description=\"PDF/UA corrigenda identifier\"\n" +
838+
" pdfaProperty:name=\"corr\"\n" +
839+
" pdfaProperty:valueType=\"Text\"/>\n" +
840+
" </rdf:Seq>\n" +
841+
" </pdfaSchema:property>\n" +
842+
" </rdf:Description>\n" +
843+
" </rdf:li>\n" +
844+
" </rdf:Bag>\n" +
845+
" </pdfaExtension:schemas>\n" +
846+
" <pdfuaid:part>1</pdfuaid:part>\n" +
847+
" </rdf:Description>\n" +
848+
" </rdf:RDF>\n" +
849+
"</x:xmpmeta><?xpacket end=\"w\"?>";
850+
DomXmpParser xmpParser1 = new DomXmpParser();
851+
XMPMetadata xmp1 = xmpParser1.parse(s.getBytes(StandardCharsets.UTF_8));
852+
XMPSchema uaSchema1 = xmp1.getSchema("http://www.aiim.org/pdfua/ns/id/");
853+
assertEquals(1, uaSchema1.getIntegerPropertyValueAsSimple("part"));
854+
XmpSerializer serializer = new XmpSerializer();
855+
ByteArrayOutputStream baos = new ByteArrayOutputStream();
856+
// make sure that nothing is lost in serialization
857+
serializer.serialize(xmp1, baos, true);
858+
DomXmpParser xmpParser2 = new DomXmpParser();
859+
XMPMetadata xmp2 = xmpParser2.parse(baos.toByteArray());
860+
XMPSchema uaSchema2 = xmp2.getSchema("http://www.aiim.org/pdfua/ns/id/");
861+
assertEquals(1, uaSchema2.getIntegerPropertyValueAsSimple("part"));
862+
}
791863
}

0 commit comments

Comments
 (0)