You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tika.apache.org by ta...@apache.org on 2014/03/06 18:08:10 UTC
svn commit: r1574964 - in /tika/trunk/tika-parsers/src:
main/java/org/apache/tika/parser/pdf/PDFParser.java
test/java/org/apache/tika/parser/pdf/PDFParserTest.java
test/resources/test-documents/testPDF_twoAuthors.pdf
Author: tallison
Date: Thu Mar 6 17:08:09 2014
New Revision: 1574964
URL: http://svn.apache.org/r1574964
Log:
TIKA-1252: handle multiple authors in PDF xmp metadata
Added:
tika/trunk/tika-parsers/src/test/resources/test-documents/testPDF_twoAuthors.pdf (with props)
Modified:
tika/trunk/tika-parsers/src/main/java/org/apache/tika/parser/pdf/PDFParser.java
tika/trunk/tika-parsers/src/test/java/org/apache/tika/parser/pdf/PDFParserTest.java
Modified: tika/trunk/tika-parsers/src/main/java/org/apache/tika/parser/pdf/PDFParser.java
URL: http://svn.apache.org/viewvc/tika/trunk/tika-parsers/src/main/java/org/apache/tika/parser/pdf/PDFParser.java?rev=1574964&r1=1574963&r2=1574964&view=diff
==============================================================================
--- tika/trunk/tika-parsers/src/main/java/org/apache/tika/parser/pdf/PDFParser.java (original)
+++ tika/trunk/tika-parsers/src/main/java/org/apache/tika/parser/pdf/PDFParser.java Thu Mar 6 17:08:09 2014
@@ -24,6 +24,8 @@ import java.util.Collections;
import java.util.List;
import java.util.Set;
+import org.apache.jempbox.xmp.XMPSchema;
+import org.apache.jempbox.xmp.XMPSchemaDublinCore;
import org.apache.jempbox.xmp.pdfa.XMPSchemaPDFAId;
import org.apache.pdfbox.cos.COSArray;
import org.apache.pdfbox.cos.COSBase;
@@ -160,10 +162,24 @@ public class PDFParser extends AbstractP
private void extractMetadata(PDDocument document, Metadata metadata)
throws TikaException {
+
+ org.apache.jempbox.xmp.XMPMetadata xmp = null;
+ XMPSchemaDublinCore dcSchema = null;
+ try{
+ if (document.getDocumentCatalog().getMetadata() != null){
+ xmp = document.getDocumentCatalog().getMetadata().exportXMPMetadata();
+ }
+ if (xmp != null){
+ dcSchema = xmp.getDublinCoreSchema();
+ }
+ } catch (IOException e){
+ //swallow
+ }
PDDocumentInformation info = document.getDocumentInformation();
metadata.set(PagedText.N_PAGES, document.getNumberOfPages());
- addMetadata(metadata, TikaCoreProperties.TITLE, info.getTitle());
- addMetadata(metadata, TikaCoreProperties.CREATOR, info.getAuthor());
+ extractDublinCoreListItems(metadata, TikaCoreProperties.TITLE, info.getTitle(), dcSchema);
+ extractDublinCoreListItems(metadata, TikaCoreProperties.CREATOR, info.getAuthor(), dcSchema);
+ extractDublinCoreListItems(metadata, TikaCoreProperties.CONTRIBUTOR, null, dcSchema);
addMetadata(metadata, TikaCoreProperties.CREATOR_TOOL, info.getCreator());
addMetadata(metadata, TikaCoreProperties.KEYWORDS, info.getKeywords());
addMetadata(metadata, "producer", info.getProducer());
@@ -215,8 +231,7 @@ public class PDFParser extends AbstractP
MEDIA_TYPE.toString()+"; version="+Float.toString(document.getDocument().getVersion()));
try {
- if( document.getDocumentCatalog().getMetadata() != null ) {
- org.apache.jempbox.xmp.XMPMetadata xmp = document.getDocumentCatalog().getMetadata().exportXMPMetadata();
+ if( xmp != null ) {
xmp.addXMLNSMapping(XMPSchemaPDFAId.NAMESPACE, XMPSchemaPDFAId.class);
XMPSchemaPDFAId pdfaxmp = (XMPSchemaPDFAId) xmp.getSchemaByClass(XMPSchemaPDFAId.class);
if( pdfaxmp != null ) {
@@ -259,6 +274,68 @@ public class PDFParser extends AbstractP
}
}
+ /**
+ * This tries to read a list from a particular property in
+ * XMPSchemaDublinCore.
+ * If it can't find the information, it falls back to the
+ * pdfboxBaseline. The pdfboxBaseline should be the value
+ * that pdfbox returns from its PDDocumentInformation object
+ * (e.g. getAuthor()) This method is designed include the pdfboxBaseline,
+ * and it should not duplicate the pdfboxBaseline.
+ * <p>
+ * Until PDFBOX-1803/TIKA-1233 are fixed, do not call this
+ * on dates!
+ *
+ * @param property
+ * @param pdfBoxBaseline
+ * @param dc
+ * @param metadata
+ */
+ private void extractDublinCoreListItems(Metadata metadata, Property property,
+ String pdfBoxBaseline, XMPSchemaDublinCore dc){
+ //if no dc, add baseline and return
+ if (dc == null){
+ if (pdfBoxBaseline != null && pdfBoxBaseline.length() > 0){
+ addMetadata(metadata, property, pdfBoxBaseline);
+ }
+ return;
+ }
+ List<String> items = getXMPBagOrSeqList(dc, property.getName());
+ if (items == null){
+ if (pdfBoxBaseline != null && pdfBoxBaseline.length() > 0){
+ addMetadata(metadata, property, pdfBoxBaseline);
+ }
+ return;
+ }
+ for (String item : items){
+ if (pdfBoxBaseline != null && ! item.equals(pdfBoxBaseline)){
+ addMetadata(metadata, property, item);
+ }
+ }
+ //finally, add the baseline
+ if (pdfBoxBaseline != null && pdfBoxBaseline.length() > 0){
+ addMetadata(metadata, property, pdfBoxBaseline);
+ }
+ }
+
+ /**
+ * As of this writing, XMPSchema can contain bags or sequence lists
+ * for some attributes...despite standards documentation.
+ * JempBox expects one or the other for specific attributes.
+ * Until more flexibility is added to JempBox, Tika will have to handle both.
+ *
+ * @param schema
+ * @param name
+ * @return list of values or null
+ */
+ private List<String> getXMPBagOrSeqList(XMPSchema schema, String name){
+ List<String> ret = schema.getBagList(name);
+ if (ret == null){
+ ret = schema.getSequenceList(name);
+ }
+ return ret;
+ }
+
private void addMetadata(Metadata metadata, Property property, String value) {
if (value != null) {
metadata.add(property, value);
Modified: tika/trunk/tika-parsers/src/test/java/org/apache/tika/parser/pdf/PDFParserTest.java
URL: http://svn.apache.org/viewvc/tika/trunk/tika-parsers/src/test/java/org/apache/tika/parser/pdf/PDFParserTest.java?rev=1574964&r1=1574963&r2=1574964&view=diff
==============================================================================
--- tika/trunk/tika-parsers/src/test/java/org/apache/tika/parser/pdf/PDFParserTest.java (original)
+++ tika/trunk/tika-parsers/src/test/java/org/apache/tika/parser/pdf/PDFParserTest.java Thu Mar 6 17:08:09 2014
@@ -737,7 +737,34 @@ public class PDFParserTest extends TikaT
assertEquals("pdfaid:part", m.get("pdfaid:part"), "1");
}
+ public void testMultipleAuthors() throws Exception {
+ String fName = "testPDF_twoAuthors.pdf";
+ InputStream is = PDFParserTest.class.getResourceAsStream(
+ "/test-documents/"+fName);
+ Parser p = new AutoDetectParser();
+ Metadata m = new Metadata();
+ ParseContext c = new ParseContext();
+ ContentHandler h = new BodyContentHandler();
+ p.parse(is, h, m, c);
+ is.close();
+
+ String[] keys = new String[]{
+ "dc:creator",
+ "meta:author",
+ "creator",
+ "Author"
+ };
+ for (String k : keys){
+ String[] vals = m.getValues(k);
+ assertEquals("number of authors == 2 for key: "+ k, 2, vals.length);
+ Set<String> set = new HashSet<String>();
+ set.add(vals[0]);
+ set.add(vals[1]);
+ assertTrue("Sample Author 1", set.contains("Sample Author 1"));
+ assertTrue("Sample Author 2", set.contains("Sample Author 2"));
+ }
+ }
/**
* This is a workaround until PDFBox-1922 is fixed.
* The goal is to test for equality but skip the version issue.
Added: tika/trunk/tika-parsers/src/test/resources/test-documents/testPDF_twoAuthors.pdf
URL: http://svn.apache.org/viewvc/tika/trunk/tika-parsers/src/test/resources/test-documents/testPDF_twoAuthors.pdf?rev=1574964&view=auto
==============================================================================
Binary file - no diff available.
Propchange: tika/trunk/tika-parsers/src/test/resources/test-documents/testPDF_twoAuthors.pdf
------------------------------------------------------------------------------
svn:mime-type = application/octet-stream