(tika) branch main updated: TIKA-4444: improve test coverage

[tilman]

This is an automated email from the ASF dual-hosted git repository.

tilman pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tika.git


The following commit(s) were added to refs/heads/main by this push:
     new b09992892 TIKA-4444: improve test coverage
b09992892 is described below

commit b09992892ad1340f8842943910135ef268b45d09
Author: Tilman Hausherr <til...@apache.org>
AuthorDate: Thu Jun 26 12:42:33 2025 +0200

    TIKA-4444: improve test coverage
---
 .../java/org/apache/tika/parser/pdf/CustomTikaXMPTest.java | 14 +++++++++++++-
 .../src/test/resources/test-documents/xmp/testPDFUA.xmp    | 10 ++++++++++
 2 files changed, 23 insertions(+), 1 deletion(-)

diff --git 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/java/org/apache/tika/parser/pdf/CustomTikaXMPTest.java
 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/java/org/apache/tika/parser/pdf/CustomTikaXMPTest.java
index dc6e3b3b4..56611c707 100644
--- 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/java/org/apache/tika/parser/pdf/CustomTikaXMPTest.java
+++ 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/java/org/apache/tika/parser/pdf/CustomTikaXMPTest.java
@@ -31,6 +31,7 @@ import org.apache.tika.exception.TikaException;
 import org.apache.tika.metadata.Metadata;
 import org.apache.tika.metadata.PDF;
 import org.apache.tika.metadata.TikaCoreProperties;
+import org.apache.tika.metadata.XMP;
 import org.apache.tika.parser.ParseContext;
 import org.apache.tika.utils.XMLReaderUtils;
 
@@ -62,6 +63,8 @@ public class CustomTikaXMPTest extends TikaTest {
         String[] subjects = metadata.getValues(TikaCoreProperties.SUBJECT);
         assertEquals("keywords", subjects[0]);
         assertEquals("subject", subjects[1]);
+        assertEquals("1234567890", metadata.get(XMP.IDENTIFIER));
+        assertEquals("Advisory", metadata.get(XMP.ADVISORY));
     }
 
     @Test
@@ -73,7 +76,7 @@ public class CustomTikaXMPTest extends TikaTest {
     }
 
     /**
-     * TIKA-4442: Test unusual dublin core properties.
+     * Test dublin core properties.
      * 
      * @throws Exception 
      */
@@ -86,12 +89,21 @@ public class CustomTikaXMPTest extends TikaTest {
         assertEquals("International Union of Thinkology", 
metadata.get(TikaCoreProperties.PUBLISHER));
         assertEquals("Relation", metadata.get(TikaCoreProperties.RELATION));
         assertEquals("Journal of Thinkology", 
metadata.get(TikaCoreProperties.SOURCE));
+        assertEquals("Thinking: is it needed?", 
metadata.get(TikaCoreProperties.DESCRIPTION));
         String[] subjects = metadata.getValues(TikaCoreProperties.SUBJECT);
+        assertEquals(5, subjects.length);
         assertEquals("THOUGHTS", subjects[0]);
         assertEquals("HAPPINESS", subjects[1]);
         assertEquals("FEAR", subjects[2]);
         assertEquals("ANGER", subjects[3]);
         assertEquals("DESPAIR", subjects[4]);
+        String[] creators = metadata.getValues(TikaCoreProperties.CREATOR);
+        assertEquals(5, creators.length);
+        assertEquals("Dorothy", creators[0]);
+        assertEquals("Toto", creators[1]);
+        assertEquals("Scarecrow", creators[2]);
+        assertEquals("Tin Man", creators[3]);
+        assertEquals("Cowardly Lion", creators[4]);
     }
 
     private Metadata extract(String xmpFileName) throws IOException, 
TikaException, SAXException {
diff --git 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/resources/test-documents/xmp/testPDFUA.xmp
 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/resources/test-documents/xmp/testPDFUA.xmp
index 20702fe38..c2275b315 100644
--- 
a/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/resources/test-documents/xmp/testPDFUA.xmp
+++ 
b/tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-pdf-module/src/test/resources/test-documents/xmp/testPDFUA.xmp
@@ -35,6 +35,16 @@
       <xmp:ModifyDate>2016-09-14T15:37:53+02:00</xmp:ModifyDate>
       <xmp:MetadataDate>2016-09-14T15:37:53+02:00</xmp:MetadataDate>
       <xmp:CreatorTool>Adobe InDesign CS6 (Macintosh)</xmp:CreatorTool>
+      <xmp:Identifier>
+        <rdf:Bag>
+          <rdf:li>1234567890</rdf:li>
+        </rdf:Bag>
+      </xmp:Identifier>
+      <xmp:Advisory>
+        <rdf:Bag>
+          <rdf:li>Advisory</rdf:li>
+        </rdf:Bag>
+      </xmp:Advisory>
       <dc:format>application/pdf</dc:format>
       <dc:title>
         <rdf:Alt>