HSEARCH-2500 Test list of PDF as Set<String>

Author: DavideD

engine/src/test/resources/org/hibernate/search/test/bridge/builtin/test-document-2.pdf

@@ -8,8 +8,11 @@
 
 import java.net.URI;
 import java.sql.Blob;
+import java.util.HashSet;
+import java.util.Set;
 
 import javax.persistence.Basic;
+import javax.persistence.ElementCollection;
 import javax.persistence.Entity;
 import javax.persistence.FetchType;
 import javax.persistence.GeneratedValue;
@@ -18,6 +21,7 @@
 
 import org.hibernate.search.annotations.Field;
 import org.hibernate.search.annotations.Indexed;
+import org.hibernate.search.annotations.IndexedEmbedded;
 import org.hibernate.search.annotations.TikaBridge;
 
 /**
@@ -32,9 +36,19 @@ public class Book {
 	private byte[] contentAsBytes;
 	private URI contentAsURI;
 
+	private Set<String> contentAsListOfString;
+
 	public Book() {
 	}
 
+	public Book(String... contents) {
+		Set<String> temp = new HashSet<>();
+		for ( String string : contents ) {
+			temp.add( string );
+		}
+		this.contentAsListOfString = temp;
+	}
+
 	public Book(Blob content) {
 		this.contentAsBlob = content;
 	}
@@ -81,7 +95,6 @@ public void setContentAsBytes(byte[] contentAsBytes) {
 		this.contentAsBytes = contentAsBytes;
 	}
 
-	@Lob
 	@Basic(fetch = FetchType.LAZY)
 	@Field(indexNullAs = "<NULL>")
 	@TikaBridge
@@ -92,4 +105,16 @@ public URI getContentAsURI() {
 	public void setContentAsURI(URI contentAsURI) {
 		this.contentAsURI = contentAsURI;
 	}
+
+	@IndexedEmbedded
+	@Field
+	@TikaBridge
+	@ElementCollection
+	public Set<String> getContentAsListOfString() {
+		return contentAsListOfString;
+	}
+
+	public void setContentAsListOfString(Set<String> contentAsListOfString) {
+		this.contentAsListOfString = contentAsListOfString;
+	}
 }

orm/src/test/java/org/hibernate/search/test/bridge/tika/Book.java

@@ -41,15 +41,50 @@
  * @author Hardy Ferentschik
  */
 public class TikaBridgeInputTypeTest extends SearchTestBase {
-	private static final String TEST_DOCUMENT_PDF = "/org/hibernate/search/test/bridge/tika/test-document-1.pdf";
+
+	private static final String TEST_DOCUMENT_PDF_1 = "/org/hibernate/search/test/bridge/tika/test-document-1.pdf";
+	private static final String TEST_DOCUMENT_PDF_2 = "/org/hibernate/search/test/bridge/tika/test-document-2.pdf";
+
+	@Rule
+	public ClasspathResourceAsFile testDocumentPdf1 = new ClasspathResourceAsFile( getClass(), TEST_DOCUMENT_PDF_1 );
 
 	@Rule
-	public ClasspathResourceAsFile testDocumentPdf = new ClasspathResourceAsFile( getClass(), TEST_DOCUMENT_PDF );
+	public ClasspathResourceAsFile testDocumentPdf2 = new ClasspathResourceAsFile( getClass(), TEST_DOCUMENT_PDF_2 );
+
+	@Test
+	public void testDefaultTikaBridgeWithListOfString() throws Exception {
+		try ( Session session = openSession() ) {
+			String content1 = testDocumentPdf1.get().getAbsolutePath();
+			String content2 = testDocumentPdf2.get().getAbsolutePath();
+
+			persistBook( session, new Book( content1, content2 ) );
+
+			indexBook( session );
+
+			List<Book> resultWithLucene = search( session, "contentAsListOfString", "Lucene" );
+			assertEquals( "there should be a match", 1, resultWithLucene.size() );
+
+			List<Book> resultWithTika = search( session, "contentAsListOfString", "Tika" );
+			assertEquals( "there should be a match", 1, resultWithTika.size() );
+		}
+	}
+
+	private List<Book> search(Session session, String field, String keyword) throws ParseException {
+		FullTextSession fullTextSession = Search.getFullTextSession( session );
+		Transaction transaction = fullTextSession.beginTransaction();
+		QueryParser parser = new QueryParser( field, TestConstants.standardAnalyzer );
+		Query query = parser.parse( keyword );
+		@SuppressWarnings("unchecked")
+		List<Book> result = fullTextSession.createFullTextQuery( query ).list();
+		transaction.commit();
+		fullTextSession.clear();
+		return result;
+	}
 
 	@Test
 	public void testDefaultTikaBridgeWithBlob() throws Exception {
 		try ( Session session = openSession() ) {
-			Blob content = dataAsBlob( testDocumentPdf.get(), session );
+			Blob content = dataAsBlob( testDocumentPdf1.get(), session );
 
 			persistBook( session, new Book( content ) );
 			persistBook( session, new Book() );
@@ -64,7 +99,7 @@ public void testDefaultTikaBridgeWithBlob() throws Exception {
 	@Test
 	public void testDefaultTikaBridgeWithByteArray() throws Exception {
 		try ( Session session = openSession() ) {
-			byte[] content = dataAsBytes( testDocumentPdf.get() );
+			byte[] content = dataAsBytes( testDocumentPdf1.get() );
 
 			persistBook( session, new Book( content ) );
 			persistBook( session, new Book() );
@@ -77,7 +112,7 @@ public void testDefaultTikaBridgeWithByteArray() throws Exception {
 	@Test
 	public void testDefaultTikaBridgeWithURI() throws Exception {
 		try ( Session session = openSession() ) {
-			URI content = testDocumentPdf.get().toURI();
+			URI content = testDocumentPdf1.get().toURI();
 
 			persistBook( session, new Book( content ) );
 			persistBook( session, new Book() );