TIKA-1974 -- remove deprecated metadata properties/keys for Tika 2.0

kaitlyn-liao · Jan 26, 2018 · 10a8eec · 10a8eec
1 parent 3ce43ad
commit 10a8eec
Show file tree

Hide file tree

Showing 162 changed files with 823 additions and 1,204 deletions.
diff --git a/CHANGES.txt b/CHANGES.txt
@@ -1,5 +1,7 @@
 Release 2.0.0 - ???
 
+   * Remove deprecated Metadata keys/properties (TIKA-1974).
+
    * Require Java 8 (TIKA-2553).
 
    * Add a parser for XPS (TIKA-2524).

diff --git a/tika-app/src/main/java/org/apache/tika/cli/TikaCLI.java b/tika-app/src/main/java/org/apache/tika/cli/TikaCLI.java
@@ -80,6 +80,7 @@
 import org.apache.tika.io.TikaInputStream;
 import org.apache.tika.language.detect.LanguageHandler;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.metadata.serialization.JsonMetadata;
 import org.apache.tika.metadata.serialization.JsonMetadataList;
 import org.apache.tika.mime.MediaType;
@@ -1036,7 +1037,7 @@ public boolean shouldParseEmbedded(Metadata metadata) {
         }
 
         public void parseEmbedded(InputStream inputStream, ContentHandler contentHandler, Metadata metadata, boolean outputHtml) throws SAXException, IOException {
-            String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
+            String name = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
 
             if (name == null) {
                 name = "file" + count++;
@@ -1053,7 +1054,7 @@ public void parseEmbedded(InputStream inputStream, ContentHandler contentHandler
                 }
             }
 
-            String relID = metadata.get(Metadata.EMBEDDED_RELATIONSHIP_ID);
+            String relID = metadata.get(TikaCoreProperties.EMBEDDED_RELATIONSHIP_ID);
             if (relID != null && !name.startsWith(relID)) {
                 name = relID + "_" + name;
             }

diff --git a/tika-app/src/main/java/org/apache/tika/gui/TikaGUI.java b/tika-app/src/main/java/org/apache/tika/gui/TikaGUI.java
@@ -67,6 +67,7 @@
 import org.apache.tika.extractor.DocumentSelector;
 import org.apache.tika.io.TikaInputStream;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.metadata.serialization.JsonMetadataList;
 import org.apache.tika.mime.MediaType;
 import org.apache.tika.parser.AbstractParser;
@@ -367,7 +368,7 @@ private void handleStream(InputStream input, Metadata md)
             }
         }
 
-        String name = md.get(Metadata.RESOURCE_NAME_KEY);
+        String name = md.get(TikaCoreProperties.RESOURCE_NAME_KEY);
         if (name != null && name.length() > 0) {
             setTitle("Apache Tika: " + name);
         } else {
@@ -653,7 +654,7 @@ public Set<MediaType> getSupportedTypes(ParseContext context) {
       public void parse(InputStream stream, ContentHandler handler,
             Metadata metadata, ParseContext context) throws IOException,
             SAXException, TikaException {
-         String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
+         String name = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
          if(name != null && wanted.containsKey(name)) {
             FileOutputStream out = new FileOutputStream(wanted.get(name));
             IOUtils.copy(stream, out);

diff --git a/tika-app/src/test/java/org/apache/tika/cli/TikaCLITest.java b/tika-app/src/test/java/org/apache/tika/cli/TikaCLITest.java
@@ -159,9 +159,9 @@ public void testJsonMetadataOutput() throws Exception {
         //test legacy alphabetic sort of keys
         int enc = json.indexOf("\"Content-Encoding\"");
         int fb = json.indexOf("fb:admins");
-        int title = json.indexOf("\"title\"");
+        int title = json.indexOf("\"dc:title\"");
         assertTrue(enc > -1 && fb > -1 && enc < fb);
-        assertTrue (fb > -1 && title > -1 && fb < title);
+        assertTrue (fb > -1 && title > -1 && fb > title);
         assertTrue(json.contains("\"X-TIKA:digest:MD2\":"));
     }
 
@@ -183,9 +183,9 @@ public void testJsonMetadataPrettyPrintOutput() throws Exception {
         //test legacy alphabetic sort of keys
         int enc = json.indexOf("\"Content-Encoding\"");
         int fb = json.indexOf("fb:admins");
-        int title = json.indexOf("\"title\"");
+        int title = json.indexOf("\"dc:title\"");
         assertTrue(enc > -1 && fb > -1 && enc < fb);
-        assertTrue (fb > -1 && title > -1 && fb < title);
+        assertTrue (fb > -1 && title > -1 && fb > title);
     }
 
     /**
@@ -398,12 +398,10 @@ public void testJsonRecursiveMetadataParserMetadataOnly() throws Exception {
         String[] params = new String[]{"-m", "-J", "-r", resourcePrefix+"test_recursive_embedded.docx"};
         TikaCLI.main(params);
         String content = outContent.toString(UTF_8.name());
-        assertTrue(content.contains("[\n" +
-                "  {\n" +
-                "    \"Application-Name\": \"Microsoft Office Word\",\n" +
-                "    \"Application-Version\": \"15.0000\",\n" +
-                "    \"Character Count\": \"28\",\n" +
-                "    \"Character-Count-With-Spaces\": \"31\","));
+        assertTrue(content.contains(
+                "\"extended-properties:AppVersion\": \"15.0000\","));
+        assertTrue(content.contains(
+                "\"extended-properties:Application\": \"Microsoft Office Word\","));
         assertTrue(content.contains("\"X-TIKA:embedded_resource_path\": \"/embed1.zip\""));
         assertFalse(content.contains("X-TIKA:content"));
     }

diff --git a/tika-batch/src/main/java/org/apache/tika/batch/fs/FSDocumentSelector.java b/tika-batch/src/main/java/org/apache/tika/batch/fs/FSDocumentSelector.java
@@ -22,11 +22,12 @@
 
 import org.apache.tika.extractor.DocumentSelector;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.util.PropsUtil;
 
 /**
  * Selector that chooses files based on their file name
- * and their size, as determined by Metadata.RESOURCE_NAME_KEY and Metadata.CONTENT_LENGTH.
+ * and their size, as determined by TikaCoreProperties.RESOURCE_NAME_KEY and Metadata.CONTENT_LENGTH.
  * <p/>
  * The {@link #excludeFileName} pattern is applied first (if it isn't null).
  * Then the {@link #includeFileName} pattern is applied (if it isn't null),
@@ -52,7 +53,7 @@ public FSDocumentSelector(Pattern includeFileName, Pattern excludeFileName, long
 
     @Override
     public boolean select(Metadata metadata) {
-        String fName = metadata.get(Metadata.RESOURCE_NAME_KEY);
+        String fName = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
         long sz = PropsUtil.getLong(metadata.get(Metadata.CONTENT_LENGTH), -1L);
         if (maxFileSizeBytes > -1 && sz > 0) {
             if (sz > maxFileSizeBytes) {

diff --git a/tika-batch/src/main/java/org/apache/tika/batch/fs/FSFileResource.java b/tika-batch/src/main/java/org/apache/tika/batch/fs/FSFileResource.java
@@ -28,13 +28,14 @@
 import org.apache.tika.batch.FileResource;
 import org.apache.tika.io.TikaInputStream;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 
 /**
  * FileSystem(FS)Resource wraps a file name.
  * <p/>
  * This class automatically sets the following keys in Metadata:
  * <ul>
- *     <li>Metadata.RESOURCE_NAME_KEY (file name)</li>
+ *     <li>TikaCoreProperties.RESOURCE_NAME_KEY (file name)</li>
  *     <li>Metadata.CONTENT_LENGTH</li>
  *     <li>FSProperties.FS_REL_PATH</li>
  *     <li>FileResource.FILE_EXTENSION</li>
@@ -76,7 +77,7 @@ public FSFileResource(Path inputRoot, Path fullPath) {
 
         //need to set these now so that the filter can determine
         //whether or not to crawl this file
-        metadata.set(Metadata.RESOURCE_NAME_KEY, fullPath.getFileName().toString());
+        metadata.set(TikaCoreProperties.RESOURCE_NAME_KEY, fullPath.getFileName().toString());
         long sz = -1;
         try {
             sz = Files.size(fullPath);

diff --git a/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java b/tika-batch/src/test/java/org/apache/tika/batch/RecursiveParserWrapperFSConsumerTest.java
@@ -46,7 +46,7 @@ public class RecursiveParserWrapperFSConsumerTest extends TikaTest {
     public void testEmbeddedWithNPE() throws Exception {
         final String path = "/test-documents/embedded_with_npe.xml";
         final Metadata metadata = new Metadata();
-        metadata.add(Metadata.RESOURCE_NAME_KEY, "embedded_with_npe.xml");
+        metadata.add(TikaCoreProperties.RESOURCE_NAME_KEY, "embedded_with_npe.xml");
 
         ArrayBlockingQueue<FileResource> queue = new ArrayBlockingQueue<FileResource>(2);
         queue.add(new FileResource() {
@@ -93,7 +93,7 @@ queue, new AutoDetectParserFactory(), new BasicContentHandlerFactory(BasicConten
     public void testEmbeddedThenNPE() throws Exception {
         final String path = "/test-documents/embedded_then_npe.xml";
         final Metadata metadata = new Metadata();
-        metadata.add(Metadata.RESOURCE_NAME_KEY, "embedded_then_npe.xml");
+        metadata.add(TikaCoreProperties.RESOURCE_NAME_KEY, "embedded_then_npe.xml");
 
         ArrayBlockingQueue<FileResource> queue = new ArrayBlockingQueue<FileResource>(2);
         queue.add(new FileResource() {

diff --git a/tika-bundle/src/test/java/org/apache/tika/bundle/BundleIT.java b/tika-bundle/src/test/java/org/apache/tika/bundle/BundleIT.java
@@ -46,6 +46,7 @@
 import org.apache.tika.detect.Detector;
 import org.apache.tika.fork.ForkParser;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.mime.MediaType;
 import org.apache.tika.parser.CompositeParser;
 import org.apache.tika.parser.DefaultParser;
@@ -130,10 +131,10 @@ public void testManifestNoJUnit() throws Exception {
     @Test
     public void testBundleDetection() throws Exception {
         Metadata metadataTXT = new Metadata();
-        metadataTXT.set(Metadata.RESOURCE_NAME_KEY, "test.txt");
+        metadataTXT.set(TikaCoreProperties.RESOURCE_NAME_KEY, "test.txt");
 
         Metadata metadataPDF = new Metadata();
-        metadataPDF.set(Metadata.RESOURCE_NAME_KEY, "test.pdf");
+        metadataPDF.set(TikaCoreProperties.RESOURCE_NAME_KEY, "test.pdf");
 
         // Simple type detection
         assertEquals(MediaType.TEXT_PLAIN, contentTypeDetector.detect(null, metadataTXT));

diff --git a/tika-core/pom.xml b/tika-core/pom.xml
@@ -125,6 +125,7 @@
                 <exlude>org/apache/tika/metadata/Metadata</exlude>
                 <exlude>org/apache/tika/metadata/MSOffice</exlude>
                 <exlude>org/apache/tika/parser/EmptyParser</exlude>
+                <exclude>org/apache/tika/metadata/TikaMetadataKeys</exclude>
               </excludes>
               <comparisonArtifacts>
                 <comparisonArtifact>

diff --git a/tika-core/src/main/java/org/apache/tika/Tika.java b/tika-core/src/main/java/org/apache/tika/Tika.java
@@ -32,6 +32,7 @@
 import org.apache.tika.io.TikaInputStream;
 import org.apache.tika.language.translate.Translator;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.parser.AutoDetectParser;
 import org.apache.tika.parser.ParseContext;
 import org.apache.tika.parser.Parser;
@@ -180,7 +181,7 @@ public String detect(InputStream stream, Metadata metadata)
      */
     public String detect(InputStream stream, String name) throws IOException {
         Metadata metadata = new Metadata();
-        metadata.set(Metadata.RESOURCE_NAME_KEY, name);
+        metadata.set(TikaCoreProperties.RESOURCE_NAME_KEY, name);
         return detect(stream, metadata);
     }
 

diff --git a/tika-core/src/main/java/org/apache/tika/detect/NameDetector.java b/tika-core/src/main/java/org/apache/tika/detect/NameDetector.java
@@ -23,6 +23,7 @@
 import java.util.regex.Pattern;
 
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.mime.MediaType;
 
 import static java.nio.charset.StandardCharsets.UTF_8;
@@ -95,7 +96,7 @@ public NameDetector(Map<Pattern, MediaType> patterns) {
      */
     public MediaType detect(InputStream input, Metadata metadata) {
         // Look for a resource name in the input metadata
-        String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
+        String name = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
         if (name != null) {
             // If the name is a URL, skip the trailing query
             int question = name.indexOf('?');

diff --git a/tika-core/src/main/java/org/apache/tika/extractor/ParserContainerExtractor.java b/tika-core/src/main/java/org/apache/tika/extractor/ParserContainerExtractor.java
@@ -28,6 +28,7 @@
 import org.apache.tika.io.TemporaryResources;
 import org.apache.tika.io.TikaInputStream;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.mime.MediaType;
 import org.apache.tika.parser.AbstractParser;
 import org.apache.tika.parser.AutoDetectParser;
@@ -111,7 +112,7 @@ public void parse(
                 TikaInputStream tis = TikaInputStream.get(stream, tmp);
 
                 // Figure out what we have to process
-                String filename = metadata.get(Metadata.RESOURCE_NAME_KEY);
+                String filename = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
                 MediaType type = detector.detect(tis, metadata);
 
                 if (extractor == null) {

diff --git a/tika-core/src/main/java/org/apache/tika/extractor/ParsingEmbeddedDocumentExtractor.java b/tika-core/src/main/java/org/apache/tika/extractor/ParsingEmbeddedDocumentExtractor.java
@@ -27,6 +27,7 @@
 import org.apache.tika.io.TemporaryResources;
 import org.apache.tika.io.TikaInputStream;
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 import org.apache.tika.parser.DelegatingParser;
 import org.apache.tika.parser.ParseContext;
 import org.apache.tika.parser.Parser;
@@ -64,7 +65,7 @@ public boolean shouldParseEmbedded(Metadata metadata) {
 
         FilenameFilter filter = context.get(FilenameFilter.class);
         if (filter != null) {
-            String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
+            String name = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
             if (name != null) {
                 return filter.accept(ABSTRACT_PATH, name);
             }
@@ -82,7 +83,7 @@ public void parseEmbedded(
            handler.startElement(XHTML, "div", "div", attributes);
         }
 
-        String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
+        String name = metadata.get(TikaCoreProperties.RESOURCE_NAME_KEY);
         if (name != null && name.length() > 0 && outputHtml) {
             handler.startElement(XHTML, "h1", "h1", new AttributesImpl());
             char[] chars = name.toCharArray();

diff --git a/tika-core/src/main/java/org/apache/tika/io/FilenameUtils.java b/tika-core/src/main/java/org/apache/tika/io/FilenameUtils.java
@@ -83,7 +83,7 @@ public static String normalize(final String name) {
      * <p>
      * The goal of this is to get a filename from a path.
      * The package parsers and some other embedded doc
-     * extractors could put anything into Metadata.RESOURCE_NAME_KEY.
+     * extractors could put anything into TikaCoreProperties.RESOURCE_NAME_KEY.
      * <p>
      * If a careless client used that filename as if it were a
      * filename and not a path when writing embedded files,

diff --git a/tika-core/src/main/java/org/apache/tika/io/TikaInputStream.java b/tika-core/src/main/java/org/apache/tika/io/TikaInputStream.java
@@ -38,6 +38,7 @@
 import java.sql.SQLException;
 
 import org.apache.tika.metadata.Metadata;
+import org.apache.tika.metadata.TikaCoreProperties;
 
 /**
  * Input stream with extended capabilities. The purpose of this class is
@@ -222,7 +223,7 @@ public static TikaInputStream get(Path path) throws IOException {
      */
     public static TikaInputStream get(Path path, Metadata metadata)
             throws IOException {
-        metadata.set(Metadata.RESOURCE_NAME_KEY, path.getFileName().toString());
+        metadata.set(TikaCoreProperties.RESOURCE_NAME_KEY, path.getFileName().toString());
         metadata.set(Metadata.CONTENT_LENGTH, Long.toString(Files.size(path)));
         return new TikaInputStream(path);
     }
@@ -262,7 +263,7 @@ public static TikaInputStream get(File file) throws FileNotFoundException {
     @Deprecated
     public static TikaInputStream get(File file, Metadata metadata)
             throws FileNotFoundException {
-        metadata.set(Metadata.RESOURCE_NAME_KEY, file.getName());
+        metadata.set(TikaCoreProperties.RESOURCE_NAME_KEY, file.getName());
         metadata.set(Metadata.CONTENT_LENGTH, Long.toString(file.length()));
         return new TikaInputStream(file);
     }
@@ -410,7 +411,7 @@ public static TikaInputStream get(URL url, Metadata metadata)
         String path = url.getPath();
         int slash = path.lastIndexOf('/');
         if (slash + 1 < path.length()) { // works even with -1!
-            metadata.set(Metadata.RESOURCE_NAME_KEY, path.substring(slash + 1));
+            metadata.set(TikaCoreProperties.RESOURCE_NAME_KEY, path.substring(slash + 1));
         }
 
         String type = connection.getContentType();

diff --git a/tika-core/src/main/java/org/apache/tika/metadata/AccessPermissions.java b/tika-core/src/main/java/org/apache/tika/metadata/AccessPermissions.java
@@ -24,7 +24,7 @@
  */
 public interface AccessPermissions {
 
-    final static String PREFIX = "access_permission"+Metadata.NAMESPACE_PREFIX_DELIMITER;
+    final static String PREFIX = "access_permission"+TikaCoreProperties.NAMESPACE_PREFIX_DELIMITER;
 
     /**
      * Can any modifications be made to the document

diff --git a/tika-core/src/main/java/org/apache/tika/metadata/Database.java b/tika-core/src/main/java/org/apache/tika/metadata/Database.java
@@ -17,7 +17,7 @@
  * limitations under the License.
  */
 public interface Database {
-    final static String PREFIX = "database"+Metadata.NAMESPACE_PREFIX_DELIMITER;
+    final static String PREFIX = "database"+TikaCoreProperties.NAMESPACE_PREFIX_DELIMITER;
 
     Property TABLE_NAME = Property.externalTextBag(PREFIX+"table_name");
     Property COLUMN_COUNT = Property.externalText(PREFIX+"column_count");