Refactor: Renamed to SemaphoreProxy

Avoids confusion over use of "MultiArticle". Main use case for this is to hide the actual call to Semaphore so it can be mocked easily.

Author: rjrudin

marklogic-spark-api/src/main/java/com/marklogic/spark/core/classifier/MultiArticleClassifier.java renamed to marklogic-spark-api/src/main/java/com/marklogic/spark/core/classifier/SemaphoreProxy.java

@@ -9,11 +9,10 @@
 
 /**
  * Provides an abstraction over the call to Semaphore that sends a document containing multiple articles and
- * then receives a document containing a classification for each article.
- *
- * This will get a new name soon now that its scope is more than just classifying a set of articles.
+ * then receives a document containing a classification for each article. Main use case is to enable easy mocking of
+ * the calls to Semaphore for testing purposes.
  */
-public interface MultiArticleClassifier extends Closeable {
+public interface SemaphoreProxy extends Closeable {
 
     byte[] classifyDocument(byte[] content, String uri);
 

marklogic-spark-api/src/main/java/com/marklogic/spark/core/classifier/SemaphoreMultiArticleClassifier.java renamed to marklogic-spark-api/src/main/java/com/marklogic/spark/core/classifier/SemaphoreProxyImpl.java

@@ -15,12 +15,12 @@
  * Not threadsafe, as the config of the underlying client is modified. This is fine in the context of the connector,
  * where it is only used by a single Spark data writer.
  */
-class SemaphoreMultiArticleClassifier implements MultiArticleClassifier {
+class SemaphoreProxyImpl implements SemaphoreProxy {
 
     private final ClassificationClient classificationClient;
     private long totalDurationOfCalls;
 
-    public SemaphoreMultiArticleClassifier(ClassificationConfiguration config) {
+    public SemaphoreProxyImpl(ClassificationConfiguration config) {
         this.classificationClient = new ClassificationClient();
         config.setMultiArticle(true);
         classificationClient.setClassificationConfiguration(config);

marklogic-spark-api/src/main/java/com/marklogic/spark/core/classifier/SemaphoreTextClassifier.java

@@ -34,15 +34,15 @@ class SemaphoreTextClassifier implements TextClassifier {
 
     static final Logger SEMAPHORE_LOGGER = LoggerFactory.getLogger("com.marklogic.semaphore.classifier");
 
-    private final MultiArticleClassifier multiArticleClassifier;
+    private final SemaphoreProxy semaphoreProxy;
     private final DOMHelper domHelper;
     private final Transformer transformer;
     private final String encoding;
     private final XPathExpression articleExpression;
     private final int batchSize;
 
-    SemaphoreTextClassifier(MultiArticleClassifier multiArticleClassifier, String encoding, int batchSize) {
-        this.multiArticleClassifier = multiArticleClassifier;
+    SemaphoreTextClassifier(SemaphoreProxy semaphoreProxy, String encoding, int batchSize) {
+        this.semaphoreProxy = semaphoreProxy;
         this.domHelper = new DOMHelper(null);
         this.transformer = newTransformer();
         this.encoding = encoding;
@@ -53,7 +53,7 @@ class SemaphoreTextClassifier implements TextClassifier {
     @Override
     public void classifyDocument(DocumentInputs inputs) {
         byte[] content = inputs.getContentAsBytes();
-        byte[] classification = multiArticleClassifier.classifyDocument(content, inputs.getInitialUri());
+        byte[] classification = semaphoreProxy.classifyDocument(content, inputs.getInitialUri());
         inputs.setDocumentClassification(classification);
     }
 
@@ -82,7 +82,7 @@ private void classifyBatch(List<ClassifiableContent> contentBatch) {
 
         Document structuredDocument;
         try {
-            structuredDocument = multiArticleClassifier.classifyArticles(documentBytes);
+            structuredDocument = semaphoreProxy.classifyArticles(documentBytes);
         } catch (Exception e) {
             throw new ConnectorException(String.format("Unable to classify content, cause: %s", e.getMessage()), e);
         }
@@ -96,7 +96,7 @@ private void classifyBatch(List<ClassifiableContent> contentBatch) {
 
     @Override
     public void close() throws IOException {
-        IOUtils.closeQuietly(multiArticleClassifier);
+        IOUtils.closeQuietly(semaphoreProxy);
     }
 
     private Document buildMultiArticleRequest(List<ClassifiableContent> classifiableContents) {

marklogic-spark-api/src/main/java/com/marklogic/spark/core/classifier/TextClassifierFactory.java

@@ -24,17 +24,17 @@ public abstract class TextClassifierFactory {
     private static final String MOCK_CLASSIFIER_OPTION = "spark.marklogic.testing.mockClassifierResponse";
 
     public static TextClassifier newTextClassifier(Context context) {
-        MultiArticleClassifier multiArticleClassifier = null;
+        SemaphoreProxy semaphoreProxy = null;
         final String host = context.getStringOption(Options.WRITE_CLASSIFIER_HOST);
 
         if (context.hasOption(MOCK_CLASSIFIER_OPTION)) {
             String mockResponse = context.getStringOption(MOCK_CLASSIFIER_OPTION);
             assert mockResponse != null;
-            multiArticleClassifier = new MockTextClassifier(mockResponse);
+            semaphoreProxy = new MockSemaphoreProxy(mockResponse);
         } else if (host != null && !host.trim().isEmpty()) {
             try {
                 ClassificationConfiguration config = buildClassificationConfiguration(context);
-                multiArticleClassifier = new SemaphoreMultiArticleClassifier(config);
+                semaphoreProxy = new SemaphoreProxyImpl(config);
             } catch (ConnectorException ex) {
                 throw ex;
             } catch (Exception e) {
@@ -43,11 +43,11 @@ public static TextClassifier newTextClassifier(Context context) {
             }
         }
 
-        if (multiArticleClassifier != null) {
+        if (semaphoreProxy != null) {
             // We may need a dedicated encoding for this
             String encoding = context.getStringOption(Options.READ_FILES_ENCODING, "UTF-8");
             int batchSize = context.getIntOption(Options.WRITE_CLASSIFIER_BATCH_SIZE, 20, 1);
-            return new SemaphoreTextClassifier(multiArticleClassifier, encoding, batchSize);
+            return new SemaphoreTextClassifier(semaphoreProxy, encoding, batchSize);
         }
         return null;
     }
@@ -133,15 +133,15 @@ private TextClassifierFactory() {
 
     // Sonar doesn't like static assignments in this class, but this class is only used as a mock for testing.
     @SuppressWarnings("java:S2696")
-    public static class MockTextClassifier implements MultiArticleClassifier {
+    public static class MockSemaphoreProxy implements SemaphoreProxy {
 
         private final Document mockResponse;
         private static int timesInvoked;
         private static boolean wasClosed;
 
         // Sonar doesn't like this static assignment, but it's fine in a class that's only used as a mock.
         @SuppressWarnings("java:S3010")
-        private MockTextClassifier(String mockResponse) {
+        private MockSemaphoreProxy(String mockResponse) {
             this.mockResponse = new DOMHelper(null).parseXmlString(mockResponse, null);
             timesInvoked = 0;
         }

tests/src/test/java/com/marklogic/spark/writer/classifier/AddClassificationToJsonTest.java

@@ -26,7 +26,7 @@ void chunkAndAddClassificationToJsonInOriginalJsonDoc() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         JsonNode doc = readJsonDocument("/split-test.json");
         assertTrue(doc.get("classification").has("STRUCTUREDDOCUMENT"));
@@ -47,7 +47,7 @@ void sidecarChunksAddClassificationToJson() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         JsonNode doc = readJsonDocument("/split-test.json");
         assertTrue(doc.get("classification").has("STRUCTUREDDOCUMENT"));
@@ -82,7 +82,7 @@ void classifyJsonContentsWithoutChunking() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         JsonNode doc = readJsonDocument("/split-test.json");
         assertTrue(doc.get("classification").has("STRUCTUREDDOCUMENT"));
@@ -98,7 +98,7 @@ void chunkAndAddClassificationOnlyToChunksInOriginalDoc() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         JsonNode doc = readJsonDocument("/split-test.json");
         assertEquals(4, doc.get("chunks").size(), "Expecting 4 chunks based on max chunk size of 500.");

tests/src/test/java/com/marklogic/spark/writer/classifier/AddClassificationToXmlTest.java

@@ -25,7 +25,7 @@ void chunkAndAddClassificationToXmlInOriginalDoc() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         XmlNode doc = readXmlDocument("/split-test.xml");
         doc.assertElementExists("Expecting each chunk to have a 'model:classification' child element",
@@ -44,7 +44,7 @@ void sidecarChunksAddClassificationToXml() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         XmlNode doc = readXmlDocument("/split-test.xml");
         doc.assertElementExists("Expecting the root of the document to have a 'model:classification' child element",
@@ -75,7 +75,7 @@ void classifyXmlContentsWithoutChunking() {
             .mode(SaveMode.Append)
             .save();
 
-        assertTrue(TextClassifierFactory.MockTextClassifier.isClosed());
+        assertTrue(TextClassifierFactory.MockSemaphoreProxy.isClosed());
 
         XmlNode doc = readXmlDocument("/split-test.xml");
         doc.assertElementExists("Expecting the root of the document to have a 'model:classification' child element",

tests/src/test/java/com/marklogic/spark/writer/classifier/ClassifyExtractedTextTest.java

@@ -73,7 +73,7 @@ void withBatchSize() {
             .mode(SaveMode.Append)
             .save();
 
-        assertEquals(3, TextClassifierFactory.MockTextClassifier.getTimesInvoked(), "The mock classifier should " +
+        assertEquals(3, TextClassifierFactory.MockSemaphoreProxy.getTimesInvoked(), "The mock classifier should " +
             "have been invoked 3 times - with 10, 10, and then 8 articles.");
     }