Adapt Java bindings to new API

native_client/java/jni/deepspeech.i

@@ -18,18 +18,32 @@
 %typemap(newfree) char* "DS_FreeString($1);";
 
 %include "carrays.i"
-%array_functions(struct MetadataItem, metadataItem_array);
+%array_functions(struct TokenMetadata, TokenMetadata_array);
+%array_functions(struct CandidateTranscript, CandidateTranscript_array);
+
+%extend struct CandidateTranscript {
+  /**
+   * Retrieve one TokenMetadata element
+   * 
+   * @param i Array index of the TokenMetadata to get
+   *
+   * @return The TokenMetadata requested or null
+   */
+  TokenMetadata getToken(int i) {
+    return TokenMetadata_array_getitem(self->tokens, i);
+  }
+}
 
 %extend struct Metadata {
   /**
-   * Retrieve one MetadataItem element
+   * Retrieve one CandidateTranscript element
    * 
-   * @param i Array index of the MetadataItem to get
+   * @param i Array index of the CandidateTranscript to get
    *
-   * @return The MetadataItem requested or null
+   * @return The CandidateTranscript requested or null
    */
-  MetadataItem getItem(int i) {
-    return metadataItem_array_getitem(self->items, i);
+  CandidateTranscript getTranscript(int i) {
+    return CandidateTranscript_array_getitem(self->transcripts, i);
   }
 
   ~Metadata() {
@@ -37,10 +51,12 @@
   }
 }
 
-%nodefaultdtor Metadata;
 %nodefaultctor Metadata;
-%nodefaultctor MetadataItem;
-%nodefaultdtor MetadataItem;
+%nodefaultdtor Metadata;
+%nodefaultctor CandidateTranscript;
+%nodefaultdtor CandidateTranscript;
+%nodefaultctor TokenMetadata;
+%nodefaultdtor TokenMetadata;
 
 %newobject DS_SpeechToText;
 %newobject DS_IntermediateDecode;

native_client/java/libdeepspeech/src/androidTest/java/org/mozilla/deepspeech/libdeepspeech/test/BasicTest.java

@@ -12,7 +12,7 @@ import org.junit.runners.MethodSorters;
 import static org.junit.Assert.*;
 
 import org.mozilla.deepspeech.libdeepspeech.DeepSpeechModel;
-import org.mozilla.deepspeech.libdeepspeech.Metadata;
+import org.mozilla.deepspeech.libdeepspeech.CandidateTranscript;
 
 import java.io.RandomAccessFile;
 import java.io.FileNotFoundException;
@@ -61,10 +61,10 @@ public class BasicTest {
         m.freeModel();
     }
 
-    private String metadataToString(Metadata m) {
+    private String candidateTranscriptToString(CandidateTranscript t) {
         String retval = "";
-        for (int i = 0; i < m.getNum_items(); ++i) {
-            retval += m.getItem(i).getCharacter();
+        for (int i = 0; i < t.getNum_tokens(); ++i) {
+            retval += t.getToken(i).getText();
         }
         return retval;
     }
@@ -97,7 +97,7 @@ public class BasicTest {
             ByteBuffer.wrap(bytes).order(ByteOrder.LITTLE_ENDIAN).asShortBuffer().get(shorts);
 
             if (extendedMetadata) {
-                return metadataToString(m.sttWithMetadata(shorts, shorts.length));
+                return candidateTranscriptToString(m.sttWithMetadata(shorts, shorts.length, 1).getTranscript(0));
             } else {
                 return m.stt(shorts, shorts.length);
             }

native_client/java/libdeepspeech/src/main/java/org/mozilla/deepspeech/libdeepspeech/DeepSpeechModel.java

@@ -117,11 +117,12 @@ public class DeepSpeechModel {
     * @param buffer A 16-bit, mono raw audio signal at the appropriate
     *                sample rate (matching what the model was trained on).
     * @param buffer_size The number of samples in the audio signal.
+    * @param num_results Number of candidate transcripts to return.
     *
     * @return Outputs a Metadata object of individual letters along with their timing information.
     */
-    public Metadata sttWithMetadata(short[] buffer, int buffer_size) {
-        return impl.SpeechToTextWithMetadata(this._msp, buffer, buffer_size);
+    public Metadata sttWithMetadata(short[] buffer, int buffer_size, int num_results) {
+        return impl.SpeechToTextWithMetadata(this._msp, buffer, buffer_size, num_results);
     }
 
    /**
@@ -160,6 +161,18 @@ public class DeepSpeechModel {
         return impl.IntermediateDecode(ctx.get());
     }
 
+   /**
+    * @brief Compute the intermediate decoding of an ongoing streaming inference.
+    *
+    * @param ctx A streaming state pointer returned by createStream().
+    * @param num_results Number of candidate transcripts to return.
+    *
+    * @return The STT intermediate result.
+    */
+    public Metadata intermediateDecodeWithMetadata(DeepSpeechStreamingState ctx, int num_results) {
+        return impl.IntermediateDecodeWithMetadata(ctx.get(), num_results);
+    }
+
    /**
     * @brief Signal the end of an audio signal to an ongoing streaming
     *        inference, returns the STT result over the whole audio signal.
@@ -179,12 +192,13 @@ public class DeepSpeechModel {
     *        inference, returns per-letter metadata.
     *
     * @param ctx A streaming state pointer returned by createStream().
+    * @param num_results Number of candidate transcripts to return.
     *
     * @return Outputs a Metadata object of individual letters along with their timing information.
     *
     * @note This method will free the state pointer (@p ctx).
     */
-    public Metadata finishStreamWithMetadata(DeepSpeechStreamingState ctx) {
-        return impl.FinishStreamWithMetadata(ctx.get());
+    public Metadata finishStreamWithMetadata(DeepSpeechStreamingState ctx, int num_results) {
+        return impl.FinishStreamWithMetadata(ctx.get(), num_results);
     }
 }