Home | History | Annotate | Download | only in speech
      1 /*
      2  * Copyright (C) 2010 The Android Open Source Project
      3  *
      4  * Licensed under the Apache License, Version 2.0 (the "License"); you may not
      5  * use this file except in compliance with the License. You may obtain a copy of
      6  * the License at
      7  *
      8  * http://www.apache.org/licenses/LICENSE-2.0
      9  *
     10  * Unless required by applicable law or agreed to in writing, software
     11  * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
     12  * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
     13  * License for the specific language governing permissions and limitations under
     14  * the License.
     15  */
     16 
     17 package android.speech;
     18 
     19 import android.annotation.SdkConstant;
     20 import android.annotation.SdkConstant.SdkConstantType;
     21 import android.app.Service;
     22 import android.content.Intent;
     23 import android.content.pm.PackageManager;
     24 import android.os.Bundle;
     25 import android.os.Handler;
     26 import android.os.IBinder;
     27 import android.os.Message;
     28 import android.os.RemoteException;
     29 import android.util.Log;
     30 
     31 /**
     32  * This class provides a base class for recognition service implementations. This class should be
     33  * extended only in case you wish to implement a new speech recognizer. Please note that the
     34  * implementation of this service is stateless.
     35  */
     36 public abstract class RecognitionService extends Service {
     37     /**
     38      * The {@link Intent} that must be declared as handled by the service.
     39      */
     40     @SdkConstant(SdkConstantType.SERVICE_ACTION)
     41     public static final String SERVICE_INTERFACE = "android.speech.RecognitionService";
     42 
     43     /**
     44      * Name under which a RecognitionService component publishes information about itself.
     45      * This meta-data should reference an XML resource containing a
     46      * <code>&lt;{@link android.R.styleable#RecognitionService recognition-service}&gt;</code> tag.
     47      */
     48     public static final String SERVICE_META_DATA = "android.speech";
     49 
     50     /** Log messages identifier */
     51     private static final String TAG = "RecognitionService";
     52 
     53     /** Debugging flag */
     54     private static final boolean DBG = false;
     55 
     56     /** Binder of the recognition service */
     57     private RecognitionServiceBinder mBinder = new RecognitionServiceBinder(this);
     58 
     59     /**
     60      * The current callback of an application that invoked the
     61      * {@link RecognitionService#onStartListening(Intent, Callback)} method
     62      */
     63     private Callback mCurrentCallback = null;
     64 
     65     private static final int MSG_START_LISTENING = 1;
     66 
     67     private static final int MSG_STOP_LISTENING = 2;
     68 
     69     private static final int MSG_CANCEL = 3;
     70 
     71     private static final int MSG_RESET = 4;
     72 
     73     private final Handler mHandler = new Handler() {
     74         @Override
     75         public void handleMessage(Message msg) {
     76             switch (msg.what) {
     77                 case MSG_START_LISTENING:
     78                     StartListeningArgs args = (StartListeningArgs) msg.obj;
     79                     dispatchStartListening(args.mIntent, args.mListener);
     80                     break;
     81                 case MSG_STOP_LISTENING:
     82                     dispatchStopListening((IRecognitionListener) msg.obj);
     83                     break;
     84                 case MSG_CANCEL:
     85                     dispatchCancel((IRecognitionListener) msg.obj);
     86                     break;
     87                 case MSG_RESET:
     88                     dispatchClearCallback();
     89                     break;
     90             }
     91         }
     92     };
     93 
     94     private void dispatchStartListening(Intent intent, IRecognitionListener listener) {
     95         if (mCurrentCallback == null) {
     96             if (DBG) Log.d(TAG, "created new mCurrentCallback, listener = " + listener.asBinder());
     97             mCurrentCallback = new Callback(listener);
     98             RecognitionService.this.onStartListening(intent, mCurrentCallback);
     99         } else {
    100             try {
    101                 listener.onError(SpeechRecognizer.ERROR_RECOGNIZER_BUSY);
    102             } catch (RemoteException e) {
    103                 Log.d(TAG, "onError call from startListening failed");
    104             }
    105             Log.i(TAG, "concurrent startListening received - ignoring this call");
    106         }
    107     }
    108 
    109     private void dispatchStopListening(IRecognitionListener listener) {
    110         try {
    111             if (mCurrentCallback == null) {
    112                 listener.onError(SpeechRecognizer.ERROR_CLIENT);
    113                 Log.w(TAG, "stopListening called with no preceding startListening - ignoring");
    114             } else if (mCurrentCallback.mListener.asBinder() != listener.asBinder()) {
    115                 listener.onError(SpeechRecognizer.ERROR_RECOGNIZER_BUSY);
    116                 Log.w(TAG, "stopListening called by other caller than startListening - ignoring");
    117             } else { // the correct state
    118                 RecognitionService.this.onStopListening(mCurrentCallback);
    119             }
    120         } catch (RemoteException e) { // occurs if onError fails
    121             Log.d(TAG, "onError call from stopListening failed");
    122         }
    123     }
    124 
    125     private void dispatchCancel(IRecognitionListener listener) {
    126         if (mCurrentCallback == null) {
    127             if (DBG) Log.d(TAG, "cancel called with no preceding startListening - ignoring");
    128         } else if (mCurrentCallback.mListener.asBinder() != listener.asBinder()) {
    129             Log.w(TAG, "cancel called by client who did not call startListening - ignoring");
    130         } else { // the correct state
    131             RecognitionService.this.onCancel(mCurrentCallback);
    132             mCurrentCallback = null;
    133             if (DBG) Log.d(TAG, "canceling - setting mCurrentCallback to null");
    134         }
    135     }
    136 
    137     private void dispatchClearCallback() {
    138         mCurrentCallback = null;
    139     }
    140 
    141     private class StartListeningArgs {
    142         public final Intent mIntent;
    143 
    144         public final IRecognitionListener mListener;
    145 
    146         public StartListeningArgs(Intent intent, IRecognitionListener listener) {
    147             this.mIntent = intent;
    148             this.mListener = listener;
    149         }
    150     }
    151 
    152     /**
    153      * Checks whether the caller has sufficient permissions
    154      *
    155      * @param listener to send the error message to in case of error
    156      * @return {@code true} if the caller has enough permissions, {@code false} otherwise
    157      */
    158     private boolean checkPermissions(IRecognitionListener listener) {
    159         if (DBG) Log.d(TAG, "checkPermissions");
    160         if (RecognitionService.this.checkCallingOrSelfPermission(android.Manifest.permission.
    161                 RECORD_AUDIO) == PackageManager.PERMISSION_GRANTED) {
    162             return true;
    163         }
    164         try {
    165             Log.e(TAG, "call for recognition service without RECORD_AUDIO permissions");
    166             listener.onError(SpeechRecognizer.ERROR_INSUFFICIENT_PERMISSIONS);
    167         } catch (RemoteException re) {
    168             Log.e(TAG, "sending ERROR_INSUFFICIENT_PERMISSIONS message failed", re);
    169         }
    170         return false;
    171     }
    172 
    173     /**
    174      * Notifies the service that it should start listening for speech.
    175      *
    176      * @param recognizerIntent contains parameters for the recognition to be performed. The intent
    177      *        may also contain optional extras, see {@link RecognizerIntent}. If these values are
    178      *        not set explicitly, default values should be used by the recognizer.
    179      * @param listener that will receive the service's callbacks
    180      */
    181     protected abstract void onStartListening(Intent recognizerIntent, Callback listener);
    182 
    183     /**
    184      * Notifies the service that it should cancel the speech recognition.
    185      */
    186     protected abstract void onCancel(Callback listener);
    187 
    188     /**
    189      * Notifies the service that it should stop listening for speech. Speech captured so far should
    190      * be recognized as if the user had stopped speaking at this point. This method is only called
    191      * if the application calls it explicitly.
    192      */
    193     protected abstract void onStopListening(Callback listener);
    194 
    195     @Override
    196     public final IBinder onBind(final Intent intent) {
    197         if (DBG) Log.d(TAG, "onBind, intent=" + intent);
    198         return mBinder;
    199     }
    200 
    201     @Override
    202     public void onDestroy() {
    203         if (DBG) Log.d(TAG, "onDestroy");
    204         mCurrentCallback = null;
    205         mBinder.clearReference();
    206         super.onDestroy();
    207     }
    208 
    209     /**
    210      * This class receives callbacks from the speech recognition service and forwards them to the
    211      * user. An instance of this class is passed to the
    212      * {@link RecognitionService#onStartListening(Intent, Callback)} method. Recognizers may call
    213      * these methods on any thread.
    214      */
    215     public class Callback {
    216         private final IRecognitionListener mListener;
    217 
    218         private Callback(IRecognitionListener listener) {
    219             mListener = listener;
    220         }
    221 
    222         /**
    223          * The service should call this method when the user has started to speak.
    224          */
    225         public void beginningOfSpeech() throws RemoteException {
    226             if (DBG) Log.d(TAG, "beginningOfSpeech");
    227             mListener.onBeginningOfSpeech();
    228         }
    229 
    230         /**
    231          * The service should call this method when sound has been received. The purpose of this
    232          * function is to allow giving feedback to the user regarding the captured audio.
    233          *
    234          * @param buffer a buffer containing a sequence of big-endian 16-bit integers representing a
    235          *        single channel audio stream. The sample rate is implementation dependent.
    236          */
    237         public void bufferReceived(byte[] buffer) throws RemoteException {
    238             mListener.onBufferReceived(buffer);
    239         }
    240 
    241         /**
    242          * The service should call this method after the user stops speaking.
    243          */
    244         public void endOfSpeech() throws RemoteException {
    245             mListener.onEndOfSpeech();
    246         }
    247 
    248         /**
    249          * The service should call this method when a network or recognition error occurred.
    250          *
    251          * @param error code is defined in {@link SpeechRecognizer}
    252          */
    253         public void error(int error) throws RemoteException {
    254             Message.obtain(mHandler, MSG_RESET).sendToTarget();
    255             mListener.onError(error);
    256         }
    257 
    258         /**
    259          * The service should call this method when partial recognition results are available. This
    260          * method can be called at any time between {@link #beginningOfSpeech()} and
    261          * {@link #results(Bundle)} when partial results are ready. This method may be called zero,
    262          * one or multiple times for each call to {@link SpeechRecognizer#startListening(Intent)},
    263          * depending on the speech recognition service implementation.
    264          *
    265          * @param partialResults the returned results. To retrieve the results in
    266          *        ArrayList&lt;String&gt; format use {@link Bundle#getStringArrayList(String)} with
    267          *        {@link SpeechRecognizer#RESULTS_RECOGNITION} as a parameter
    268          */
    269         public void partialResults(Bundle partialResults) throws RemoteException {
    270             mListener.onPartialResults(partialResults);
    271         }
    272 
    273         /**
    274          * The service should call this method when the endpointer is ready for the user to start
    275          * speaking.
    276          *
    277          * @param params parameters set by the recognition service. Reserved for future use.
    278          */
    279         public void readyForSpeech(Bundle params) throws RemoteException {
    280             mListener.onReadyForSpeech(params);
    281         }
    282 
    283         /**
    284          * The service should call this method when recognition results are ready.
    285          *
    286          * @param results the recognition results. To retrieve the results in {@code
    287          *        ArrayList&lt;String&gt;} format use {@link Bundle#getStringArrayList(String)} with
    288          *        {@link SpeechRecognizer#RESULTS_RECOGNITION} as a parameter
    289          */
    290         public void results(Bundle results) throws RemoteException {
    291             Message.obtain(mHandler, MSG_RESET).sendToTarget();
    292             mListener.onResults(results);
    293         }
    294 
    295         /**
    296          * The service should call this method when the sound level in the audio stream has changed.
    297          * There is no guarantee that this method will be called.
    298          *
    299          * @param rmsdB the new RMS dB value
    300          */
    301         public void rmsChanged(float rmsdB) throws RemoteException {
    302             mListener.onRmsChanged(rmsdB);
    303         }
    304     }
    305 
    306     /** Binder of the recognition service */
    307     private static class RecognitionServiceBinder extends IRecognitionService.Stub {
    308         private RecognitionService mInternalService;
    309 
    310         public RecognitionServiceBinder(RecognitionService service) {
    311             mInternalService = service;
    312         }
    313 
    314         public void startListening(Intent recognizerIntent, IRecognitionListener listener) {
    315             if (DBG) Log.d(TAG, "startListening called by:" + listener.asBinder());
    316             if (mInternalService != null && mInternalService.checkPermissions(listener)) {
    317                 mInternalService.mHandler.sendMessage(Message.obtain(mInternalService.mHandler,
    318                         MSG_START_LISTENING, mInternalService.new StartListeningArgs(
    319                                 recognizerIntent, listener)));
    320             }
    321         }
    322 
    323         public void stopListening(IRecognitionListener listener) {
    324             if (DBG) Log.d(TAG, "stopListening called by:" + listener.asBinder());
    325             if (mInternalService != null && mInternalService.checkPermissions(listener)) {
    326                 mInternalService.mHandler.sendMessage(Message.obtain(mInternalService.mHandler,
    327                         MSG_STOP_LISTENING, listener));
    328             }
    329         }
    330 
    331         public void cancel(IRecognitionListener listener) {
    332             if (DBG) Log.d(TAG, "cancel called by:" + listener.asBinder());
    333             if (mInternalService != null && mInternalService.checkPermissions(listener)) {
    334                 mInternalService.mHandler.sendMessage(Message.obtain(mInternalService.mHandler,
    335                         MSG_CANCEL, listener));
    336             }
    337         }
    338 
    339         public void clearReference() {
    340             mInternalService = null;
    341         }
    342     }
    343 }
    344