core/java/android/speech/tts/TextToSpeechClient.java

/*
 * Copyright (C) 2013 The Android Open Source Project
 *
 * Licensed under the Apache License, Version 2.0 (the "License"); you may not
 * use this file except in compliance with the License. You may obtain a copy of
 * the License at
 *
 * http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
 * License for the specific language governing permissions and limitations under
 * the License.
 */

package android.speech.tts;

import android.app.Activity;
import android.app.Application;
import android.content.ComponentName;
import android.content.Context;
import android.content.Intent;
import android.content.ServiceConnection;
import android.media.AudioManager;
import android.net.Uri;
import android.os.AsyncTask;
import android.os.IBinder;
import android.os.ParcelFileDescriptor;
import android.os.RemoteException;
import android.speech.tts.ITextToSpeechCallback;
import android.speech.tts.ITextToSpeechService;
import android.util.Log;
import android.util.Pair;

import java.io.File;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.concurrent.atomic.AtomicInteger;

/**
 * Synthesizes speech from text for immediate playback or to create a sound
 * file.
 * <p>
 * This is an updated version of the speech synthesis client that supersedes
 * {@link android.speech.tts.TextToSpeech}.
 * <p>
 * A TextToSpeechClient instance can only be used to synthesize text once it has
 * connected to the service. The TextToSpeechClient instance will start establishing
 * the connection after a call to the {@link #connect()} method. This is usually done in
 * {@link Application#onCreate()} or {@link Activity#onCreate}. When the connection
 * is established, the instance will call back using the
 * {@link TextToSpeechClient.ConnectionCallbacks} interface. Only after a
 * successful callback is the client usable.
 * <p>
 * After successful connection, the list of all available voices can be obtained
 * by calling the {@link TextToSpeechClient#getEngineStatus() method. The client can
 * choose a voice using some custom heuristic and build a {@link RequestConfig} object
 * using {@link RequestConfig.Builder}, or can use one of the common heuristics found
 * in ({@link RequestConfigHelper}.
 * <p>
 * When you are done using the TextToSpeechClient instance, call the
 * {@link #disconnect()} method to release the connection.
 * <p>
 * In the rare case of a change to the set of available voices, the service will call to the
 * {@link ConnectionCallbacks#onEngineStatusChange} with new set of available voices as argument.
 * In response, the client HAVE to recreate all {@link RequestConfig} instances in use.
 */
public final class TextToSpeechClient {
    private static final String TAG = TextToSpeechClient.class.getSimpleName();

    private final Object mLock = new Object();
    private final TtsEngines mEnginesHelper;
    private final Context mContext;

    // Guarded by mLock
    private Connection mServiceConnection;
    private final RequestCallbacks mDefaultRequestCallbacks;
    private final ConnectionCallbacks mConnectionCallbacks;
    private EngineStatus mEngineStatus;
    private String mRequestedEngine;
    private boolean mFallbackToDefault;
    private HashMap<String, Pair<UtteranceId, RequestCallbacks>> mCallbacks;
    // Guarded by mLock

    /** Common voices parameters */
    public static final class Params {
        private Params() {}

        /**
         * Maximum allowed time for a single request attempt, in milliseconds, before synthesis
         * fails (or fallback request starts, if requested using
         * {@link #FALLBACK_VOICE_NAME}).
         */
        public static final String NETWORK_TIMEOUT_MS = "networkTimeoutMs";

        /**
         * Number of network request retries that are attempted in case of failure
         */
        public static final String NETWORK_RETRIES_COUNT = "networkRetriesCount";

        /**
         * Should synthesizer report sub-utterance progress on synthesis. Only applicable
         * for the {@link TextToSpeechClient#queueSpeak} method.
         */
        public static final String TRACK_SUBUTTERANCE_PROGRESS = "trackSubutteranceProgress";

        /**
         * If a voice exposes this parameter then it supports the fallback request feature.
         *
         * If it is set to a valid name of some other voice ({@link VoiceInfo#getName()}) then
         * in case of request failure (due to network problems or missing data), fallback request
         * will be attempted. Request will be done using the voice referenced by this parameter.
         * If it is the case, the client will be informed by a callback to the {@link
         * RequestCallbacks#onSynthesisFallback(UtteranceId)}.
         */
        public static final String FALLBACK_VOICE_NAME = "fallbackVoiceName";

        /**
         * Audio parameter for specifying a linear multiplier to the speaking speed of the voice.
         * The value is a float. Values below zero decrease speed of the synthesized speech
         * values above one increase it. If the value of this parameter is equal to zero,
         * then it will be replaced by a settings-configurable default before it reaches
         * TTS service.
         */
        public static final String SPEECH_SPEED = "speechSpeed";

        /**
         * Audio parameter for controlling the pitch of the output. The Value is a positive float,
         * with default of {@code 1.0}. The value is used to scale the primary frequency linearly.
         * Lower values lower the tone of the synthesized voice, greater values increase it.
         */
        public static final String SPEECH_PITCH = "speechPitch";

        /**
         * Audio parameter for controlling output volume. Value is a float with scale of 0 to 1
         */
        public static final String AUDIO_PARAM_VOLUME = TextToSpeech.Engine.KEY_PARAM_VOLUME;

        /**
         * Audio parameter for controlling output pan.
         * Value is a float ranging from -1 to +1 where -1 maps to a hard-left pan,
         * 0 to center (the default behavior), and +1 to hard-right.
         */
        public static final String AUDIO_PARAM_PAN = TextToSpeech.Engine.KEY_PARAM_PAN;

        /**
         * Audio parameter for specifying the audio stream type to be used when speaking text
         * or playing back a file. The value should be one of the STREAM_ constants
         * defined in {@link AudioManager}.
         */
        public static final String AUDIO_PARAM_STREAM = TextToSpeech.Engine.KEY_PARAM_STREAM;
    }

    /**
     * Result codes for TTS operations.
     */
    public static final class Status {
        private Status() {}

        /**
         * Denotes a successful operation.
         */
        public static final int SUCCESS = 0;

        /**
         * Denotes a stop requested by a client. It's used only on the service side of the API,
         * client should never expect to see this result code.
         */
        public static final int STOPPED = 100;

        /**
         * Denotes a generic failure.
         */
        public static final int ERROR_UNKNOWN = -1;

        /**
         * Denotes a failure of a TTS engine to synthesize the given input.
         */
        public static final int ERROR_SYNTHESIS = 10;

        /**
         * Denotes a failure of a TTS service.
         */
        public static final int ERROR_SERVICE = 11;

        /**
         * Denotes a failure related to the output (audio device or a file).
         */
        public static final int ERROR_OUTPUT = 12;

        /**
         * Denotes a failure caused by a network connectivity problems.
         */
        public static final int ERROR_NETWORK = 13;

        /**
         * Denotes a failure caused by network timeout.
         */
        public static final int ERROR_NETWORK_TIMEOUT = 14;

        /**
         * Denotes a failure caused by an invalid request.
         */
        public static final int ERROR_INVALID_REQUEST = 15;

        /**
         * Denotes a failure related to passing a non-unique utterance id.
         */
        public static final int ERROR_NON_UNIQUE_UTTERANCE_ID = 16;

        /**
         * Denotes a failure related to missing data. The TTS implementation may download
         * the missing data, and if so, request will succeed in future. This error can only happen
         * for voices with {@link VoiceInfo#FEATURE_MAY_AUTOINSTALL} feature.
         * Note: the recommended way to avoid this error is to create a request with the fallback
         * voice.
         */
        public static final int ERROR_DOWNLOADING_ADDITIONAL_DATA = 17;
    }

    /**
     * Set of callbacks for the events related to the progress of a synthesis request
     * through the synthesis queue. Each synthesis request is associated with a call to
     * {@link #queueSpeak} or {@link #queueSynthesizeToFile}.
     *
     * The callbacks specified in this method will NOT be called on UI thread.
     */
    public static abstract class RequestCallbacks  {
        /**
         * Called after synthesis of utterance successfully starts.
         */
        public void onSynthesisStart(UtteranceId utteranceId) {}

        /**
         * Called after synthesis successfully finishes.
         * @param utteranceId
         *            Unique identifier of synthesized utterance.
         */
        public void onSynthesisSuccess(UtteranceId utteranceId) {}

        /**
         * Called after synthesis was stopped in middle of synthesis process.
         * @param utteranceId
         *            Unique identifier of synthesized utterance.
         */
        public void onSynthesisStop(UtteranceId utteranceId) {}

        /**
         * Called when requested synthesis failed and fallback synthesis is about to be attempted.
         *
         * Requires voice with available {@link TextToSpeechClient.Params#FALLBACK_VOICE_NAME}
         * parameter, and request with this parameter enabled.
         *
         * This callback will be followed by callback to the {@link #onSynthesisStart},
         * {@link #onSynthesisFailure} or {@link #onSynthesisSuccess} that depends on the
         * fallback outcome.
         *
         * For more fallback feature reference, look at the
         * {@link TextToSpeechClient.Params#FALLBACK_VOICE_NAME}.
         *
         * @param utteranceId
         *            Unique identifier of synthesized utterance.
         */
        public void onSynthesisFallback(UtteranceId utteranceId) {}

        /**
         * Called after synthesis of utterance fails.
         *
         * It may be called instead or after a {@link #onSynthesisStart} callback.
         *
         * @param utteranceId
         *            Unique identifier of synthesized utterance.
         * @param errorCode
         *            One of the values from {@link Status}.
         */
        public void onSynthesisFailure(UtteranceId utteranceId, int errorCode) {}

        /**
         * Called during synthesis to mark synthesis progress.
         *
         * Requires voice with available
         * {@link TextToSpeechClient.Params#TRACK_SUBUTTERANCE_PROGRESS} parameter, and
         * request with this parameter enabled.
         *
         * @param utteranceId
         *            Unique identifier of synthesized utterance.
         * @param charIndex
         *            String index (java char offset) of recently synthesized character.
         * @param msFromStart
         *            Miliseconds from the start of the synthesis.
         */
        public void onSynthesisProgress(UtteranceId utteranceId, int charIndex,
                int msFromStart) {}
    }

    /**
     * Interface definition of callbacks that are called when the client is
     * connected or disconnected from the TTS service.
     */
    public static interface ConnectionCallbacks {
        /**
         * After calling {@link TextToSpeechClient#connect()}, this method will be invoked
         * asynchronously when the connect request has successfully completed.
         *
         * Clients are strongly encouraged to call {@link TextToSpeechClient#getEngineStatus()}
         * and create {@link RequestConfig} objects used in subsequent synthesis requests.
         */
        public void onConnectionSuccess();

        /**
         * After calling {@link TextToSpeechClient#connect()}, this method may be invoked
         * asynchronously when the connect request has failed to complete.
         *
         * It may be also invoked synchronously, from the body of
         * {@link TextToSpeechClient#connect()} method.
         */
        public void onConnectionFailure();

        /**
         * Called when the connection to the service is lost. This can happen if there is a problem
         * with the speech service (e.g. a crash or resource problem causes it to be killed by the
         * system). When called, all requests have been canceled and no outstanding listeners will
         * be executed. Applications should disable UI components that require the service.
         */
        public void onServiceDisconnected();

        /**
         * After receiving {@link #onConnectionSuccess()} callback, this method may be invoked
         * if engine status obtained from {@link TextToSpeechClient#getEngineStatus()}) changes.
         * It usually means that some voices were removed, changed or added.
         *
         * Clients are required to recreate {@link RequestConfig} objects used in subsequent
         * synthesis requests.
         */
        public void onEngineStatusChange(EngineStatus newEngineStatus);
    }

    /** State of voices as provided by engine and user. */
    public static final class EngineStatus {
        /** All available voices. */
        private final List<VoiceInfo> mVoices;

        /** Name of the TTS engine package */
        private final String mPackageName;

        private EngineStatus(String packageName, List<VoiceInfo> voices) {
            this.mVoices =  Collections.unmodifiableList(voices);
            this.mPackageName = packageName;
        }

        /**
         * Get an immutable list of all Voices exposed by the TTS engine.
         */
        public List<VoiceInfo> getVoices() {
            return mVoices;
        }

        /**
         * Get name of the TTS engine package currently in use.
         */
        public String getEnginePackage() {
            return mPackageName;
        }
    }

    /** Unique synthesis request identifier. */
    public static class UtteranceId {
        /** Unique identifier */
        private final int id;

        /** Unique identifier generator */
        private static final AtomicInteger ID_GENERATOR = new AtomicInteger();

        /**
         * Create new, unique UtteranceId instance.
         */
        public UtteranceId() {
            id = ID_GENERATOR.getAndIncrement();
        }

        /**
         * Returns a unique string associated with an instance of this object.
         *
         * This string will be used to identify the synthesis request/utterance inside the
         * TTS service.
         */
        public final String toUniqueString() {
            return "UID" + id;
        }
    }

    /**
     * Create TextToSpeech service client.
     *
     * Will connect to the default TTS service. In order to be usable, {@link #connect()} need
     * to be called first and successful connection callback need to be received.
     *
     * @param context
     *            The context this instance is running in.
     * @param engine
     *            Package name of requested TTS engine. If it's null, then default engine will
     *            be selected regardless of {@code fallbackToDefaultEngine} parameter value.
     * @param fallbackToDefaultEngine
     *            If requested engine is not available, should we fallback to the default engine?
     * @param defaultRequestCallbacks
     *            Default request callbacks, it will be used for all synthesis requests without
     *            supplied RequestCallbacks instance. Can't be null.
     * @param connectionCallbacks
     *            Callbacks for connecting and disconnecting from the service. Can't be null.
     */
    public TextToSpeechClient(Context context,
            String engine, boolean fallbackToDefaultEngine,
            RequestCallbacks defaultRequestCallbacks,
            ConnectionCallbacks connectionCallbacks) {
        if (context == null)
            throw new IllegalArgumentException("context can't be null");
        if (defaultRequestCallbacks == null)
            throw new IllegalArgumentException("defaultRequestCallbacks can't be null");
        if (connectionCallbacks == null)
            throw new IllegalArgumentException("connectionCallbacks can't be null");
        mContext = context;
        mEnginesHelper = new TtsEngines(mContext);
        mCallbacks = new HashMap<String, Pair<UtteranceId, RequestCallbacks>>();
        mDefaultRequestCallbacks = defaultRequestCallbacks;
        mConnectionCallbacks = connectionCallbacks;

        mRequestedEngine = engine;
        mFallbackToDefault = fallbackToDefaultEngine;
    }

    /**
     * Create TextToSpeech service client. Will connect to the default TTS
     * service. In order to be usable, {@link #connect()} need to be called
     * first and successful connection callback need to be received.
     *
     * @param context Context this instance is running in.
     * @param defaultRequestCallbacks Default request callbacks, it
     *            will be used for all synthesis requests without supplied
     *            RequestCallbacks instance. Can't be null.
     * @param connectionCallbacks Callbacks for connecting and disconnecting
     *            from the service. Can't be null.
     */
    public TextToSpeechClient(Context context, RequestCallbacks defaultRequestCallbacks,
            ConnectionCallbacks connectionCallbacks) {
        this(context, null, true, defaultRequestCallbacks, connectionCallbacks);
    }


    private boolean initTts(String requestedEngine, boolean fallbackToDefaultEngine) {
        // Step 1: Try connecting to the engine that was requested.
        if (requestedEngine != null) {
            if (mEnginesHelper.isEngineInstalled(requestedEngine)) {
                if ((mServiceConnection = connectToEngine(requestedEngine)) != null) {
                    return true;
                } else if (!fallbackToDefaultEngine) {
                    Log.w(TAG, "Couldn't connect to requested engine: " + requestedEngine);
                    return false;
                }
            } else if (!fallbackToDefaultEngine) {
                Log.w(TAG, "Requested engine not installed: " + requestedEngine);
                return false;
            }
        }

        // Step 2: Try connecting to the user's default engine.
        final String defaultEngine = mEnginesHelper.getDefaultEngine();
        if (defaultEngine != null && !defaultEngine.equals(requestedEngine)) {
            if ((mServiceConnection = connectToEngine(defaultEngine)) != null) {
                return true;
            }
        }

        // Step 3: Try connecting to the highest ranked engine in the
        // system.
        final String highestRanked = mEnginesHelper.getHighestRankedEngineName();
        if (highestRanked != null && !highestRanked.equals(requestedEngine) &&
                !highestRanked.equals(defaultEngine)) {
            if ((mServiceConnection = connectToEngine(highestRanked)) != null) {
                return true;
            }
        }

        Log.w(TAG, "Couldn't find working TTS engine");
        return false;
    }

    private Connection connectToEngine(String engine) {
        Connection connection = new Connection(engine);
        Intent intent = new Intent(TextToSpeech.Engine.INTENT_ACTION_TTS_SERVICE);
        intent.setPackage(engine);
        boolean bound = mContext.bindService(intent, connection, Context.BIND_AUTO_CREATE);
        if (!bound) {
            Log.e(TAG, "Failed to bind to " + engine);
            return null;
        } else {
            Log.i(TAG, "Successfully bound to " + engine);
            return connection;
        }
    }


    /**
     * Connects the client to TTS service. This method returns immediately, and connects to the
     * service in the background.
     *
     * After connection initializes successfully, {@link ConnectionCallbacks#onConnectionSuccess()}
     * is called. On a failure {@link ConnectionCallbacks#onConnectionFailure} is called.
     *
     * Both of those callback may be called asynchronously on the main thread,
     * {@link ConnectionCallbacks#onConnectionFailure} may be called synchronously, before
     * this method returns.
     */
    public void connect() {
        synchronized (mLock) {
            if (mServiceConnection != null) {
                return;
            }
            if(!initTts(mRequestedEngine, mFallbackToDefault)) {
                mConnectionCallbacks.onConnectionFailure();
            }
        }
    }

    /**
     * Checks if the client is currently connected to the service, so that
     * requests to other methods will succeed.
     */
    public boolean isConnected() {
        synchronized (mLock) {
            return mServiceConnection != null && mServiceConnection.isEstablished();
        }
    }

    /**
     * Closes the connection to TextToSpeech service. No calls can be made on this object after
     * calling this method.
     * It is good practice to call this method in the onDestroy() method of an Activity
     * so the TextToSpeech engine can be cleanly stopped.
     */
    public void disconnect() {
        synchronized (mLock) {
            if (mServiceConnection != null) {
                mServiceConnection.disconnect();
                mServiceConnection = null;
                mCallbacks.clear();
            }
        }
    }

    /**
     * Register callback.
     *
     * @param utteranceId Non-null UtteranceId instance.
     * @param callback Non-null callbacks for the request
     * @return Status.SUCCESS or error code in case of invalid arguments.
     */
    private int addCallback(UtteranceId utteranceId, RequestCallbacks callback) {
        synchronized (mLock) {
            if (utteranceId == null || callback == null) {
                return Status.ERROR_INVALID_REQUEST;
            }
            if (mCallbacks.put(utteranceId.toUniqueString(),
                    new Pair<UtteranceId, RequestCallbacks>(utteranceId, callback)) != null) {
                return Status.ERROR_NON_UNIQUE_UTTERANCE_ID;
            }
            return Status.SUCCESS;
        }
    }

    /**
     * Remove and return callback.
     *
     * @param utteranceIdStr Unique string obtained from {@link UtteranceId#toUniqueString}.
     */
    private Pair<UtteranceId, RequestCallbacks> removeCallback(String utteranceIdStr) {
        synchronized (mLock) {
            return mCallbacks.remove(utteranceIdStr);
        }
    }

    /**
     * Get callback and utterance id.
     *
     * @param utteranceIdStr Unique string obtained from {@link UtteranceId#toUniqueString}.
     */
    private Pair<UtteranceId, RequestCallbacks> getCallback(String utteranceIdStr) {
        synchronized (mLock) {
            return mCallbacks.get(utteranceIdStr);
        }
    }

    /**
     * Remove callback and call {@link RequestCallbacks#onSynthesisFailure} with passed
     * error code.
     *
     * @param utteranceIdStr Unique string obtained from {@link UtteranceId#toUniqueString}.
     * @param errorCode argument to {@link RequestCallbacks#onSynthesisFailure} call.
     */
    private void removeCallbackAndErr(String utteranceIdStr, int errorCode) {
        synchronized (mLock) {
            Pair<UtteranceId, RequestCallbacks> c = mCallbacks.remove(utteranceIdStr);
            c.second.onSynthesisFailure(c.first, errorCode);
        }
    }

    /**
     * Retrieve TTS engine status {@link EngineStatus}. Requires connected client.
     */
    public EngineStatus getEngineStatus() {
        synchronized (mLock) {
            return mEngineStatus;
        }
    }

    /**
     * Query TTS engine about available voices and defaults.
     *
     * @return EngineStatus is connected or null if client is disconnected.
     */
    private EngineStatus requestEngineStatus(ITextToSpeechService service)
            throws RemoteException {
        List<VoiceInfo> voices = service.getVoicesInfo();
        if (voices == null) {
            Log.e(TAG, "Requested engine doesn't support TTS V2 API");
            return null;
        }

        return new EngineStatus(mServiceConnection.getEngineName(), voices);
    }

    private class Connection implements ServiceConnection {
        private final String mEngineName;

        private ITextToSpeechService mService;

        private boolean mEstablished;

        private PrepareConnectionAsyncTask mSetupConnectionAsyncTask;

        public Connection(String engineName) {
            this.mEngineName = engineName;
        }

        private final ITextToSpeechCallback.Stub mCallback = new ITextToSpeechCallback.Stub() {

            @Override
            public void onStart(String utteranceIdStr) {
                synchronized (mLock) {
                    Pair<UtteranceId, RequestCallbacks> callbacks = getCallback(utteranceIdStr);
                    callbacks.second.onSynthesisStart(callbacks.first);
                }
            }

            public void onStop(String utteranceIdStr) {
                synchronized (mLock) {
                    Pair<UtteranceId, RequestCallbacks> callbacks = removeCallback(utteranceIdStr);
                    callbacks.second.onSynthesisStop(callbacks.first);
                }
            }

            @Override
            public void onSuccess(String utteranceIdStr) {
                synchronized (mLock) {
                    Pair<UtteranceId, RequestCallbacks> callbacks = removeCallback(utteranceIdStr);
                    callbacks.second.onSynthesisSuccess(callbacks.first);
                }
            }

            public void onFallback(String utteranceIdStr) {
                synchronized (mLock) {
                    Pair<UtteranceId, RequestCallbacks> callbacks = getCallback(
                            utteranceIdStr);
                    callbacks.second.onSynthesisFallback(callbacks.first);
                }
            };

            @Override
            public void onError(String utteranceIdStr, int errorCode) {
                removeCallbackAndErr(utteranceIdStr, errorCode);
            }

            @Override
            public void onVoicesInfoChange(List<VoiceInfo> voicesInfo) {
                synchronized (mLock) {
                    mEngineStatus = new EngineStatus(mServiceConnection.getEngineName(),
                            voicesInfo);
                    mConnectionCallbacks.onEngineStatusChange(mEngineStatus);
                }
            }
        };

        private class PrepareConnectionAsyncTask extends AsyncTask<Void, Void, EngineStatus> {

            private final ComponentName mName;

            public PrepareConnectionAsyncTask(ComponentName name) {
                mName = name;
            }

            @Override
            protected EngineStatus doInBackground(Void... params) {
                synchronized(mLock) {
                    if (isCancelled()) {
                        return null;
                    }
                    try {
                        mService.setCallback(getCallerIdentity(), mCallback);
                        return requestEngineStatus(mService);
                    } catch (RemoteException re) {
                        Log.e(TAG, "Error setting up the TTS service");
                        return null;
                    }
                }
            }

            @Override
            protected void onPostExecute(EngineStatus result) {
                synchronized(mLock) {
                    if (mSetupConnectionAsyncTask == this) {
                        mSetupConnectionAsyncTask = null;
                    }
                    if (result == null) {
                        Log.e(TAG, "Setup task failed");
                        disconnect();
                        mConnectionCallbacks.onConnectionFailure();
                        return;
                    }

                    mEngineStatus = result;
                    mEstablished = true;
                }
                mConnectionCallbacks.onConnectionSuccess();
            }
        }

        @Override
        public void onServiceConnected(ComponentName name, IBinder service) {
            Log.i(TAG, "Connected to " + name);

            synchronized(mLock) {
                mEstablished = false;
                mService = ITextToSpeechService.Stub.asInterface(service);
                startSetupConnectionTask(name);
            }
        }

        @Override
        public void onServiceDisconnected(ComponentName name) {
            Log.i(TAG, "Asked to disconnect from " + name);

            synchronized(mLock) {
                stopSetupConnectionTask();
            }
            mConnectionCallbacks.onServiceDisconnected();
        }

        private void startSetupConnectionTask(ComponentName name) {
            stopSetupConnectionTask();
            mSetupConnectionAsyncTask = new PrepareConnectionAsyncTask(name);
            mSetupConnectionAsyncTask.execute();
        }

        private boolean stopSetupConnectionTask() {
            boolean result = false;
            if (mSetupConnectionAsyncTask != null) {
                result = mSetupConnectionAsyncTask.cancel(false);
                mSetupConnectionAsyncTask = null;
            }
            return result;
        }

        IBinder getCallerIdentity() {
            return mCallback;
        }

        boolean isEstablished() {
            return mService != null && mEstablished;
        }

        boolean runAction(Action action) {
            synchronized (mLock) {
                try {
                    action.run(mService);
                    return true;
                } catch (Exception ex) {
                    Log.e(TAG, action.getName() + " failed", ex);
                    disconnect();
                    return false;
                }
            }
        }

        void disconnect() {
            mContext.unbindService(this);
            stopSetupConnectionTask();
            mService = null;
            mEstablished = false;
            if (mServiceConnection == this) {
                mServiceConnection = null;
            }
        }

        String getEngineName() {
            return mEngineName;
        }
    }

    private abstract class Action {
        private final String mName;

        public Action(String name) {
            mName = name;
        }

        public String getName() {return mName;}
        abstract void run(ITextToSpeechService service) throws RemoteException;
    }

    private IBinder getCallerIdentity() {
        if (mServiceConnection != null) {
            return mServiceConnection.getCallerIdentity();
        }
        return null;
    }

    private boolean runAction(Action action) {
        synchronized (mLock) {
            if (mServiceConnection == null) {
                return false;
            }
            if (!mServiceConnection.isEstablished()) {
                return false;
            }
            mServiceConnection.runAction(action);
            return true;
        }
    }

    private static final String ACTION_STOP_NAME = "stop";

    /**
     * Interrupts the current utterance spoken (whether played or rendered to file) and discards
     * other utterances in the queue.
     */
    public void stop() {
        runAction(new Action(ACTION_STOP_NAME) {
            @Override
            public void run(ITextToSpeechService service) throws RemoteException {
               if (service.stop(getCallerIdentity()) != Status.SUCCESS) {
                   Log.e(TAG, "Stop failed");
               }
               mCallbacks.clear();
            }
        });
    }

    private static final String ACTION_QUEUE_SPEAK_NAME = "queueSpeak";

    /**
     * Speaks the string using the specified queuing strategy using current
     * voice. This method is asynchronous, i.e. the method just adds the request
     * to the queue of TTS requests and then returns. The synthesis might not
     * have finished (or even started!) at the time when this method returns.
     *
     * @param utterance The string of text to be spoken. No longer than
     *            1000 characters.
     * @param utteranceId Unique identificator used to track the synthesis progress
     *            in {@link RequestCallbacks}.
     * @param config Synthesis request configuration. Can't be null. Has to contain a
     *            voice.
     * @param callbacks Synthesis request callbacks. If null, default request
     *            callbacks object will be used.
     */
    public void queueSpeak(final String utterance, final UtteranceId utteranceId,
            final RequestConfig config,
            final RequestCallbacks callbacks) {
        runAction(new Action(ACTION_QUEUE_SPEAK_NAME) {
            @Override
            public void run(ITextToSpeechService service) throws RemoteException {
                RequestCallbacks c = mDefaultRequestCallbacks;
                if (callbacks != null) {
                    c = callbacks;
                }
                int addCallbackStatus = addCallback(utteranceId, c);
                if (addCallbackStatus != Status.SUCCESS) {
                    c.onSynthesisFailure(utteranceId, Status.ERROR_INVALID_REQUEST);
                    return;
                }

                int queueResult = service.speakV2(
                        getCallerIdentity(),
                        new SynthesisRequestV2(utterance, utteranceId.toUniqueString(), config));
                if (queueResult != Status.SUCCESS) {
                    removeCallbackAndErr(utteranceId.toUniqueString(), queueResult);
                }
            }
        });
    }

    private static final String ACTION_QUEUE_SYNTHESIZE_TO_FILE = "queueSynthesizeToFile";

    /**
     * Synthesizes the given text to a file using the specified parameters. This
     * method is asynchronous, i.e. the method just adds the request to the
     * queue of TTS requests and then returns. The synthesis might not have
     * finished (or even started!) at the time when this method returns.
     *
     * @param utterance The text that should be synthesized. No longer than
     *            1000 characters.
     * @param utteranceId Unique identificator used to track the synthesis progress
     *            in {@link RequestCallbacks}.
     * @param outputFile File to write the generated audio data to.
     * @param config Synthesis request configuration. Can't be null. Have to contain a
     *            voice.
     * @param callbacks Synthesis request callbacks. If null, default request
     *            callbacks object will be used.
     */
    public void queueSynthesizeToFile(final String utterance, final UtteranceId utteranceId,
            final File outputFile, final RequestConfig config,
            final RequestCallbacks callbacks) {
        runAction(new Action(ACTION_QUEUE_SYNTHESIZE_TO_FILE) {
            @Override
            public void run(ITextToSpeechService service) throws RemoteException {
                RequestCallbacks c = mDefaultRequestCallbacks;
                if (callbacks != null) {
                    c = callbacks;
                }
                int addCallbackStatus = addCallback(utteranceId, c);
                if (addCallbackStatus != Status.SUCCESS) {
                    c.onSynthesisFailure(utteranceId, Status.ERROR_INVALID_REQUEST);
                    return;
                }

                ParcelFileDescriptor fileDescriptor = null;
                try {
                    if (outputFile.exists() && !outputFile.canWrite()) {
                        Log.e(TAG, "No permissions to write to " + outputFile);
                        removeCallbackAndErr(utteranceId.toUniqueString(), Status.ERROR_OUTPUT);
                        return;
                    }
                    fileDescriptor = ParcelFileDescriptor.open(outputFile,
                            ParcelFileDescriptor.MODE_WRITE_ONLY |
                            ParcelFileDescriptor.MODE_CREATE |
                            ParcelFileDescriptor.MODE_TRUNCATE);

                    int queueResult = service.synthesizeToFileDescriptorV2(getCallerIdentity(),
                            fileDescriptor,
                            new SynthesisRequestV2(utterance, utteranceId.toUniqueString(),
                                    config));
                    fileDescriptor.close();
                    if (queueResult != Status.SUCCESS) {
                        removeCallbackAndErr(utteranceId.toUniqueString(), queueResult);
                    }
                } catch (FileNotFoundException e) {
                    Log.e(TAG, "Opening file " + outputFile + " failed", e);
                    removeCallbackAndErr(utteranceId.toUniqueString(), Status.ERROR_OUTPUT);
                } catch (IOException e) {
                    Log.e(TAG, "Closing file " + outputFile + " failed", e);
                    removeCallbackAndErr(utteranceId.toUniqueString(), Status.ERROR_OUTPUT);
                }
            }
        });
    }

    private static final String ACTION_QUEUE_SILENCE_NAME = "queueSilence";

    /**
     * Plays silence for the specified amount of time. This method is asynchronous,
     * i.e. the method just adds the request to the queue of TTS requests and then
     * returns. The synthesis might not have finished (or even started!) at the time
     * when this method returns.
     *
     * @param durationInMs The duration of the silence in milliseconds.
     * @param utteranceId Unique identificator used to track the synthesis progress
     *            in {@link RequestCallbacks}.
     * @param callbacks Synthesis request callbacks. If null, default request
     *            callbacks object will be used.
     */
    public void queueSilence(final long durationInMs, final UtteranceId utteranceId,
            final RequestCallbacks callbacks) {
        runAction(new Action(ACTION_QUEUE_SILENCE_NAME) {
            @Override
            public void run(ITextToSpeechService service) throws RemoteException {
                RequestCallbacks c = mDefaultRequestCallbacks;
                if (callbacks != null) {
                    c = callbacks;
                }
                int addCallbackStatus = addCallback(utteranceId, c);
                if (addCallbackStatus != Status.SUCCESS) {
                    c.onSynthesisFailure(utteranceId, Status.ERROR_INVALID_REQUEST);
                }

                int queueResult = service.playSilence(getCallerIdentity(), durationInMs,
                        TextToSpeech.QUEUE_ADD, utteranceId.toUniqueString());

                if (queueResult != Status.SUCCESS) {
                    removeCallbackAndErr(utteranceId.toUniqueString(), queueResult);
                }
            }
        });
    }


    private static final String ACTION_QUEUE_AUDIO_NAME = "queueAudio";

    /**
     * Plays the audio resource using the specified parameters.
     * This method is asynchronous, i.e. the method just adds the request to the queue of TTS
     * requests and then returns. The synthesis might not have finished (or even started!) at the
     * time when this method returns.
     *
     * @param audioUrl The audio resource that should be played
     * @param utteranceId Unique identificator used to track synthesis progress
     *            in {@link RequestCallbacks}.
     * @param config Synthesis request configuration. Can't be null. Doesn't have to contain a
     *            voice (only system parameters are used).
     * @param callbacks Synthesis request callbacks. If null, default request
     *            callbacks object will be used.
     */
    public void queueAudio(final Uri audioUrl, final UtteranceId utteranceId,
            final RequestConfig config, final RequestCallbacks callbacks) {
        runAction(new Action(ACTION_QUEUE_AUDIO_NAME) {
            @Override
            public void run(ITextToSpeechService service) throws RemoteException {
                RequestCallbacks c = mDefaultRequestCallbacks;
                if (callbacks != null) {
                    c = callbacks;
                }
                int addCallbackStatus = addCallback(utteranceId, c);
                if (addCallbackStatus != Status.SUCCESS) {
                    c.onSynthesisFailure(utteranceId, Status.ERROR_INVALID_REQUEST);
                }

                int queueResult = service.playAudioV2(getCallerIdentity(), audioUrl,
                        utteranceId.toUniqueString(), config.getVoiceParams());

                if (queueResult != Status.SUCCESS) {
                    removeCallbackAndErr(utteranceId.toUniqueString(), queueResult);
                }
            }
        });
    }
}