eSpeak::eSpeak(std::unique_ptr<NUClear::Environment> environment) : Reactor(std::move(environment)) {
// Initialize espeak, and set it to play out the speakers, and not exit if it can't find it's directory
espeak_Initialize(AUDIO_OUTPUT_PLAYBACK, 500, nullptr, 1 << 15);
espeak_SetVoiceByName("default");
espeak_SetParameter(espeakVOLUME, 100, 0);
espeak_SetParameter(espeakCAPITALS, 6, 0);
on<Trigger<messages::output::Say>, Options<Sync<eSpeak>>>([](const messages::output::Say& message) {
// Wait to finish the current message (if any)
// By waiting here this reaction can finish and return to the pool
// if it does not have to wait for another say message
espeak_Synchronize();
// Say the new message
espeak_Synth(message.c_str(), // Text
message.size() + 1, // Size (including null at end)
0, // Start position
POS_CHARACTER, // Position Type (irrelevant since we start at the beginning)
0, // End position (0 means no end position)
espeakCHARS_AUTO, // Flags (auto encoding)
nullptr, // User identifier for callback
nullptr // Callback
);
});
on<Trigger<Shutdown>>([](const Shutdown&) {
// Stop espeak
espeak_Terminate();
});
}
void espeak_callback(const std_msgs::String::ConstPtr& line) {
// lock mutex before calling espeak functions
boost::mutex::scoped_lock u_lock(mtx);
/* Speak the string */
ROS_INFO("%s", line->data.c_str());
espeak_Synth(line->data.c_str(), line->data.length()+1, 0, POS_CHARACTER, 0,
espeakCHARS_AUTO | espeakPHONEMES | espeakENDPAUSE, NULL, NULL);
espeak_Synchronize();
//ROS_INFO("Speaking: \"%s\"", line->data.c_str());
}
int synchronize() {
espeak_ERROR rc;
if ( 1 == initFlag ) {
rc = espeak_Synchronize();
if ( EE_INTERNAL_ERROR == rc ) {
return -1;
}
}
return 0;
}
int main(int argc, char* argv[] )
{
output = AUDIO_OUTPUT_PLAYBACK;
int I, Run = 1, L;
espeak_Initialize(output, Buflength, path, Options );
espeak_SetVoiceByName(Voice);
Size = strlen(text)+1;
printf("Saying '%s'",text);
espeak_Synth( text, Size, position, position_type, end_position, flags,
unique_identifier, user_data );
espeak_Synchronize( );
printf("\n:Done\n");
return 0;
}
/* TTS annoncement should be in thread otherwise
* it will freez the game till announcemrnt finishes */
int tts_thread_func(void *arg)
{
espeak_POSITION_TYPE position_type = POS_CHARACTER;
tts_argument recived = *((tts_argument*)(arg));
fprintf(stderr,"\nSpeaking : %s - %d",recived.text,recived.mode);
if (recived.mode == INTERRUPT)
T4K_Tts_cancel();
else
T4K_Tts_wait();
int Size = strlen(recived.text)+1;
espeak_Synth(recived.text, Size, 0, position_type, 0, espeakCHARS_AUTO,0, NULL);
espeak_Synchronize();
return 1;
}
/** synthesizeText
* Synthesizes a text string.
* The text string could be annotated with SSML tags.
* @text - text to synthesize
* @buffer - buffer which will receive generated samples
* @bufferSize - size of buffer
* @userdata - pointer to user data which will be passed back to callback function
* return tts_result
*/
tts_result TtsEngine::synthesizeText( const char * text, int8_t * buffer, size_t bufferSize, void * userdata )
{
espeak_SetSynthCallback(eSpeakCallback);
unsigned int unique_identifier;
espeak_ERROR err;
err = espeak_Synth(text,
strlen(text),
0, // position
POS_CHARACTER,
0, // end position (0 means no end position)
espeakCHARS_UTF8,
&unique_identifier,
userdata);
err = espeak_Synchronize();
return TTS_SUCCESS;
}
JNIEXPORT jboolean
JNICALL Java_com_googlecode_eyesfree_espeak_SpeechSynthesis_nativeSynthesize(
JNIEnv *env, jobject object, jstring text) {
if (DEBUG) LOGV("%s", __FUNCTION__);
native_data_t *nat = getNativeData(env, object);
const char *c_text = env->GetStringUTFChars(text, NULL);
unsigned int unique_identifier;
nat->env = env;
espeak_SetSynthCallback(SynthCallback);
espeak_Synth(c_text, strlen(c_text), 0, // position
POS_CHARACTER, 0, // end position (0 means no end position)
espeakCHARS_UTF8 | espeakSSML, // use or ignore xml tags
&unique_identifier, nat);
espeak_Synchronize();
env->ReleaseStringUTFChars(text, c_text);
return JNI_TRUE;
}
int main(int argc, char* argv[] )
{
output = AUDIO_OUTPUT_PLAYBACK;
int I, Run = 1, L;
espeak_Initialize(output, Buflength, path, Options );
//espeak_SetVoiceByName(Voice);
const char *langNativeString = "lt"; //Default to US English
espeak_VOICE voice;
memset(&voice, 0, sizeof(espeak_VOICE)); // Zero out the voice first
voice.languages = langNativeString;
voice.name = "klatt";
voice.variant = 2;
voice.gender = 1;
espeak_SetVoiceByProperties(&voice);
Size = strlen(text)+1;
printf("Saying '%s'",text);
espeak_Synth( text, Size, position, position_type, end_position, flags,
unique_identifier, user_data );
espeak_Synchronize( );
printf("\n:Done\n");
return 0;
}
//-----------------------------------------------------------------------------
// Data definitions
//-----------------------------------------------------------------------------
void RobotSpeak::Speak(const char *psz, bool fWait)
{
// Tell msound to release itself if necessary...
printf("Speak %s\n", psz);
// See if we need to initialize.
if (!_fSpeakInit)
InitSpeak();
else
{
// if it is still playing cancel any active stuff...
if (espeak_IsPlaying())
espeak_Cancel();
}
if (psz)
{
espeak_Synth( psz, strlen(psz)+1, 0, POS_CHARACTER, 0, espeakCHARS_AUTO, &_uSpeakIdentifier, 0 );
if (fWait)
espeak_Synchronize();
}
}
int main (int argc, char **argv)
//==============================
{
static struct option long_options[] =
{
/* These options set a flag. */
// {"verbose", no_argument, &verbose_flag, 1},
// {"brief", no_argument, &verbose_flag, 0},
/* These options don't set a flag.
We distinguish them by their indices. */
{"help", no_argument, 0, 'h'},
{"stdin", no_argument, 0, 0x100},
{"compile-debug", optional_argument, 0, 0x101},
{"compile", optional_argument, 0, 0x102},
{"punct", optional_argument, 0, 0x103},
{"voices", optional_argument, 0, 0x104},
{"stdout", no_argument, 0, 0x105},
{"split", optional_argument, 0, 0x106},
{0, 0, 0, 0}
};
static const char* err_load = "Failed to read ";
FILE *f_text=NULL;
char *p_text=NULL;
int option_index = 0;
int c;
int ix;
int flag_stdin = 0;
int flag_compile = 0;
int filesize = 0;
int synth_flags = espeakCHARS_AUTO | espeakPHONEMES | espeakENDPAUSE;
int volume = -1;
int speed = -1;
int pitch = -1;
int wordgap = -1;
int option_capitals = -1;
int option_punctuation = -1;
int option_phonemes = -1;
int option_linelength = 0;
int option_waveout = 0;
char filename[120];
char voicename[40];
char voice_mbrola[20];
char dictname[40];
#define N_PUNCTLIST 100
wchar_t option_punctlist[N_PUNCTLIST];
voicename[0] = 0;
voice_mbrola[0] = 0;
dictname[0] = 0;
wavefile[0] = 0;
filename[0] = 0;
option_punctlist[0] = 0;
while(true)
{
c = getopt_long (argc, argv, "a:bf:g:hk:l:mp:qs:v:w:xXz",
long_options, &option_index);
/* Detect the end of the options. */
if (c == -1)
break;
switch (c)
{
case 'b':
synth_flags |= espeakCHARS_8BIT;
break;
case 'h':
printf("\n");
printf("eSpeak text-to-speech: %s\n%s",espeak_Info(NULL),help_text);
exit(0);
break;
case 'k':
option_capitals = atoi(optarg);
break;
case 'x':
option_phonemes = 1;
break;
case 'X':
option_phonemes = 2;
break;
case 'm':
synth_flags |= espeakSSML;
break;
case 'p':
pitch = atoi(optarg);
break;
case 'q':
quiet = 1;
break;
case 'f':
strncpy0(filename,optarg,sizeof(filename));
break;
case 'l':
option_linelength = atoi(optarg);
break;
case 'a':
volume = atoi(optarg);
break;
case 's':
speed = atoi(optarg);
break;
case 'g':
wordgap = atoi(optarg);
break;
case 'v':
strncpy0(voicename,optarg,sizeof(voicename));
break;
case 'w':
option_waveout = 1;
strncpy0(wavefile,optarg,sizeof(filename));
break;
case 'z': // remove pause from the end of a sentence
synth_flags &= ~espeakENDPAUSE;
break;
case 0x100: // --stdin
flag_stdin = 1;
break;
case 0x105: // --stdout
option_waveout = 1;
strcpy(wavefile,"stdout");
break;
case 0x101: // --compile-debug
case 0x102: // --compile
strncpy0(voicename,optarg,sizeof(voicename));
flag_compile = c;
quiet = 1;
break;
case 0x103: // --punct
option_punctuation = 1;
if(optarg != NULL)
{
ix = 0;
while((ix < N_PUNCTLIST) && ((option_punctlist[ix] = optarg[ix]) != 0)) ix++;
option_punctlist[N_PUNCTLIST-1] = 0;
option_punctuation = 2;
}
break;
case 0x104: // --voices
espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS,0,NULL,0);
DisplayVoices(stdout,optarg);
exit(0);
case 0x106: // -- split
if(optarg == NULL)
samples_split = 30; // default 30 minutes
else
samples_split = atoi(optarg);
break;
default:
exit(0);
}
}
if(option_waveout || quiet)
{
// writing to a file (or no output), we can use synchronous mode
samplerate = espeak_Initialize(AUDIO_OUTPUT_SYNCHRONOUS,0,NULL,0);
samples_split = (samplerate * samples_split) * 60;
espeak_SetSynthCallback(SynthCallback);
if(samples_split)
{
char *extn;
extn = strrchr(wavefile,'.');
if((extn != NULL) && ((wavefile + strlen(wavefile) - extn) <= 4))
{
strcpy(filetype,extn);
*extn = 0;
}
}
else
if(option_waveout)
{
if(OpenWavFile(wavefile,samplerate) != 0)
exit(4);
}
}
else
{
// play the sound output
samplerate = espeak_Initialize(AUDIO_OUTPUT_PLAYBACK,0,NULL,0);
}
if(voicename[0] == 0)
strcpy(voicename,"default");
if(espeak_SetVoiceByName(voicename) != EE_OK)
{
fprintf(stderr,"%svoice '%s'\n",err_load,voicename);
exit(2);
}
if(flag_compile)
{
// This must be done after the voice is set
espeak_CompileDictionary("", stderr, flag_compile & 0x1);
exit(0);
}
// set any non-default values of parameters. This must be done after espeak_Initialize()
if(speed > 0)
espeak_SetParameter(espeakRATE,speed,0);
if(volume >= 0)
espeak_SetParameter(espeakVOLUME,volume,0);
if(pitch >= 0)
espeak_SetParameter(espeakPITCH,pitch,0);
if(option_capitals >= 0)
espeak_SetParameter(espeakCAPITALS,option_capitals,0);
if(option_punctuation >= 0)
espeak_SetParameter(espeakPUNCTUATION,option_punctuation,0);
if(wordgap >= 0)
espeak_SetParameter(espeakWORDGAP,wordgap,0);
if(option_linelength > 0)
espeak_SetParameter(espeakLINELENGTH,option_linelength,0);
if(option_punctuation == 2)
espeak_SetPunctuationList(option_punctlist);
if(option_phonemes >= 0)
espeak_SetPhonemeTrace(option_phonemes,stderr);
if(filename[0]==0)
{
if((optind < argc) && (flag_stdin == 0))
{
// there's a non-option parameter, and no -f or --stdin
// use it as text
p_text = argv[optind];
}
else
{
f_text = stdin;
if(flag_stdin == 0)
{
flag_stdin = 2;
}
}
}
else
{
filesize = GetFileLength(filename);
f_text = fopen(filename,"r");
}
if((f_text == NULL) && (p_text == NULL))
{
fprintf(stderr,"%sfile '%s'\n",err_load,filename);
exit(1);
}
if(p_text != NULL)
{
int size;
size = strlen(p_text);
espeak_Synth(p_text,size+1,0,POS_CHARACTER,0,synth_flags,NULL,NULL);
}
else
if(flag_stdin)
{
int max = 1000;
p_text = (char *)malloc(max);
if(flag_stdin == 2)
{
// line by line input on stdin
while(fgets(p_text,max,stdin) != NULL)
{
p_text[max-1] = 0;
espeak_Synth(p_text,max,0,POS_CHARACTER,0,synth_flags,NULL,NULL);
}
}
else
{
// bulk input on stdin
ix = 0;
while(!feof(stdin))
{
p_text[ix++] = fgetc(stdin);
if(ix >= (max-1))
{
max += 1000;
p_text = (char *)realloc(p_text,max);
}
}
if(ix > 0)
{
p_text[ix-1] = 0;
espeak_Synth(p_text,ix+1,0,POS_CHARACTER,0,synth_flags,NULL,NULL);
}
}
}
else
if(f_text != NULL)
{
if((p_text = (char *)malloc(filesize+1)) == NULL)
{
fprintf(stderr,"Failed to allocate memory %d bytes",filesize);
exit(3);
}
fread(p_text,1,filesize,f_text);
p_text[filesize]=0;
espeak_Synth(p_text,filesize+1,0,POS_CHARACTER,0,synth_flags,NULL,NULL);
fclose(f_text);
}
espeak_Synchronize();
return(0);
}
static void
spk_drain(volatile SpeechSynthesizer *spk)
{
espeak_Synchronize();
}
