00001 #include "rospeex_if/rospeex.h"
00002
00003 #include <stdio.h>
00004 #include <stdint.h>
00005 #include <sstream>
00006
00007 #include <sys/stat.h>
00008
00009 #include <ros/ros.h>
00010 #include <ros/console.h>
00011 #include <ros/package.h>
00012
00013 #include <boost/format.hpp>
00014 #include <boost/filesystem.hpp>
00015
00016
00017 #include "rospeex_msgs/SpeechSynthesisResponse.h"
00018 #include "rospeex_msgs/SpeechSynthesisRequest.h"
00019 #include "rospeex_msgs/SpeechSynthesisHeader.h"
00020 #include "rospeex_msgs/SpeechSynthesisState.h"
00021 #include "rospeex_msgs/SpeechRecognitionResponse.h"
00022 #include "rospeex_msgs/SpeechRecognitionRequest.h"
00023 #include "rospeex_msgs/SignalProcessingResponse.h"
00024 #include "rospeex_msgs/SpeechRecognitionConfig.h"
00025
00026 using namespace rospeex;
00027 using namespace rospeex_msgs;
00028
00029 #define SR_RESPONSE_TOPIC_NAME "sr_res"
00030 #define SS_RESPONSE_TOPIC_NAME "ss_res"
00031 #define SPI_RESPONSE_TOPIC_NAME "spi_res"
00032 #define SR_REQUEST_TOPIC_NAME "sr_req"
00033 #define SS_REQUEST_TOPIC_NAME "ss_req"
00034 #define SS_STATE_TOPIC_NAME "ss_state"
00035 #define SPI_CONFIG_SERVICE_NAME "spi_config"
00036
00037
00038 class Interface::Impl
00039 {
00040 public:
00041 Impl()
00042 : sr_func_()
00043 , ss_func_()
00044 , ss_request_id_(0)
00045 , sr_request_id_(0)
00046 , sr_pub_()
00047 , ss_pub_()
00048 , ss_state_pub_()
00049 , sr_sub_()
00050 , ss_sub_()
00051 , spi_sub_()
00052 , ss_enable_()
00053 , sr_enable_()
00054 , spi_enable_()
00055 {}
00056
00061 void SPIResponse( const SignalProcessingResponse::ConstPtr& response );
00062
00067 void SRResponse( const SpeechRecognitionResponse::ConstPtr& response );
00068
00073 void SSResponse( const SpeechSynthesisResponse::ConstPtr& response );
00074
00080 void playPackageSound( const std::string& package_name );
00081
00086 void playSound( const std::string& file_name );
00087
00092 void sendSpeechSynthesisState( bool state );
00093
00094 uint32_t ss_request_id_;
00095 uint32_t sr_request_id_;
00096
00097 SpeechRecognizeCallback sr_func_;
00098 SpeechSynthesisCallback ss_func_;
00099
00100 ros::Publisher sr_pub_;
00101 ros::Publisher ss_pub_;
00102 ros::Publisher ss_state_pub_;
00103
00104 ros::Subscriber sr_sub_;
00105 ros::Subscriber ss_sub_;
00106 ros::Subscriber spi_sub_;
00107
00108 std::string spi_engine_;
00109 std::string spi_language_;
00110
00111 bool ss_enable_, sr_enable_, spi_enable_;
00112 };
00113
00114
00115 void Interface::Impl::SPIResponse( const SignalProcessingResponse::ConstPtr& response )
00116 {
00117 if ( sr_func_.empty() == false ) {
00118 playPackageSound("accept.wav");
00119 sr_request_id_++;
00120
00121 } else {
00122
00123 }
00124 }
00125
00126
00127 void Interface::Impl::SRResponse( const SpeechRecognitionResponse::ConstPtr& response )
00128 {
00129
00130 if( response->message == "" ) {
00131 playPackageSound("nomessage.wav");
00132
00133 } else if ( spi_enable_ != true ) {
00134 playPackageSound("accept.wav");
00135 }
00136
00137 ros::NodeHandle n;
00138 std::string user = response->header.user;
00139 bool check_user = ( user == "spi" || user == n.getNamespace() );
00140 if ( sr_func_.empty() == false && check_user == true) {
00141 sr_func_( response->message );
00142
00143 } else {
00144
00145 }
00146 }
00147
00148
00149 void Interface::Impl::SSResponse( const SpeechSynthesisResponse::ConstPtr& response )
00150 {
00151 ros::NodeHandle n;
00152 if ( ss_func_.empty() == false && response->header.user == n.getNamespace() ) {
00153 ss_func_( response->data );
00154
00155 } else {
00156
00157 }
00158 }
00159
00160
00161 Interface::Interface()
00162 : impl_( new Impl() )
00163 {}
00164
00165
00166 Interface::~Interface()
00167 {}
00168
00169
00170 bool Interface::init( bool ss_enable, bool sr_enable, bool spi_enable )
00171 {
00172 impl_ = boost::shared_ptr<Impl>(new Impl());
00173
00174 ros::NodeHandle n;
00175 impl_->ss_enable_ = ss_enable;
00176 impl_->sr_enable_ = sr_enable;
00177 impl_->spi_enable_ = spi_enable;
00178
00179
00180 if ( sr_enable == true ) {
00181 ROS_INFO("enable speeech recognition.");
00182 impl_->sr_pub_ = n.advertise<SpeechRecognitionRequest>(SR_REQUEST_TOPIC_NAME, 10);
00183 boost::function<void(const SpeechRecognitionResponse::ConstPtr&)> sr_func = boost::bind(&Impl::SRResponse, this->impl_, _1);
00184 impl_->sr_sub_ = n.subscribe(SR_RESPONSE_TOPIC_NAME, 10, sr_func);
00185 }
00186
00187
00188 if ( ss_enable == true ) {
00189 ROS_INFO("enable speech synthsis.");
00190 impl_->ss_pub_ = n.advertise<SpeechSynthesisRequest>(SS_REQUEST_TOPIC_NAME, 10);
00191 boost::function<void(const SpeechSynthesisResponse::ConstPtr&)> ss_func = boost::bind(&Impl::SSResponse, this->impl_, _1);
00192 impl_->ss_sub_ = n.subscribe(SS_RESPONSE_TOPIC_NAME, 10, ss_func);
00193 }
00194
00195
00196 if ( spi_enable == true ) {
00197 ROS_INFO("enable signal processing interface.");
00198 boost::function<void(const SignalProcessingResponse::ConstPtr&)> spi_func = boost::bind(&Impl::SPIResponse, this->impl_, _1);
00199 impl_->ss_state_pub_ = n.advertise<SpeechSynthesisState>(SS_STATE_TOPIC_NAME, 10);
00200 impl_->spi_sub_ = n.subscribe(SPI_RESPONSE_TOPIC_NAME, 10, spi_func);
00201 }
00202
00203 return true;
00204 }
00205
00206
00207 void Interface::Impl::playPackageSound( const std::string& sound_file )
00208 {
00209
00210 sendSpeechSynthesisState(true);
00211
00212
00213 std::string package_path = ros::package::getPath("rospeex_if");
00214 boost::filesystem::path sound_path(package_path + "/sound/" + sound_file);
00215 playSound(sound_path.string());
00216
00217
00218 sendSpeechSynthesisState(false);
00219 }
00220
00221
00222 void Interface::Impl::sendSpeechSynthesisState( bool state )
00223 {
00224
00225 if ( ss_state_pub_ ) {
00226 ros::NodeHandle n;
00227 SpeechSynthesisState ss_state;
00228 ss_state.header.language = "";
00229 ss_state.header.voice_font = "";
00230 ss_state.header.engine = "";
00231 ss_state.header.user = n.getNamespace();
00232 ss_state.header.request_id = "0";
00233 ss_state.header.request_type = SpeechSynthesisHeader::REQUEST_TYPE_SAY;
00234 ss_state.play_state = state;
00235 ss_state_pub_.publish(ss_state);
00236 }
00237 }
00238
00239
00240 void Interface::Impl::playSound( const std::string& file_path )
00241 {
00242
00243 boost::system::error_code error;
00244 bool is_exist = boost::filesystem::exists(file_path, error);
00245
00246 if ( error ) {
00247 ROS_ERROR("file[%s] open error. %s",file_path.c_str() ,error.message().c_str());
00248 return;
00249 }
00250
00251 if ( !is_exist ) {
00252 ROS_ERROR("[%s] is not exist.", file_path.c_str());
00253 return;
00254 }
00255
00256
00257 std::string command = (boost::format("aplay -q %s") % file_path).str();
00258 int ret = system(command.c_str());
00259 if(ret != 0){
00260 ROS_ERROR("playing sound error.");
00261 }
00262 }
00263
00264
00265 void Interface::say( const std::string& msg,
00266 const std::string& language,
00267 const std::string& engine,
00268 const std::string& voice_font)
00269 {
00270 if ( impl_->ss_pub_ ) {
00271 ros::NodeHandle n;
00272 std::stringstream ss;
00273 ss << impl_->ss_request_id_;
00274 SpeechSynthesisRequest request;
00275 request.header.language = language;
00276 request.header.voice_font = voice_font;
00277 request.header.engine = engine;
00278 request.header.user = n.getNamespace();
00279 request.header.request_id = ss.str();
00280 request.header.request_type = SpeechSynthesisHeader::REQUEST_TYPE_SAY;
00281 request.message = msg;
00282 impl_->ss_pub_.publish(request);
00283 impl_->ss_request_id_++;
00284
00285 } else {
00286 ROS_INFO("ss interface is disabled.");
00287 }
00288 }
00289
00290
00291 void Interface::tts( const std::string& msg,
00292 const std::string& file,
00293 const std::string& language,
00294 const std::string& engine,
00295 const std::string& voice_font )
00296 {
00297 if ( impl_->ss_pub_ ) {
00298 ros::NodeHandle n;
00299 std::stringstream ss;
00300 ss << impl_->ss_request_id_;
00301 SpeechSynthesisRequest request;
00302 request.header.language = language;
00303 request.header.voice_font = voice_font;
00304 request.header.engine = engine;
00305 request.header.user = n.getNamespace();
00306 request.header.request_id = ss.str();
00307 request.header.request_type = SpeechSynthesisHeader::REQUEST_TYPE_TTS;
00308 request.message = msg;
00309 impl_->ss_pub_.publish(request);
00310 impl_->ss_request_id_++;
00311
00312 } else {
00313 ROS_INFO("ss interface is disabled.");
00314 }
00315 }
00316
00317
00318 void Interface::recognize( const std::string& data,
00319 const std::string& language,
00320 const std::string& engine )
00321 {
00322 if ( impl_->sr_pub_ ) {
00323 ros::NodeHandle n;
00324 std::stringstream ss;
00325 ss << impl_->sr_request_id_;
00326 SpeechRecognitionRequest request;
00327 request.header.language = language;
00328 request.header.engine = engine;
00329 request.header.user = n.getNamespace();
00330 request.header.request_id = ss.str();
00331 request.data = data;
00332 impl_->sr_pub_.publish(request);
00333 impl_->sr_request_id_++;
00334
00335 } else {
00336 ROS_INFO("sr interface is disabled.");
00337 }
00338 }
00339
00340
00341 void Interface::registerSRResponse( const SpeechRecognizeCallback& callback )
00342 {
00343 impl_->sr_func_ = callback;
00344 }
00345
00346
00347 void Interface::registerSRResponse( void(*fp)(const std::string& msg) )
00348 {
00349 impl_->sr_func_ = SpeechRecognizeCallback(fp);
00350 }
00351
00352
00353 void Interface::registerSSResponse( const SpeechSynthesisCallback& callback )
00354 {
00355 impl_->ss_func_ = callback;
00356 }
00357
00358
00359 void Interface::registerSSResponse( void(*fp)(const std::string& data) )
00360 {
00361 impl_->ss_func_ = SpeechSynthesisCallback(fp);
00362 }
00363
00364
00365 void Interface::setSPIConfig( const std::string& language, const std::string& engine)
00366 {
00367 impl_->spi_language_ = language;
00368 impl_->spi_engine_ = engine;
00369
00370 ros::service::waitForService(SPI_CONFIG_SERVICE_NAME, -1);
00371 ros::NodeHandle nh;
00372 ros::ServiceClient client = nh.serviceClient<SpeechRecognitionConfig>(SPI_CONFIG_SERVICE_NAME);
00373 SpeechRecognitionConfig srv;
00374 srv.request.engine = engine;
00375 srv.request.language = language;
00376 if ( client.call(srv) ) {
00377 ROS_DEBUG("Success to call service %s", SPI_CONFIG_SERVICE_NAME);
00378
00379 } else {
00380 ROS_ERROR("Failed to call service %s", SPI_CONFIG_SERVICE_NAME);
00381 }
00382 }
00383
00384
00385 void Interface::playSound( const std::string& file_name )
00386 {
00387 impl_->playSound(file_name);
00388 }