1 /*---------------------------------------------------------------------------* 2 * vocab.cpp * 3 * * 4 * Copyright 2007, 2008 Nuance Communciations, Inc. * 5 * * 6 * Licensed under the Apache License, Version 2.0 (the 'License'); * 7 * you may not use this file except in compliance with the License. * 8 * * 9 * You may obtain a copy of the License at * 10 * http://www.apache.org/licenses/LICENSE-2.0 * 11 * * 12 * Unless required by applicable law or agreed to in writing, software * 13 * distributed under the License is distributed on an 'AS IS' BASIS, * 14 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * 15 * See the License for the specific language governing permissions and * 16 * limitations under the License. * 17 * * 18 *---------------------------------------------------------------------------*/ 19 20 #include <string> 21 #include <iostream> 22 #include <stdexcept> 23 #include "ESR_Locale.h" 24 #include "LCHAR.h" 25 #include "pstdio.h" 26 #include "ESR_Session.h" 27 #include "SR_Vocabulary.h" 28 29 #include "vocab.h" 30 31 #define MAX_LINE_LENGTH 256 32 #define MAX_PRONS_LENGTH 1024 33 34 #define DEBUG 0 35 36 #define GENERIC CONTEXT "#" 37 38 Vocabulary::Vocabulary( std::string const & vocFileName ) 39 { 40 ESR_ReturnCode rc; 41 rc = SR_VocabularyLoad(vocFileName.c_str(), &m_hVocab); 42 if (rc != ESR_SUCCESS) 43 { 44 std::cout << "Error: " << ESR_rc2str(rc) <<std::endl; 45 exit (-1); 46 } 47 } 48 49 Vocabulary::~Vocabulary() 50 { 51 SR_VocabularyDestroy(m_hVocab); 52 } 53 54 Pronunciation::Pronunciation() 55 { 56 } 57 58 Pronunciation::~Pronunciation() 59 { 60 } 61 62 void Pronunciation::clear() 63 { 64 m_Prons.clear(); 65 for (unsigned int ii=0;ii<m_ModelIDs.size();ii++ ) 66 { 67 m_ModelIDs[ii].clear(); 68 } 69 m_ModelIDs.clear(); 70 } 71 72 int Pronunciation::lookup( Vocabulary & vocab, std::string & phrase ) 73 { 74 ESR_ReturnCode rc; 75 LCHAR prons[MAX_PRONS_LENGTH]; 76 LCHAR* c_phrase; 77 size_t len; 78 79 LCHAR s[MAX_LINE_LENGTH]; 80 strcpy (s, phrase.c_str() ); // No conversion for std::string to wchar 81 //clear(); 82 83 memset (prons, 0x00, sizeof(LCHAR)); 84 85 c_phrase = s; 86 SR_Vocabulary *p_SRVocab = vocab.getSRVocabularyHandle(); 87 #if DEBUG 88 std::cout << "DEBUG: " << phrase <<" to be looked up" << std::endl; 89 #endif 90 rc = SR_VocabularyGetPronunciation( p_SRVocab, c_phrase, prons, &len ); 91 if (rc != ESR_SUCCESS) 92 // std::cout <<"ERORORORORROOR!" <<std::endl; 93 std::cout <<"ERROR: " << ESR_rc2str(rc) << std::endl; 94 else { 95 #if DEBUG 96 std::cout <<"OUTPUT: " << prons << " num " << len << std::endl; 97 #endif 98 size_t len_used; 99 LCHAR *pron = 0; 100 for(len_used=0; len_used <len; ) { 101 pron = &prons[0]+len_used; 102 len_used += LSTRLEN(pron)+1; 103 #if DEBUG 104 std::cout << "DEBUG: used " << len_used << " now " << LSTRLEN(pron) << std::endl; 105 #endif 106 std::string pronString( pron ); // wstring conversion if needed 107 addPron( pronString ); 108 #if DEBUG 109 std::cout << "DEBUG: " << phrase << " " << pron << std::endl; 110 #endif 111 } 112 } 113 return getPronCount(); 114 } 115 116 117 int Pronunciation::addPron( std::string & s ) 118 { 119 m_Prons.push_back( s ); 120 return m_Prons.size(); 121 } 122 123 int Pronunciation::getPronCount() 124 { // returns number of prons 125 return m_Prons.size(); 126 } 127 128 bool Pronunciation::getPron( int index, std::string &s ) 129 { 130 // returns string length used 131 try { 132 s = m_Prons.at(index); 133 } 134 catch(std::out_of_range& err) { 135 std::cerr << "out_of_range: " << err.what() << std::endl; 136 } 137 return true; 138 } 139 140 void Pronunciation::print() 141 { 142 std::string s; 143 for (int ii=0; ii< getPronCount(); ii++) { 144 getPron(ii, s); 145 #if DEBUG 146 std::cout << "Pron #" << ii << ": " << s << std::endl; 147 #endif 148 } 149 } 150 151 void Pronunciation::printModelIDs() 152 { 153 std::string s; 154 for (int ii=0; ii< getPronCount(); ii++) { 155 getPron(ii, s); 156 #if DEBUG 157 std::cout << " Pron #" << ii << ": " << s << std::endl; 158 std::cout << " Model IDs: "; 159 #endif 160 for (int jj=0;jj<getModelCount(ii);jj++) { 161 std::cout << " " << getModelID(ii,jj); 162 } 163 #if DEBUG 164 std::cout << std::endl; 165 #endif 166 } 167 } 168 169 int Pronunciation::getPhonemeCount( int pronIndex ) 170 { 171 std::string s; 172 getPron(pronIndex, s); 173 return s.size(); 174 } 175 176 bool Pronunciation::getPhoneme( int pronIndex, int picIndex , std::string &phoneme ) 177 { 178 std::string s; 179 getPron(pronIndex, s); 180 phoneme= s.at(picIndex); 181 return true; 182 } 183 184 185 bool Pronunciation::getPIC( int pronIndex, int picIndex, std::string &pic ) 186 { 187 std::string pron; 188 char lphon; 189 char cphon; 190 char rphon; 191 192 getPron( pronIndex, pron ); 193 int numPhonemes = pron.size(); 194 if ( 1==numPhonemes ) { 195 lphon=GENERIC_CONTEXT; 196 rphon=GENERIC_CONTEXT; 197 cphon = pron.at(0); 198 } 199 else 200 { 201 if ( 0==picIndex ) { 202 lphon=GENERIC_CONTEXT; 203 rphon=GENERIC_CONTEXT; 204 } 205 else if( numPhonemes-1==picIndex ) { 206 lphon = pron.at(picIndex-1); 207 rphon=GENERIC_CONTEXT; 208 } 209 else { 210 lphon = pron.at(picIndex-1); 211 rphon = pron.at(picIndex+1); 212 } 213 cphon = pron.at(picIndex); 214 pic = lphon + cphon + rphon; 215 } 216 return true; 217 } 218 219 int Pronunciation::lookupModelIDs( AcousticModel &acoustic ) 220 { 221 // Looks up all hmms for all prons 222 std::string pron; 223 char lphon; 224 char cphon; 225 char rphon; 226 227 int numProns = getPronCount(); 228 int totalCount=0; 229 for (int ii=0;ii < numProns; ii++ ) 230 { 231 getPron( ii, pron ); 232 std::vector<int> idList; // Create storage 233 int numPhonemes = getPhonemeCount(ii); 234 if (1==numPhonemes) { 235 lphon=GENERIC_CONTEXT; 236 rphon=GENERIC_CONTEXT; 237 cphon = pron.at(0); 238 } 239 else 240 for ( int jj=0;jj<numPhonemes;jj++ ) 241 { 242 std::string pic; 243 getPIC(ii, jj, pic); 244 lphon = pron.at(0); 245 cphon = pron.at(1); 246 rphon = pron.at(2); 247 int id = CA_ArbdataGetModelIdsForPIC( acoustic.getCAModelHandle(), lphon, cphon, rphon ); 248 #if DEBUG 249 std::cout <<"DEBUG model id: " << lphon <<cphon << rphon << " "<< id << std::endl; 250 #endif 251 252 idList.push_back(id); 253 } 254 m_ModelIDs.push_back(idList); 255 totalCount+=numPhonemes; 256 } 257 return totalCount; 258 } 259 260 int Pronunciation::getModelCount( int pronIndex ) 261 { 262 return m_ModelIDs[pronIndex].size(); 263 } 264 265 int Pronunciation::getModelID( int pronIndex, int modelPos ) 266 { 267 return m_ModelIDs[pronIndex][modelPos]; 268 } 269 270 AcousticModel::AcousticModel( std::string & arbFileName ) 271 { 272 m_CA_Arbdata = CA_LoadArbdata( arbFileName.c_str() ); 273 if (!m_CA_Arbdata) 274 { 275 std::cout << "Error: while trying to load " << arbFileName.c_str() << std::endl; 276 exit (-1); 277 } 278 279 } 280 281 AcousticModel::~AcousticModel() 282 { 283 CA_FreeArbdata( m_CA_Arbdata); 284 } 285 286 int AcousticModel::getStateIndices(int id, std::vector<int> & stateIDs) 287 { 288 srec_arbdata *allotree = (srec_arbdata*) m_CA_Arbdata; 289 int numStates = allotree->hmm_infos[id].num_states; 290 #if DEBUG 291 std::cout << "getStateIndices: count = " << numStates <<std::endl; 292 #endif 293 for (int ii=0; ii <numStates; ii++ ) { 294 stateIDs.push_back( allotree->hmm_infos[id].state_indices[ii] ); 295 #if DEBUG 296 std::cout << allotree->hmm_infos[id].state_indices[ii] ; 297 #endif 298 } 299 #if DEBUG 300 std::cout << std::endl; 301 #endif 302 return stateIDs.size(); 303 } 304 305
