1 /* 2 * Copyright (C) 2009 Google Inc. 3 * 4 * Licensed under the Apache License, Version 2.0 (the "License"); 5 * you may not use this file except in compliance with the License. 6 * You may obtain a copy of 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, 12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13 * See the License for the specific language governing permissions and 14 * limitations under the License. 15 */ 16 #ifndef ANDROID_TTS_H 17 #define ANDROID_TTS_H 18 19 // This header defines the interface used by the Android platform 20 // to access Text-To-Speech functionality in shared libraries that implement 21 // speech synthesis and the management of resources associated with the 22 // synthesis. 23 24 // The shared library must contain a function named "android_getTtsEngine" 25 // that returns an 'android_tts_engine_t' instance. 26 27 #ifdef __cplusplus 28 extern "C" { 29 #endif 30 31 #include <stdint.h> 32 33 #define ANDROID_TTS_ENGINE_PROPERTY_CONFIG "engineConfig" 34 #define ANDROID_TTS_ENGINE_PROPERTY_PITCH "pitch" 35 #define ANDROID_TTS_ENGINE_PROPERTY_RATE "rate" 36 #define ANDROID_TTS_ENGINE_PROPERTY_VOLUME "volume" 37 38 typedef enum { 39 ANDROID_TTS_SUCCESS = 0, 40 ANDROID_TTS_FAILURE = -1, 41 ANDROID_TTS_FEATURE_UNSUPPORTED = -2, 42 ANDROID_TTS_VALUE_INVALID = -3, 43 ANDROID_TTS_PROPERTY_UNSUPPORTED = -4, 44 ANDROID_TTS_PROPERTY_SIZE_TOO_SMALL = -5, 45 ANDROID_TTS_MISSING_RESOURCES = -6 46 } android_tts_result_t; 47 48 typedef enum { 49 ANDROID_TTS_LANG_COUNTRY_VAR_AVAILABLE = 2, 50 ANDROID_TTS_LANG_COUNTRY_AVAILABLE = 1, 51 ANDROID_TTS_LANG_AVAILABLE = 0, 52 ANDROID_TTS_LANG_MISSING_DATA = -1, 53 ANDROID_TTS_LANG_NOT_SUPPORTED = -2 54 } android_tts_support_result_t; 55 56 typedef enum { 57 ANDROID_TTS_SYNTH_DONE = 0, 58 ANDROID_TTS_SYNTH_PENDING = 1 59 } android_tts_synth_status_t; 60 61 typedef enum { 62 ANDROID_TTS_CALLBACK_HALT = 0, 63 ANDROID_TTS_CALLBACK_CONTINUE = 1 64 } android_tts_callback_status_t; 65 66 // Supported audio formats 67 typedef enum { 68 ANDROID_TTS_AUDIO_FORMAT_INVALID = -1, 69 ANDROID_TTS_AUDIO_FORMAT_DEFAULT = 0, 70 ANDROID_TTS_AUDIO_FORMAT_PCM_16_BIT = 1, 71 ANDROID_TTS_AUDIO_FORMAT_PCM_8_BIT = 2, 72 } android_tts_audio_format_t; 73 74 75 /* An android_tts_engine_t object can be anything, but must have, 76 * as its first field, a pointer to a table of functions. 77 * 78 * See the full definition of struct android_tts_engine_t_funcs_t 79 * below for details. 80 */ 81 typedef struct android_tts_engine_funcs_t android_tts_engine_funcs_t; 82 83 typedef struct { 84 android_tts_engine_funcs_t *funcs; 85 } android_tts_engine_t; 86 87 /* This function must be located in the TTS Engine shared library 88 * and must return the address of an android_tts_engine_t library. 89 */ 90 extern android_tts_engine_t *android_getTtsEngine(); 91 92 /* Including the old version for legacy support (Froyo compatibility). 93 * This should return the same thing as android_getTtsEngine. 94 */ 95 android_tts_engine_t *getTtsEngine(); 96 97 // A callback type used to notify the framework of new synthetized 98 // audio samples, status will be SYNTH_DONE for the last sample of 99 // the last request, of SYNTH_PENDING otherwise. 100 // 101 // This is passed by the framework to the engine through the 102 // 'engine_init' function (see below). 103 // 104 // The callback for synthesis completed takes: 105 // @param [inout] void *& - The userdata pointer set in the original 106 // synth call 107 // @param [in] uint32_t - Track sampling rate in Hz 108 // @param [in] uint32_t - The audio format 109 // @param [in] int - The number of channels 110 // @param [inout] int8_t *& - A buffer of audio data only valid during the 111 // execution of the callback 112 // @param [inout] size_t & - The size of the buffer 113 // @param [in] tts_synth_status - indicate whether the synthesis is done, or 114 // if more data is to be synthesized. 115 // @return TTS_CALLBACK_HALT to indicate the synthesis must stop, 116 // TTS_CALLBACK_CONTINUE to indicate the synthesis must continue if 117 // there is more data to produce. 118 typedef android_tts_callback_status_t (*android_tts_synth_cb_t) 119 (void **pUserData, 120 uint32_t trackSamplingHz, 121 android_tts_audio_format_t audioFormat, 122 int channelCount, 123 int8_t **pAudioBuffer, 124 size_t *pBufferSize, 125 android_tts_synth_status_t status); 126 127 128 // The table of function pointers that the android_tts_engine_t must point to. 129 // Note that each of these functions will take a handle to the engine itself 130 // as their first parameter. 131 // 132 133 struct android_tts_engine_funcs_t { 134 // reserved fields, ignored by the framework 135 // they must be placed here to ensure binary compatibility 136 // of legacy binary plugins. 137 void *reserved[2]; 138 139 // Initialize the TTS engine and returns whether initialization succeeded. 140 // @param synthDoneCBPtr synthesis callback function pointer 141 // @return TTS_SUCCESS, or TTS_FAILURE 142 android_tts_result_t (*init) 143 (void *engine, 144 android_tts_synth_cb_t synthDonePtr, 145 const char *engineConfig); 146 147 // Shut down the TTS engine and releases all associated resources. 148 // @return TTS_SUCCESS, or TTS_FAILURE 149 android_tts_result_t (*shutdown) 150 (void *engine); 151 152 // Interrupt synthesis and flushes any synthesized data that hasn't been 153 // output yet. This will block until callbacks underway are completed. 154 // @return TTS_SUCCESS, or TTS_FAILURE 155 android_tts_result_t (*stop) 156 (void *engine); 157 158 // Returns the level of support for the language, country and variant. 159 // @return TTS_LANG_COUNTRY_VAR_AVAILABLE if the language, country and variant are supported, 160 // and the corresponding resources are correctly installed 161 // TTS_LANG_COUNTRY_AVAILABLE if the language and country are supported and the 162 // corresponding resources are correctly installed, but there is no match for 163 // the specified variant 164 // TTS_LANG_AVAILABLE if the language is supported and the 165 // corresponding resources are correctly installed, but there is no match for 166 // the specified country and variant 167 // TTS_LANG_MISSING_DATA if the required resources to provide any level of support 168 // for the language are not correctly installed 169 // TTS_LANG_NOT_SUPPORTED if the language is not supported by the TTS engine. 170 android_tts_support_result_t (*isLanguageAvailable) 171 (void *engine, 172 const char *lang, 173 const char *country, 174 const char *variant); 175 176 // Load the resources associated with the specified language. The loaded 177 // language will only be used once a call to setLanguage() with the same 178 // language value is issued. Language and country values are coded according to the ISO three 179 // letter codes for languages and countries, as can be retrieved from a java.util.Locale 180 // instance. The variant value is encoded as the variant string retrieved from a 181 // java.util.Locale instance built with that variant data. 182 // @param lang pointer to the ISO three letter code for the language 183 // @param country pointer to the ISO three letter code for the country 184 // @param variant pointer to the variant code 185 // @return TTS_SUCCESS, or TTS_FAILURE 186 android_tts_result_t (*loadLanguage) 187 (void *engine, 188 const char *lang, 189 const char *country, 190 const char *variant); 191 192 // Load the resources associated with the specified language, country and Locale variant. 193 // The loaded language will only be used once a call to setLanguageFromLocale() with the same 194 // language value is issued. Language and country values are coded according to the ISO three 195 // letter codes for languages and countries, as can be retrieved from a java.util.Locale 196 // instance. The variant value is encoded as the variant string retrieved from a 197 // java.util.Locale instance built with that variant data. 198 // @param lang pointer to the ISO three letter code for the language 199 // @param country pointer to the ISO three letter code for the country 200 // @param variant pointer to the variant code 201 // @return TTS_SUCCESS, or TTS_FAILURE 202 android_tts_result_t (*setLanguage) 203 (void *engine, 204 const char *lang, 205 const char *country, 206 const char *variant); 207 208 // Retrieve the currently set language, country and variant, or empty strings if none of 209 // parameters have been set. Language and country are represented by their 3-letter ISO code 210 // @param[out] pointer to the retrieved 3-letter code language value 211 // @param[out] pointer to the retrieved 3-letter code country value 212 // @param[out] pointer to the retrieved variant value 213 // @return TTS_SUCCESS, or TTS_FAILURE 214 android_tts_result_t (*getLanguage) 215 (void *engine, 216 char *language, 217 char *country, 218 char *variant); 219 220 // Notifies the engine what audio parameters should be used for the synthesis. 221 // This is meant to be used as a hint, the engine implementation will set the output values 222 // to those of the synthesis format, based on a given hint. 223 // @param[inout] encoding in: the desired audio sample format 224 // out: the format used by the TTS engine 225 // @param[inout] rate in: the desired audio sample rate 226 // out: the sample rate used by the TTS engine 227 // @param[inout] channels in: the desired number of audio channels 228 // out: the number of channels used by the TTS engine 229 // @return TTS_SUCCESS, or TTS_FAILURE 230 android_tts_result_t (*setAudioFormat) 231 (void *engine, 232 android_tts_audio_format_t* pEncoding, 233 uint32_t* pRate, 234 int* pChannels); 235 236 // Set a property for the the TTS engine 237 // "size" is the maximum size of "value" for properties "property" 238 // @param property pointer to the property name 239 // @param value pointer to the property value 240 // @param size maximum size required to store this type of property 241 // @return TTS_PROPERTY_UNSUPPORTED, or TTS_SUCCESS, or TTS_FAILURE, 242 // or TTS_VALUE_INVALID 243 android_tts_result_t (*setProperty) 244 (void *engine, 245 const char *property, 246 const char *value, 247 const size_t size); 248 249 // Retrieve a property from the TTS engine 250 // @param property pointer to the property name 251 // @param[out] value pointer to the retrieved language value 252 // @param[inout] iosize in: stores the size available to store the 253 // property value. 254 // out: stores the size required to hold the language 255 // value if getLanguage() returned 256 // TTS_PROPERTY_SIZE_TOO_SMALL, unchanged otherwise 257 // @return TTS_PROPERTY_UNSUPPORTED, or TTS_SUCCESS, 258 // or TTS_PROPERTY_SIZE_TOO_SMALL 259 android_tts_result_t (*getProperty) 260 (void *engine, 261 const char *property, 262 char *value, 263 size_t *iosize); 264 265 // Synthesize the text. 266 // As the synthesis is performed, the engine invokes the callback to notify 267 // the TTS framework that it has filled the given buffer, and indicates how 268 // many bytes it wrote. The callback is called repeatedly until the engine 269 // has generated all the audio data corresponding to the text. 270 // Note about the format of the input: the text parameter may use the 271 // following elements 272 // and their respective attributes as defined in the SSML 1.0 specification: 273 // * lang 274 // * say-as: 275 // o interpret-as 276 // * phoneme 277 // * voice: 278 // o gender, 279 // o age, 280 // o variant, 281 // o name 282 // * emphasis 283 // * break: 284 // o strength, 285 // o time 286 // * prosody: 287 // o pitch, 288 // o contour, 289 // o range, 290 // o rate, 291 // o duration, 292 // o volume 293 // * mark 294 // Differences between this text format and SSML are: 295 // * full SSML documents are not supported 296 // * namespaces are not supported 297 // Text is coded in UTF-8. 298 // @param text the UTF-8 text to synthesize 299 // @param userdata pointer to be returned when the call is invoked 300 // @param buffer the location where the synthesized data must be written 301 // @param bufferSize the number of bytes that can be written in buffer 302 // @return TTS_SUCCESS or TTS_FAILURE 303 android_tts_result_t (*synthesizeText) 304 (void *engine, 305 const char *text, 306 int8_t *buffer, 307 size_t bufferSize, 308 void *userdata); 309 }; 310 311 #ifdef __cplusplus 312 } 313 #endif 314 315 #endif /* ANDROID_TTS_H */ 316