|
1 /* |
|
2 * Copyright (c) 2004-2007 Nokia Corporation and/or its subsidiary(-ies). |
|
3 * All rights reserved. |
|
4 * This component and the accompanying materials are made available |
|
5 * under the terms of "Eclipse Public License v1.0" |
|
6 * which accompanies this distribution, and is available |
|
7 * at the URL "http://www.eclipse.org/legal/epl-v10.html". |
|
8 * |
|
9 * Initial Contributors: |
|
10 * Nokia Corporation - initial contribution. |
|
11 * |
|
12 * Contributors: |
|
13 * |
|
14 * Description: This file contains definition of the DevASR API. |
|
15 * |
|
16 */ |
|
17 |
|
18 |
|
19 #ifndef NSSDEVASRAPI_H |
|
20 #define NSSDEVASRAPI_H |
|
21 |
|
22 //#define AUDIOBUFFER_TO_FILE |
|
23 |
|
24 // INCLUDES |
|
25 #include <e32base.h> |
|
26 #include <nsssispeechrecognitiondatadevasr.h> |
|
27 #include <nssdevasrcommon.h> |
|
28 |
|
29 // FORWARD DECLARATIONS |
|
30 class CSRSAlgorithmManager; |
|
31 class CVMAlgorithmManager; |
|
32 class TMMFPrioritySettings; |
|
33 |
|
34 /** |
|
35 * Interface for the DevASR callbacks that are implemented by DevASR observers. |
|
36 * |
|
37 * @lib NssDevASR.lib |
|
38 * @since 2.8 |
|
39 */ |
|
40 class MDevASRObserver |
|
41 { |
|
42 public: |
|
43 /** |
|
44 * DevASR calls this method to get configuration data. |
|
45 * |
|
46 * @since 2.8 |
|
47 * @param "TUint32 aPackageType" Type identifier. |
|
48 * @param "TUint32 aPackageID" Identifier of package. |
|
49 * @param "TUint32 aStartPosition" Starting position within package in bytes. |
|
50 * @param "TUint32 aEndPosition" Ending position within package in bytes. |
|
51 * @return Pointer to created buffer containing the request data. |
|
52 */ |
|
53 virtual HBufC8* ConfigurationData( TUint32 aPackageType, |
|
54 TUint32 aPackageID, |
|
55 TUint32 aStartPosition = 0, |
|
56 TUint32 aEndPosition = KMaxTUint32 ) = 0; |
|
57 |
|
58 /** |
|
59 * DevASR notifies object implementing this interface whenever a spontaneous or non-spontaneous |
|
60 * event occurs in DevASR. |
|
61 * |
|
62 * @since 2.8 |
|
63 * @param "TDevASREvent aEvent" DevASR event code. |
|
64 * @param "TDevASRError aError" DevASR error code. |
|
65 */ |
|
66 virtual void DevASREvent( TDevASREvent aEvent, TDevASRError aError ) = 0; |
|
67 |
|
68 /** |
|
69 * DevASR notifies object implementing this interface when feature vectors are available. |
|
70 * |
|
71 * @since 2.8 |
|
72 * @param "TDesC16& aFV" Buffer containing the feature vector data. |
|
73 * @param "TInt32 aSNR" Signal-to-noise ratio. |
|
74 * @param "TInt32 aPosition" Indicates whether the feature vector is the first, middle, or end. |
|
75 */ |
|
76 virtual void FeatureVectorDataRcvd( const TDesC8& aFV, TInt32 aSNR, TInt32 aPosition ) = 0; |
|
77 |
|
78 /** |
|
79 * DevASR calls this method to send a message in response to a custom command. |
|
80 * |
|
81 * @since 2.8 |
|
82 * @param "TDesC8& aMsg" Buffer containing the message. |
|
83 */ |
|
84 virtual void DevASRMessage( TDesC8& aMsg ) = 0; |
|
85 |
|
86 /** |
|
87 * Called when DevASR needs a SI lexicon. |
|
88 * |
|
89 * @since 2.8 |
|
90 * @param "TSILexiconID aLexiconID" Lexicon ID of requested lexicon. |
|
91 * @return Pointer to created lexicon. |
|
92 */ |
|
93 virtual CSILexicon* SILexiconL( TSILexiconID aLexiconID ) = 0; |
|
94 |
|
95 /** |
|
96 * Invoked by DevASR when it needs utterance data. The response is sent |
|
97 * using SendSpeechData(). |
|
98 * |
|
99 * @since 2.8 |
|
100 */ |
|
101 virtual void RequestSpeechData() = 0; |
|
102 }; |
|
103 |
|
104 /** |
|
105 * Interface Class for the DevASR API. |
|
106 * |
|
107 * @lib NssDevASR.lib |
|
108 * @since 2.8 |
|
109 */ |
|
110 class CDevASR : public CBase |
|
111 { |
|
112 public: // Constructors and destructor |
|
113 |
|
114 /** |
|
115 * Two-phased constructors. |
|
116 * |
|
117 * @param "MDevASRObserver& aObserver" Reference to callback observer. |
|
118 */ |
|
119 IMPORT_C static CDevASR* NewL( MDevASRObserver& aObserver ); |
|
120 IMPORT_C static CDevASR* NewLC( MDevASRObserver& aObserver ); |
|
121 |
|
122 /** |
|
123 * Destructor. |
|
124 */ |
|
125 IMPORT_C virtual ~CDevASR(); |
|
126 |
|
127 public: // New functions |
|
128 |
|
129 /** |
|
130 * Activates a grammar. |
|
131 * |
|
132 * @since 2.8 |
|
133 * @param "TSIGrammarID aGrammarID" A grammar identifier. |
|
134 */ |
|
135 IMPORT_C void ActivateGrammarL( TSIGrammarID aGrammarID ); |
|
136 |
|
137 /** |
|
138 * Adapts vocabulary and models after successful recognition. |
|
139 * |
|
140 * @since 2.8 |
|
141 * @param "CSIResultSet& aResultSet" Reco result set. |
|
142 * @param "TInt aResultIndex" Index of correct result. |
|
143 */ |
|
144 IMPORT_C void AdaptL( const CSIResultSet& aResultSet, TInt aResultIndex ); |
|
145 |
|
146 /** |
|
147 * Cancels the current or started tasks. |
|
148 * @since 2.8 |
|
149 */ |
|
150 IMPORT_C void Cancel(); |
|
151 |
|
152 /** |
|
153 * Compiles an SI grammar. |
|
154 * |
|
155 * @since 2.8 |
|
156 * @param "CSICompiledGrammar& aGrammar" Reference to a grammar to be compiled. |
|
157 */ |
|
158 IMPORT_C void CompileGrammarL( CSICompiledGrammar& aGrammar ); |
|
159 |
|
160 /** |
|
161 * Computes number of rule variants for grammar. |
|
162 * |
|
163 * @since 2.8 |
|
164 * @param "CSIGrammar& aGrammar" A grammar where to add new rules. |
|
165 * @param "TUint32 aTargetNRuleVariants" Preferable target number of all |
|
166 * rule variants after adding new rule variants. |
|
167 * @param "TUint32 aMaxNRuleVariants" Maximum number of all rule variants. |
|
168 * @param "RArray<TUint>& aNewRuleScores" Scores for each new rule. The |
|
169 * size of array tells also how many new rules will be added. |
|
170 * Based on the scores some rules can get more rule variants than |
|
171 * others. If there is no scores available, the array should |
|
172 * initialize with constant values. |
|
173 * @param "RArray<TUint>& aNewRuleVariants" Array where to add number |
|
174 * of new rule variants for new rules. The order is equivalent |
|
175 * to aNewRuleScores. |
|
176 * @param "TUint32& aNPrune" Number of rule variants which should be |
|
177 * pruned before adding new rule variants. |
|
178 */ |
|
179 IMPORT_C void ComputeNewGrammarSizeL( const CSIGrammar& aGrammar, |
|
180 const TUint32 aTargetNRuleVariants, |
|
181 const TUint32 aMaxNRuleVariants, |
|
182 const RArray<TUint>& aNewRuleScores, |
|
183 RArray<TUint>& aNewRuleVariants, |
|
184 TUint32& aNPrune ); |
|
185 |
|
186 /** |
|
187 * Use to request a custom interface implementation from the component. |
|
188 * through the DevASR API. |
|
189 * |
|
190 * @since 2.8 |
|
191 * @param "TUid aInterfaceId" Interface UID, defined with the custom |
|
192 * interface. |
|
193 * @return A pointer to the interface implementation, or NULL if the |
|
194 * requested interface is not implemented. The return value must |
|
195 * be cast to the correct type by the user. |
|
196 */ |
|
197 IMPORT_C TAny* CustomInterface( TUid aInterfaceId ); |
|
198 |
|
199 /** |
|
200 * Deactivates a grammar. |
|
201 * |
|
202 * @since 2.8 |
|
203 * @param "TSIGrammarID aGrammarID" A grammar identifier. |
|
204 */ |
|
205 IMPORT_C void DeactivateGrammarL( TSIGrammarID aGrammarID ); |
|
206 |
|
207 /** |
|
208 * Requests to end recording and a result will be resolved. |
|
209 * |
|
210 * @since 2.8 |
|
211 */ |
|
212 IMPORT_C void EndRecord(); |
|
213 |
|
214 /** |
|
215 * This method is used to indicate the end of a recognition session. All loaded grammars, |
|
216 * lexicon, and models become invalid after this call. |
|
217 * |
|
218 * @since 2.8 |
|
219 */ |
|
220 IMPORT_C void EndRecSession(); |
|
221 |
|
222 /** |
|
223 * Retreive the properties of the underlying speech recognition engine. |
|
224 * |
|
225 * @since 2.8 |
|
226 * @param "RArray<TInt>& aPropertyId" An array of identifiers being querried. |
|
227 * @param "RArray<TInt>& aPropertyValue" An array of values corresponding |
|
228 * to the querried identifiers. |
|
229 */ |
|
230 IMPORT_C void GetEnginePropertiesL( const RArray<TInt>& aPropertyId, |
|
231 RArray<TInt>& aPropertyValue ); |
|
232 |
|
233 /** |
|
234 * Retreive the raw audio data accumulated during recording. |
|
235 * |
|
236 * @since 2.8 |
|
237 * @param "TDes8* aBuffer" The buffer where utterance data will be |
|
238 * written. |
|
239 * @param "TTimeIntervalMicroSeconds32& aDuration" Duration of the |
|
240 * utterance. |
|
241 */ |
|
242 IMPORT_C void GetUtteranceData( TDes8* aBuffer, |
|
243 TTimeIntervalMicroSeconds32& aDuration ); |
|
244 |
|
245 /** |
|
246 * Initializes the front-end module in the speech recognition engine. |
|
247 * The frontend module used during training/recognition functions is started |
|
248 * as a result. This method is intended to be used in conjunction with InitTrainBE(). |
|
249 * |
|
250 * @since 2.8 |
|
251 * @param "TRecognizerMode aMode" Mode of the recognizer. |
|
252 */ |
|
253 IMPORT_C void InitFrontEnd( TRecognizerMode aMode ); |
|
254 |
|
255 /** |
|
256 * Initialize the recognition engine back-end. The module responsible for recognition |
|
257 * function is started as a result. This method must be used before any recognition |
|
258 * operations and intended to be used in conjunction with InitFrontEnd(). |
|
259 * |
|
260 * @since 2.8 |
|
261 * @param "CSIResultSet& aResult" A reference to an object where the |
|
262 * recognition result will be written. |
|
263 */ |
|
264 IMPORT_C void InitRecognizerBE( CSIResultSet& aResult ); |
|
265 |
|
266 /** |
|
267 * Load the specified recognizer parameter(s). These parameters are used to alter the |
|
268 * recognizer's default parameters. The parameters are specified as attribute-value pairs. |
|
269 * |
|
270 * @since 2.8 |
|
271 * @param "RArray<TInt>& aParameterId" An array of parameter identifiers. |
|
272 * @param "RArray<TInt>& aParameterValue" An array of parameter values. |
|
273 */ |
|
274 IMPORT_C void LoadEngineParametersL( const RArray<TInt>& aParameterId, |
|
275 const RArray<TInt>& aParameterValue ); |
|
276 |
|
277 /** |
|
278 * Load the specified grammar into the recognizer. |
|
279 * |
|
280 * @since 2.8 |
|
281 * @param "CSIGrammar& aGrammar" A reference to a SI grammar. |
|
282 */ |
|
283 IMPORT_C void LoadGrammar( const CSIGrammar& aGrammar ); |
|
284 |
|
285 /** |
|
286 * Load the specified grammar into the recognizer. |
|
287 * |
|
288 * @since 2.8 |
|
289 * @param "CSICompiledGrammar& aGrammar" A reference to a compiled |
|
290 * SI grammar |
|
291 */ |
|
292 IMPORT_C void LoadGrammar( const CSICompiledGrammar& aGrammar ); |
|
293 |
|
294 /** |
|
295 * Unloads the specified grammar from the recognizer. |
|
296 * |
|
297 * @since 2.8 |
|
298 * @param "CSIGrammar& aGrammar" A reference to a SI grammar. |
|
299 */ |
|
300 IMPORT_C void UnloadGrammar( const CSIGrammar& aGrammar ); |
|
301 |
|
302 /** |
|
303 * Unloads the specified grammar from the recognizer. |
|
304 * |
|
305 * @since 2.8 |
|
306 * @param "CSICompiledGrammar& aGrammar" A reference to a compiled |
|
307 * SI grammar |
|
308 */ |
|
309 IMPORT_C void UnloadGrammar( const CSICompiledGrammar& aGrammar ); |
|
310 |
|
311 /** |
|
312 * Load the specified SI lexicon into the recognizer. |
|
313 * |
|
314 * @since 2.8 |
|
315 * @param "CSILexicon& aLexicon" A reference to a SI lexicon. |
|
316 */ |
|
317 IMPORT_C void LoadLexicon( const CSILexicon& aLexicon ); |
|
318 |
|
319 /** |
|
320 * Load the specified model bank into the recognizer. |
|
321 * |
|
322 * @since 2.8 |
|
323 * @param "CSIModelBank& aModels" A reference to a SI model bank. |
|
324 */ |
|
325 IMPORT_C void LoadModels( const CSIModelBank& aModels ); |
|
326 |
|
327 /** |
|
328 * Determines which rule variants can be removed. |
|
329 * |
|
330 * @since 2.8 |
|
331 * @param "CSIGrammar& aGrammar" A grammar to be pruned. |
|
332 * @param "TUint32 aMinNumber" Minimum number of pruned items. |
|
333 * @param "RArray<TSIRuleVariantInfo>" A array where to put information |
|
334 * which variants can be pruned. |
|
335 */ |
|
336 IMPORT_C TBool PruneGrammar( const CSIGrammar& aGrammar, |
|
337 const TUint32 aMinNumber, |
|
338 RArray<TSIRuleVariantInfo>& aPrunedRuleVariants ); |
|
339 |
|
340 /** |
|
341 * Request to begin recognition. |
|
342 * |
|
343 * @since 2.8 |
|
344 * @param "TTimeIntervalMicroSeconds32 aRecordDuration" Length of time |
|
345 * to record, expressed in microseconds. Zero means infinite |
|
346 * duration. |
|
347 */ |
|
348 IMPORT_C void Record( TTimeIntervalMicroSeconds32 aRecordDuration ); |
|
349 |
|
350 /** |
|
351 * Starts sampling before Record call. |
|
352 * |
|
353 * @since 3.2 |
|
354 */ |
|
355 IMPORT_C void PreStartSamplingL(); |
|
356 |
|
357 /** |
|
358 * Use to send the utterance data as a response to the RequestSpeechData(). |
|
359 * |
|
360 * @since 2.8 |
|
361 * @param "TDesC8& aBuffer" Buffer containing utterance data. |
|
362 * @param "TBool aEnd" Flag to tell if given buffer is the last one. |
|
363 */ |
|
364 IMPORT_C void SendSpeechData( TPtrC8& aBuffer, TBool aEnd ); |
|
365 |
|
366 /** |
|
367 * Use to set the priority of the sound device. |
|
368 * |
|
369 * @since 2.8 |
|
370 * @param "TMMFPrioritySettings& aPrioritySettings" Priority settings |
|
371 * structure. |
|
372 */ |
|
373 IMPORT_C void SetPrioritySettings( const TMMFPrioritySettings& aPrioritySettings ); |
|
374 |
|
375 /** |
|
376 * This method is used to start a recognition session with the recognition engine. |
|
377 * |
|
378 * @since 2.8 |
|
379 * @param "TRecognizerMode aMode" Mode of the recognizer. |
|
380 * @return Result code of operation. |
|
381 */ |
|
382 IMPORT_C TInt StartRecSession( TRecognizerMode aMode ); |
|
383 |
|
384 /** |
|
385 * Starts training from text |
|
386 * |
|
387 * @since 2.8 |
|
388 * @param "CSITtpWordList& aWordList" List of text entries for training. |
|
389 * @param "RArray<TLanguage>& aDefaultLanguage" Default language for |
|
390 * pronunciations. |
|
391 * @param "RArray<TUint>& aMaxNPronunsForWord" Maximum number of |
|
392 * pronunciation per word. |
|
393 */ |
|
394 IMPORT_C void StartTrainingFromTextL( CSITtpWordList& aWordList, |
|
395 const RArray<TLanguage>& aDefaultLanguage, |
|
396 const RArray<TUint32>& aMaxNPronunsForWord ); |
|
397 |
|
398 /** |
|
399 * Request to unload the specified rule in the grammar from recognizer. |
|
400 * |
|
401 * @since 2.8 |
|
402 * @param "TSIGrammarID aGrammarID" Identifer of the grammar to operate on. |
|
403 * @param "TSIRuleID aRuleID" Identifier of the rule to remove. |
|
404 */ |
|
405 IMPORT_C void UnloadRule( TSIGrammarID aGrammarID, TSIRuleID aRuleID ); |
|
406 |
|
407 /** |
|
408 * Checks if grammar is loaded or not |
|
409 * |
|
410 * @since 2.8 |
|
411 * @param "TSIGrammarID aGrammarID" Grammar ID |
|
412 * @return ETrue if grammar is currently loaded, EFalse otherwise. |
|
413 */ |
|
414 IMPORT_C TBool IsGrammarLoaded( TSIGrammarID aGrammarID ); |
|
415 |
|
416 /** |
|
417 * Checks if grammar is active or not |
|
418 * |
|
419 * @since 2.8 |
|
420 * @param "TSIGrammarID aGrammarID" Grammar ID |
|
421 * @return ETrue if grammar is active, EFalse otherwise. |
|
422 */ |
|
423 IMPORT_C TBool IsGrammarActive( TSIGrammarID aGrammarID ); |
|
424 |
|
425 private: |
|
426 |
|
427 /** |
|
428 * C++ default constructor. |
|
429 */ |
|
430 CDevASR(); |
|
431 |
|
432 /** |
|
433 * By default Symbian 2nd phase constructor is private. |
|
434 */ |
|
435 void ConstructL(MDevASRObserver& aObserver); |
|
436 |
|
437 private: // Data |
|
438 |
|
439 // Pointer to recognition algorithm manager |
|
440 CSRSAlgorithmManager* iSRSAlgorithmManager; |
|
441 |
|
442 // Reserved for future use |
|
443 TAny* iReserved1; |
|
444 TAny* iReserved2; |
|
445 |
|
446 }; |
|
447 |
|
448 #endif // NSSDEVASRAPI_H |
|
449 |
|
450 // End of File |