react-native-davoice-tts 1.0.404 → 1.0.406

package/android/libs/com/davoice/tts/1.0.0/tts-1.0.0.aar.md5

@@ -1 +1 @@
-d219717b8d1f9541dea0d60e7954c1c2  tts-1.0.0.aar
+e92b94f52a72a258f8b1e37289e0a779  tts-1.0.0.aar

package/android/libs/com/davoice/tts/1.0.0/tts-1.0.0.aar.sha1

@@ -1 +1 @@
-fb63148803a298828cf7b574378c1ca275812753  tts-1.0.0.aar
+83fce3d52b29ef2d53b9fc34694262264af77495  tts-1.0.0.aar

package/android/src/main/java/com/davoice/stt/rn/STTModule.kt

@@ -35,19 +35,81 @@ class STTModule(private val rc: ReactApplicationContext)
     return stt!!
   }
 
+  // The native side accepts the JSON itself or a path to it.
   private fun readOnboardingJson(options: ReadableMap?): String? {
-    var onboardingJson: String? = null
-    try {
-      if (options != null && options.hasKey("onboardingJsonPath") && !options.isNull("onboardingJsonPath")) {
-        val p = options.getString("onboardingJsonPath")
-        if (!p.isNullOrBlank()) {
-          onboardingJson = java.io.File(p).readText(Charsets.UTF_8)
-        }
+    if (options == null || !options.hasKey("onboardingJsonPath") || options.isNull("onboardingJsonPath")) return null
+    return options.getString("onboardingJsonPath")?.takeIf { it.isNotBlank() }
+  }
+
+  private fun readSpeakerId(options: ReadableMap?): String =
+    options?.takeIf { it.hasKey("speakerId") && !it.isNull("speakerId") }?.getString("speakerId")
+      ?.takeIf { it.isNotBlank() } ?: "default"
+
+  // JS [{id, json}] -> List<Map<String, String?>> (json = JSON string or path).
+  private fun toSpeakerItems(arr: ReadableArray?): List<Map<String, String?>> {
+    if (arr == null) return emptyList()
+    val out = ArrayList<Map<String, String?>>()
+    for (i in 0 until arr.size()) {
+      val m = arr.getMap(i) ?: continue
+      val item = HashMap<String, String?>()
+      val it = m.keySetIterator()
+      while (it.hasNextKey()) {
+        val k = it.nextKey()
+        if (m.getType(k) == ReadableType.String) item[k] = m.getString(k)
       }
-    } catch (_: Throwable) {
-      onboardingJson = null
+      out.add(item)
     }
-    return onboardingJson
+    return out
+  }
+
+  // `speakers` replaces the set; else `onboardingJsonPath` (+ `speakerId`) adds one
+  // speaker; else the current set (if any) is kept.
+  private fun startWithOptions(locale: String?, options: ReadableMap?) {
+    val speakers = options?.takeIf { it.hasKey("speakers") && !it.isNull("speakers") }?.getArray("speakers")
+    if (speakers != null && speakers.size() > 0) {
+      ensure().startSpeech(locale, toSpeakerItems(speakers))
+    } else {
+      ensure().startSpeech(locale, readOnboardingJson(options), null, readSpeakerId(options))
+    }
+  }
+
+  private fun speakerIdsArray(): WritableArray =
+    Arguments.createArray().apply { (stt?.getSpeakerIds() ?: emptyList()).forEach { pushString(it) } }
+
+  @ReactMethod
+  fun addSpeaker(speakerId: String?, jsonOrPath: String?, promise: Promise) {
+    try {
+      ensure().addSpeaker(speakerId ?: "default", jsonOrPath ?: "")
+      promise.resolve(speakerIdsArray())
+    } catch (t: Throwable) {
+      promise.reject("sv_add_failed", t.message, t)
+    }
+  }
+
+  @ReactMethod
+  fun setSpeakers(speakers: ReadableArray?, promise: Promise) {
+    try {
+      ensure().setSpeakers(toSpeakerItems(speakers))
+      promise.resolve(speakerIdsArray())
+    } catch (t: Throwable) {
+      promise.reject("sv_set_failed", t.message, t)
+    }
+  }
+
+  @ReactMethod
+  fun removeSpeaker(speakerId: String?, promise: Promise) {
+    promise.resolve(stt?.removeSpeaker(speakerId ?: "") ?: false)
+  }
+
+  @ReactMethod
+  fun getSpeakerIds(promise: Promise) {
+    promise.resolve(speakerIdsArray())
+  }
+
+  @ReactMethod
+  fun startSpeechWithSVSpeakers(locale: String?, speakers: ReadableArray?, cb: Callback) {
+    ensure().startSpeech(locale, toSpeakerItems(speakers))
+    cb.invoke(false)
   }
 
   // ===== JS API =====
@@ -83,15 +145,13 @@ class STTModule(private val rc: ReactApplicationContext)
   // ANDROID expects (locale, options, callback) like @react-native-voice
   @ReactMethod
   fun startSpeech(locale: String?, options: ReadableMap, cb: Callback) {
-    val onboardingJson = readOnboardingJson(options)
-    ensure().startSpeech(locale, onboardingJson, null)
+    startWithOptions(locale, options)
     cb.invoke(false) // parity with voice: callback(false) on success
   }
 
   @ReactMethod
   fun initWithoutModel(locale: String?, options: ReadableMap?, cb: Callback) {
-    val onboardingJson = readOnboardingJson(options)
-    ensure().startSpeech(locale, onboardingJson, null)
+    startWithOptions(locale, options)
     cb.invoke(false)
   }
 
@@ -222,16 +282,24 @@ class STTModule(private val rc: ReactApplicationContext)
     send("onSpeechError", m)
   }
 
-  override fun onSpeechResults(best: String, all: List<String>) {
-    val arr = Arguments.createArray().apply { all.forEach { pushString(it) } }
-    val m = Arguments.createMap().apply { putArray("value", arr) }
-    send("onSpeechResults", m)
+  // speakerIds/speakerId are added only when speaker verification is on.
+  private fun resultMap(values: List<String>, speakerIds: List<String>, speakerId: String?): WritableMap {
+    val arr = Arguments.createArray().apply { values.forEach { pushString(it) } }
+    return Arguments.createMap().apply {
+      putArray("value", arr)
+      if (speakerIds.isNotEmpty()) {
+        putArray("speakerIds", Arguments.createArray().apply { speakerIds.forEach { pushString(it) } })
+        if (speakerId != null) putString("speakerId", speakerId)
+      }
+    }
+  }
+
+  override fun onSpeechResults(best: String, all: List<String>, speakerIds: List<String>, speakerId: String?) {
+    send("onSpeechResults", resultMap(all, speakerIds, speakerId))
   }
 
-  override fun onSpeechPartialResults(list: List<String>) {
-    val arr = Arguments.createArray().apply { list.forEach { pushString(it) } }
-    val m = Arguments.createMap().apply { putArray("value", arr) }
-    send("onSpeechPartialResults", m)
+  override fun onSpeechPartialResults(list: List<String>, speakerIds: List<String>, speakerId: String?) {
+    send("onSpeechPartialResults", resultMap(list, speakerIds, speakerId))
   }
 
   override fun onSpeechVolumeChanged(value: Float) {

package/ios/SpeechBridge/SpeechBridge.m

@@ -603,6 +603,106 @@ RCT_EXPORT_METHOD(clearAudioRoutingConfig:(RCTPromiseResolveBlock)resolve
 }
 
 /// initAll({ locale: "en-US", model: "/path/model.onnx", timeoutMs?: 8000 })
+#pragma mark - Speaker verification (live multi-speaker)
+
+// JS arrays of {id, json} -> NSArray<NSDictionary<NSString*,NSString*>*> with
+// string values only (json may be the JSON itself or a path to it).
+static NSArray<NSDictionary<NSString *, NSString *> *> *SBSpeakerItems(id raw) {
+  NSMutableArray<NSDictionary<NSString *, NSString *> *> *out = [NSMutableArray array];
+  if (![raw isKindOfClass:[NSArray class]]) return out;
+  for (id item in (NSArray *)raw) {
+    if (![item isKindOfClass:[NSDictionary class]]) continue;
+    NSMutableDictionary<NSString *, NSString *> *d = [NSMutableDictionary dictionary];
+    [(NSDictionary *)item enumerateKeysAndObjectsUsingBlock:^(id k, id v, BOOL *stop) {
+      if ([k isKindOfClass:[NSString class]] && [v isKindOfClass:[NSString class]]) d[k] = v;
+    }];
+    [out addObject:d];
+  }
+  return out;
+}
+
+static BOOL SBNonEmptyString(id v) {
+  return [v isKindOfClass:[NSString class]] && [(NSString *)v length] > 0;
+}
+
+// Starts STT from initAll-style options: `speakers` (array) replaces the set;
+// otherwise `onboardingJsonPath` (+ optional `speakerId`, default "default") adds
+// one speaker; otherwise the current set (if any) is kept.
+- (void)startSTTWithLocale:(NSString *)locale opts:(NSDictionary *)opts
+{
+  NSArray *speakers = SBSpeakerItems(opts[@"speakers"]);
+  id onboardingJsonPath = opts[@"onboardingJsonPath"];
+  id speakerId = opts[@"speakerId"];
+  if (speakers.count > 0) {
+    [self.stt startSpeechWithLocaleStr:locale speakers:speakers];
+  } else if (SBNonEmptyString(onboardingJsonPath)) {
+    [self.stt startSpeechWithLocaleStr:locale
+                    onboardingJsonPath:onboardingJsonPath
+                             speakerId:(SBNonEmptyString(speakerId) ? speakerId : @"default")];
+  } else {
+    [self.stt startSpeechWithLocaleStr:locale];
+  }
+}
+
+RCT_EXPORT_METHOD(addSpeaker:(NSString *)speakerId
+                  jsonOrPath:(NSString *)jsonOrPath
+                  resolver:(RCTPromiseResolveBlock)resolve
+                  rejecter:(RCTPromiseRejectBlock)reject)
+{
+  [self ensureSTT];
+  NSError *error = nil;
+  if (![self.stt addSpeakerWithId:(speakerId ?: @"default") jsonOrPath:(jsonOrPath ?: @"") error:&error]) {
+    reject(@"sv_add_failed", error.localizedDescription ?: @"addSpeaker failed", error);
+    return;
+  }
+  resolve([self.stt getSpeakerIds]);
+}
+
+RCT_EXPORT_METHOD(setSpeakers:(NSArray *)speakers
+                  resolver:(RCTPromiseResolveBlock)resolve
+                  rejecter:(RCTPromiseRejectBlock)reject)
+{
+  [self ensureSTT];
+  NSError *error = nil;
+  if (![self.stt setSpeakers:SBSpeakerItems(speakers) error:&error]) {
+    reject(@"sv_set_failed", error.localizedDescription ?: @"setSpeakers failed", error);
+    return;
+  }
+  resolve([self.stt getSpeakerIds]);
+}
+
+RCT_EXPORT_METHOD(removeSpeaker:(NSString *)speakerId
+                  resolver:(RCTPromiseResolveBlock)resolve
+                  rejecter:(RCTPromiseRejectBlock)reject)
+{
+  if (!self.stt) { resolve(@(NO)); return; }
+  resolve(@([self.stt removeSpeakerWithId:(speakerId ?: @"")]));
+}
+
+RCT_EXPORT_METHOD(getSpeakerIds:(RCTPromiseResolveBlock)resolve
+                  rejecter:(RCTPromiseRejectBlock)reject)
+{
+  resolve(self.stt ? [self.stt getSpeakerIds] : @[]);
+}
+
+RCT_EXPORT_METHOD(startSpeechWithSVSpeakers:(NSString *)locale
+                  speakers:(NSArray *)speakers
+                  timeoutMs:(nonnull NSNumber *)timeoutMs
+                  resolver:(RCTPromiseResolveBlock)resolve
+                  rejecter:(RCTPromiseRejectBlock)reject)
+{
+  [self ensureSTT];
+  self.sttEngineHot = NO;
+  [self.stt startSpeechWithLocaleStr:locale speakers:SBSpeakerItems(speakers)];
+  [self waitForSTTEngineHotWithTimeoutMs:timeoutMs completion:^(BOOL ok) {
+    if (!ok) {
+      reject(@"stt_start_timeout", @"STT did not become ready before timeout", nil);
+      return;
+    }
+    resolve(@{@"ok": @(YES)});
+  }];
+}
+
 RCT_EXPORT_METHOD(initAll:(NSDictionary *)opts
                   resolver:(RCTPromiseResolveBlock)resolve
                   rejecter:(RCTPromiseRejectBlock)reject)
@@ -615,7 +715,6 @@ RCT_EXPORT_METHOD(initAll:(NSDictionary *)opts
     self.sttEngineHot = NO;
 
     NSString *locale = opts[@"locale"] ?: @"en-US";
-    NSString *onboardingJsonPath = opts[@"onboardingJsonPath"];
     NSString *modelPath = opts[@"model"];
     NSNumber *timeoutMs = opts[@"timeoutMs"];
     if (modelPath.length == 0) {
@@ -632,11 +731,7 @@ RCT_EXPORT_METHOD(initAll:(NSDictionary *)opts
         [self.stt setLicenseWithLicenseKey:self.lastLicenseKey];
       }
     }
-    if (onboardingJsonPath && (id)onboardingJsonPath != [NSNull null] && onboardingJsonPath.length > 0) {
-      [self.stt startSpeechWithLocaleStr:locale onboardingJsonPath:onboardingJsonPath];
-    } else {
-      [self.stt startSpeechWithLocaleStr:locale];
-    }
+    [self startSTTWithLocale:locale opts:opts];
 
     // 2) TTS next
     NSURL *modelURL = nil;
@@ -740,17 +835,12 @@ RCT_EXPORT_METHOD(initWithoutModel:(NSDictionary *)opts
     self.sttEngineHot = NO;
 
     NSString *locale = opts[@"locale"] ?: @"en-US";
-    NSString *onboardingJsonPath = opts[@"onboardingJsonPath"];
     NSNumber *timeoutMs = opts[@"timeoutMs"];
 
     [self tearDownSTT];
 
     [self ensureSTT];
-    if (onboardingJsonPath && (id)onboardingJsonPath != [NSNull null] && onboardingJsonPath.length > 0) {
-      [self.stt startSpeechWithLocaleStr:locale onboardingJsonPath:onboardingJsonPath];
-    } else {
-      [self.stt startSpeechWithLocaleStr:locale];
-    }
+    [self startSTTWithLocale:locale opts:opts];
 
     [self waitForSTTEngineHotWithTimeoutMs:timeoutMs completion:^(BOOL ok) {
       if (!ok) {

package/ios/TTSRNBridge/DavoiceTTS.xcframework/Info.plist

@@ -8,32 +8,32 @@
 			<key>BinaryPath</key>
 			<string>DavoiceTTS.framework/DavoiceTTS</string>
 			<key>LibraryIdentifier</key>
-			<string>ios-arm64</string>
+			<string>ios-arm64_x86_64-simulator</string>
 			<key>LibraryPath</key>
 			<string>DavoiceTTS.framework</string>
 			<key>SupportedArchitectures</key>
 			<array>
 				<string>arm64</string>
+				<string>x86_64</string>
 			</array>
 			<key>SupportedPlatform</key>
 			<string>ios</string>
+			<key>SupportedPlatformVariant</key>
+			<string>simulator</string>
 		</dict>
 		<dict>
 			<key>BinaryPath</key>
 			<string>DavoiceTTS.framework/DavoiceTTS</string>
 			<key>LibraryIdentifier</key>
-			<string>ios-arm64_x86_64-simulator</string>
+			<string>ios-arm64</string>
 			<key>LibraryPath</key>
 			<string>DavoiceTTS.framework</string>
 			<key>SupportedArchitectures</key>
 			<array>
 				<string>arm64</string>
-				<string>x86_64</string>
 			</array>
 			<key>SupportedPlatform</key>
 			<string>ios</string>
-			<key>SupportedPlatformVariant</key>
-			<string>simulator</string>
 		</dict>
 	</array>
 	<key>CFBundlePackageType</key>

package/ios/TTSRNBridge/DavoiceTTS.xcframework/ios-arm64/DavoiceTTS.framework/Headers/DavoiceTTS-Swift.h

@@ -416,7 +416,27 @@ SWIFT_CLASS_PROPERTY(@property (nonatomic, class, readonly, copy) NSArray<NSStri
 - (void)isSpeechAvailable:(void (^ _Nonnull)(BOOL))completion;
 - (BOOL)isRecognizing SWIFT_WARN_UNUSED_RESULT;
 - (void)startSpeechWithLocaleStr:(NSString * _Nullable)localeStr;
+/// Starts STT with one speaker. <code>onboardingJsonPath</code> may be a path or the JSON
+/// itself. The speaker is added to the current set (same id or same voiceprint
+/// replaces the older entry) under the id “default”.
 - (void)startSpeechWithLocaleStr:(NSString * _Nullable)localeStr onboardingJsonPath:(NSString * _Nonnull)onboardingJsonPath;
+- (void)startSpeechWithLocaleStr:(NSString * _Nullable)localeStr onboardingJsonPath:(NSString * _Nonnull)onboardingJsonPath speakerId:(NSString * _Nonnull)speakerId;
+/// Starts STT with an array of speakers, replacing the current set.
+/// Each item: [“id”: speakerId, “json”: JSON string or path to the JSON file].
+- (void)startSpeechWithLocaleStr:(NSString * _Nullable)localeStr speakers:(NSArray<NSDictionary<NSString *, NSString *> *> * _Nonnull)speakers;
+/// Adds one speaker (JSON string or path). Same id or same voiceprint replaces
+/// the older entry. Applies immediately when STT is running, otherwise at the
+/// next start. Throws if the JSON can’t be read or its audio-feature settings
+/// differ from the running engine’s.
+- (BOOL)addSpeakerWithId:(NSString * _Nonnull)id jsonOrPath:(NSString * _Nonnull)jsonOrPath error:(NSError * _Nullable * _Nullable)error;
+/// Replaces the whole speaker set. Items: [“id”: …, “json”: JSON string or path].
+/// An empty array turns speaker verification off.
+- (BOOL)setSpeakers:(NSArray<NSDictionary<NSString *, NSString *> *> * _Nonnull)speakers error:(NSError * _Nullable * _Nullable)error;
+/// Removes a speaker by id. Removing the last one turns the gate off (all audio
+/// passes). Returns false if the id wasn’t enrolled.
+- (BOOL)removeSpeakerWithId:(NSString * _Nonnull)id;
+/// Ids of the enrolled speakers, in insertion order.
+- (NSArray<NSString *> * _Nonnull)getSpeakerIds SWIFT_WARN_UNUSED_RESULT;
 - (void)stopSpeech:(void (^ _Nullable)(BOOL))completion;
 - (void)cancelSpeech:(void (^ _Nullable)(BOOL))completion;
 - (void)destroySpeech:(void (^ _Nullable)(BOOL))completion;