From e24ef426174e1759b4a513d620cf94b649cc4aa8 Mon Sep 17 00:00:00 2001 From: patrick-motsch Date: Sun, 15 Feb 2026 10:08:50 +0100 Subject: [PATCH] feat(teamsbot): AI voice test, config save, camelCase mapping, default voices - Voice test endpoint generates sample text dynamically via AI in selected language - Fixed config save: added "config" to allowed update fields in interfaceFeatures - Clean camelCase mapping in interfaceVoiceObjects (audio_content -> audioContent) - Default TTS voices for common languages in connectorVoiceGoogle - Fixed updateFeatureInstanceConfig -> updateFeatureInstance with config field Co-authored-by: Cursor --- modules/connectors/connectorVoiceGoogle.py | 12 +++++-- .../features/teamsbot/routeFeatureTeamsbot.py | 34 ++++++++++++++++--- modules/interfaces/interfaceFeatures.py | 2 +- modules/interfaces/interfaceVoiceObjects.py | 11 ++++-- 4 files changed, 50 insertions(+), 9 deletions(-) diff --git a/modules/connectors/connectorVoiceGoogle.py b/modules/connectors/connectorVoiceGoogle.py index 85a19bf7..42efffcf 100644 --- a/modules/connectors/connectorVoiceGoogle.py +++ b/modules/connectors/connectorVoiceGoogle.py @@ -778,9 +778,17 @@ class ConnectorGoogleSpeech: def _getDefaultVoice(self, languageCode: str) -> str: """ Get default voice name for a language code. - Returns None - no defaults, let the frontend handle voice selection. + Falls back to a Wavenet voice for common languages. """ - return None + _defaults = { + "de-DE": "de-DE-Wavenet-A", + "de-CH": "de-DE-Wavenet-A", + "en-US": "en-US-Wavenet-C", + "en-GB": "en-GB-Wavenet-A", + "fr-FR": "fr-FR-Wavenet-A", + "it-IT": "it-IT-Wavenet-A", + } + return _defaults.get(languageCode) async def getAvailableLanguages(self) -> Dict[str, Any]: """ diff --git a/modules/features/teamsbot/routeFeatureTeamsbot.py b/modules/features/teamsbot/routeFeatureTeamsbot.py index 91cd4cec..2abd2cf7 100644 --- a/modules/features/teamsbot/routeFeatureTeamsbot.py +++ b/modules/features/teamsbot/routeFeatureTeamsbot.py @@ -391,7 +391,7 @@ async def updateConfig( # Save to FeatureInstance.config rootInterface = getRootInterface() featureInterface = getFeatureInterface(rootInterface.db) - featureInterface.updateFeatureInstanceConfig(instanceId, mergedConfig.model_dump()) + featureInterface.updateFeatureInstance(instanceId, {"config": mergedConfig.model_dump()}) logger.info(f"Teamsbot config updated for instance {instanceId}: {list(updateDict.keys())}") return {"config": mergedConfig.model_dump()} @@ -408,20 +408,45 @@ async def testVoice( instanceId: str, context: RequestContext = Depends(getRequestContext), ): - """Test TTS voice with a sample text. Returns base64-encoded audio.""" + """Test TTS voice with AI-generated sample text in the correct language.""" from modules.interfaces.interfaceVoiceObjects import getVoiceInterface + from modules.services.serviceAi.mainServiceAi import AiService + from modules.datamodels.datamodelAi import AiCallRequest, AiCallOptions, OperationTypeEnum, PriorityEnum mandateId = _validateInstanceAccess(instanceId, context) body = await request.json() - text = body.get("text", "Hallo, ich bin der AI-Assistent. Wie kann ich helfen?") language = body.get("language", "de-DE") voiceId = body.get("voiceId") + botName = body.get("botName", "AI Assistant") try: + # Generate test text dynamically via AI in the correct language + serviceContext = type('Ctx', (), { + 'user': context.user, 'mandateId': mandateId, + 'featureInstanceId': instanceId, 'featureCode': 'teamsbot' + })() + aiService = AiService(serviceCenter=serviceContext) + await aiService.ensureAiObjectsInitialized() + + aiRequest = AiCallRequest( + prompt=f"Generate a short, friendly introduction sentence (max 2 sentences) for a meeting bot named '{botName}'. " + f"Write ONLY in the language '{language}'. No quotes, no explanation, just the text to speak.", + context="", + options=AiCallOptions( + operationType=OperationTypeEnum.DATA_ANALYSE, + priority=PriorityEnum.SPEED, + ) + ) + aiResponse = await aiService.callAi(aiRequest) + testText = aiResponse.content.strip().strip('"').strip("'") if aiResponse and aiResponse.errorCount == 0 else f"Hello, I am {botName}." + + logger.info(f"Voice test: generated text in {language}: '{testText[:60]}...'") + + # Convert to speech voiceInterface = getVoiceInterface(context.user, mandateId) result = await voiceInterface.textToSpeech( - text=text, + text=testText, languageCode=language, voiceName=voiceId, ) @@ -439,6 +464,7 @@ async def testVoice( "format": "mp3", "language": language, "voiceId": voiceId, + "text": testText, } return {"success": False, "error": "TTS returned no audio"} diff --git a/modules/interfaces/interfaceFeatures.py b/modules/interfaces/interfaceFeatures.py index 4866a0d6..b261e76e 100644 --- a/modules/interfaces/interfaceFeatures.py +++ b/modules/interfaces/interfaceFeatures.py @@ -406,7 +406,7 @@ class FeatureInterface: return None # Only allow updating specific fields - allowedFields = {"label", "enabled"} + allowedFields = {"label", "enabled", "config"} filteredData = {k: v for k, v in updateData.items() if k in allowedFields} if not filteredData: diff --git a/modules/interfaces/interfaceVoiceObjects.py b/modules/interfaces/interfaceVoiceObjects.py index cccebce4..6efe51cf 100644 --- a/modules/interfaces/interfaceVoiceObjects.py +++ b/modules/interfaces/interfaceVoiceObjects.py @@ -271,10 +271,17 @@ class VoiceObjects: if result["success"]: logger.info(f"✅ Text-to-Speech successful: {len(result['audio_content'])} bytes") + # Map connector snake_case keys to camelCase for consistent API + return { + "success": True, + "audioContent": result["audio_content"], + "audioFormat": result.get("audio_format", "mp3"), + "languageCode": result.get("language_code", languageCode), + "voiceName": result.get("voice_name", voiceName), + } else: logger.warning(f"⚠️ Text-to-Speech failed: {result.get('error', 'Unknown error')}") - - return result + return result except Exception as e: logger.error(f"❌ Text-to-Speech error: {e}")