Create tts_google.py

tts_google.py

@@ -0,0 +1,64 @@
+# tts_google.py
+from google.cloud import texttospeech
+from ssml_converter import SSMLConverter
+
+class GoogleCloudTTS(TTSInterface):
+    """Google Cloud Text-to-Speech implementation"""
+    
+    def __init__(self, credentials_path: str):
+        super().__init__()
+        self.supports_ssml = True
+        self.credentials_path = credentials_path
+        
+        # Google TTS doesn't need preprocessing with SSML
+        self.preprocessing_flags = set()
+        
+        # Initialize client
+        os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = credentials_path
+        self.client = texttospeech.TextToSpeechClient()
+        
+        # SSML converter
+        self.ssml_converter = SSMLConverter(language="tr-TR")
+        
+    async def synthesize(self, text: str, voice_id: Optional[str] = None, **kwargs) -> bytes:
+        """Convert text to speech using Google Cloud TTS"""
+        try:
+            # Check if SSML should be used
+            use_ssml = kwargs.get("use_ssml", True)
+            
+            if use_ssml and not text.startswith("<speak>"):
+                # Convert to SSML
+                text = self.ssml_converter.convert_to_ssml(text)
+                log(f"📝 Converted to SSML: {text[:200]}...")
+                input_text = texttospeech.SynthesisInput(ssml=text)
+            else:
+                input_text = texttospeech.SynthesisInput(text=text)
+            
+            # Voice selection
+            voice = texttospeech.VoiceSelectionParams(
+                language_code=kwargs.get("language_code", "tr-TR"),
+                name=voice_id or "tr-TR-Wavenet-B",
+                ssml_gender=texttospeech.SsmlVoiceGender.FEMALE
+            )
+            
+            # Audio config
+            audio_config = texttospeech.AudioConfig(
+                audio_encoding=texttospeech.AudioEncoding.MP3,
+                speaking_rate=kwargs.get("speaking_rate", 1.0),
+                pitch=kwargs.get("pitch", 0.0),
+                volume_gain_db=kwargs.get("volume_gain_db", 0.0)
+            )
+            
+            # Perform synthesis
+            response = self.client.synthesize_speech(
+                input=input_text,
+                voice=voice,
+                audio_config=audio_config
+            )
+            
+            log(f"✅ Google TTS returned {len(response.audio_content)} bytes")
+            return response.audio_content
+            
+        except Exception as e:
+            log(f"❌ Google TTS error: {e}")
+            raise