From 7fe118ce639aeac6bd84c7c2e721f3cf6dd298e0 Mon Sep 17 00:00:00 2001
From: Mark Backman <mark@daily.co>
Date: Fri, 27 Sep 2024 11:22:03 -0400
Subject: [PATCH] Align use of language param across TTS services

---
 src/pipecat/services/azure.py      | 17 ++++++++++-------
 src/pipecat/services/elevenlabs.py | 10 +++++-----
 2 files changed, 15 insertions(+), 12 deletions(-)

diff --git a/src/pipecat/services/azure.py b/src/pipecat/services/azure.py
index c8fa095ab..a1349cefe 100644
--- a/src/pipecat/services/azure.py
+++ b/src/pipecat/services/azure.py
@@ -41,7 +41,10 @@ try:
         SpeechRecognizer,
         SpeechSynthesizer,
     )
-    from azure.cognitiveservices.speech.audio import AudioStreamFormat, PushAudioInputStream
+    from azure.cognitiveservices.speech.audio import (
+        AudioStreamFormat,
+        PushAudioInputStream,
+    )
     from azure.cognitiveservices.speech.dialog import AudioConfig
     from openai import AsyncAzureOpenAI
 except ModuleNotFoundError as e:
@@ -73,7 +76,7 @@ class AzureLLMService(BaseOpenAILLMService):
 class AzureTTSService(TTSService):
     class InputParams(BaseModel):
         emphasis: Optional[str] = None
-        language_code: Optional[str] = "en-US"
+        language: Optional[str] = "en-US"
         pitch: Optional[str] = None
         rate: Optional[str] = "1.05"
         role: Optional[str] = None
@@ -105,7 +108,7 @@ class AzureTTSService(TTSService):
 
     def _construct_ssml(self, text: str) -> str:
         ssml = (
-            f"<speak version='1.0' xml:lang='{self._params.language_code}' "
+            f"<speak version='1.0' xml:lang='{self._params.language}' "
             "xmlns='http://www.w3.org/2001/10/synthesis' "
             "xmlns:mstts='http://www.w3.org/2001/mstts'>"
             f"<voice name='{self._voice}'>"
@@ -155,9 +158,9 @@ class AzureTTSService(TTSService):
         logger.debug(f"Setting TTS emphasis to: [{emphasis}]")
         self._params.emphasis = emphasis
 
-    async def set_language_code(self, language_code: str):
-        logger.debug(f"Setting TTS language code to: [{language_code}]")
-        self._params.language_code = language_code
+    async def set_language(self, language: str):
+        logger.debug(f"Setting TTS language code to: [{language}]")
+        self._params.language = language
 
     async def set_pitch(self, pitch: str):
         logger.debug(f"Setting TTS pitch to: [{pitch}]")
@@ -187,7 +190,7 @@ class AzureTTSService(TTSService):
         valid_params = {
             "voice": self.set_voice,
             "emphasis": self.set_emphasis,
-            "language_code": self.set_language_code,
+            "language_code": self.set_language,
             "pitch": self.set_pitch,
             "rate": self.set_rate,
             "role": self.set_role,
diff --git a/src/pipecat/services/elevenlabs.py b/src/pipecat/services/elevenlabs.py
index 79d90bc58..ca4713f5f 100644
--- a/src/pipecat/services/elevenlabs.py
+++ b/src/pipecat/services/elevenlabs.py
@@ -72,7 +72,7 @@ def calculate_word_times(
 
 class ElevenLabsTTSService(AsyncWordTTSService):
     class InputParams(BaseModel):
-        language_code: Optional[str] = None
+        language: Optional[str] = None
         output_format: Literal["pcm_16000", "pcm_22050", "pcm_24000", "pcm_44100"] = "pcm_16000"
         optimize_streaming_latency: Optional[str] = None
         stability: Optional[float] = None
@@ -229,13 +229,13 @@ class ElevenLabsTTSService(AsyncWordTTSService):
             if self._params.optimize_streaming_latency:
                 url += f"&optimize_streaming_latency={self._params.optimize_streaming_latency}"
 
-            # language_code can only be used with the 'eleven_turbo_v2_5' model
-            if self._params.language_code:
+            # language can only be used with the 'eleven_turbo_v2_5' model
+            if self._params.language:
                 if model == "eleven_turbo_v2_5":
-                    url += f"&language_code={self._params.language_code}"
+                    url += f"&language_code={self._params.language}"
                 else:
                     logger.debug(
-                        f"Language code [{self._params.language_code}] not applied. Language codes can only be used with the 'eleven_turbo_v2_5' model."
+                        f"Language code [{self._params.language}] not applied. Language codes can only be used with the 'eleven_turbo_v2_5' model."
                     )
 
             self._websocket = await websockets.connect(url)