Update utils.py

utils.py

@@ -8,12 +8,10 @@ import re
 import tempfile
 import requests
 from bs4 import BeautifulSoup
-import subprocess
-import pyttsx3
-from pydub import AudioSegment
 
 groq_client = Groq(api_key=os.environ["GROQ_API_KEY"])
 tokenizer = tiktoken.get_encoding("cl100k_base")
+VOICERSS_API_KEY = os.environ["VOICERSS_API_KEY"]
 
 class DialogueItem(BaseModel):
     speaker: Literal["Maria", "Sarah"]
@@ -96,23 +94,14 @@ def generate_script(system_prompt: str, input_text: str, tone: str, target_lengt
     
     return dialogue
 
-def generate_audio_espeak(text: str, speaker: str) -> str:
-    voice = "en-us+f3" if speaker == "Maria" else "en-gb+f3"
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
-        subprocess.call(['espeak-ng', '-v', voice, '-w', temp_audio.name, text])
-        return temp_audio.name
-
-def generate_audio_pyttsx3(text: str, speaker: str) -> str:
-    engine = pyttsx3.init()
-    voices = engine.getProperty('voices')
-    engine.setProperty('voice', voices[1].id if speaker == "Maria" else voices[0].id)
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
-        engine.save_to_file(text, temp_audio.name)
-        engine.runAndWait()
-        return temp_audio.name
-
 def generate_audio(text: str, speaker: str) -> str:
-    try:
-        return generate_audio_espeak(text, speaker)
-    except Exception:
-        return generate_audio_pyttsx3(text, speaker)
+    voice = "en-us" if speaker == "Maria" else "en-gb"
+    url = f"http://api.voicerss.org/?key={VOICERSS_API_KEY}&hl={voice}&src={text}"
+    
+    response = requests.get(url)
+    if response.status_code != 200:
+        raise Exception(f"Error generating audio: {response.text}")
+    
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as temp_audio:
+        temp_audio.write(response.content)
+        return temp_audio.name