Upload 3 files

app.py

@@ -1,71 +1,64 @@
 import os
 import uuid
 import json
-import logging
 import requests
 import streamlit as st
 from dotenv import load_dotenv
 from utils import voice_map, get_voice_prompt_style, AUDIO_DIR
 from generate_audio import generate_audio
+from logger_setup import logger
 
-# Load API keys from .env
+# Load API keys
 load_dotenv()
 OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
-ELEVENLABS_API_KEY = os.getenv("ELEVENLABS_API_KEY")
 
-# Ensure audio output folder exists
-os.makedirs(AUDIO_DIR, exist_ok=True)
-logging.basicConfig(filename="frontend.log", level=logging.INFO)
-
-# Streamlit UI config
+# Streamlit config
 st.set_page_config(page_title="Voice Agent Pro", page_icon="🎤")
 st.title("🎙️ Voice Agent Pro")
 st.markdown("Summarized answers with expressive AI voices.")
+logger.info("🎬 Streamlit app started")
 
-# Sidebar: select voice
+# Sidebar: voice picker
 st.sidebar.header("🎚️ Voice Settings")
 voice_label = st.sidebar.selectbox("Choose a voice:", list(voice_map.keys()))
 voice_id = voice_map[voice_label]
 tone_prompt = get_voice_prompt_style(voice_label)
 
-# App state
+# Session state setup
 if "answer" not in st.session_state: st.session_state.answer = ""
 if "audio_key" not in st.session_state: st.session_state.audio_key = None
 if "file_text" not in st.session_state: st.session_state.file_text = ""
 if "key_points" not in st.session_state: st.session_state.key_points = []
 
-# Input fields
+# Inputs
 query = st.text_area("🗨️ Ask or refine something based on the bullets:", value="", placeholder="e.g., What makes you so cool, Grandma?", key="query")
 url = st.text_input("🌐 Optional URL to summarize:")
 uploaded_file = st.file_uploader("📎 Or upload a file (PDF, TXT, DOCX)", type=["pdf", "txt", "docx"])
 
-# Reset app state
+# Reset state
+# Reset state safely without error
 if st.button("🧹 Clear All"):
-    st.session_state.query = ""
-    st.session_state.file_text = ""
-    st.session_state.answer = ""
-    st.session_state.audio_key = None
-    st.session_state.key_points = []
+    logger.info("🧼 Clear All clicked — reloading app")
+    st.rerun()
+
 
-# Helper: OpenAI response streaming
+# Helper: GPT streaming
 def stream_openai_response(payload, headers):
     with requests.post("https://api.openai.com/v1/chat/completions", headers=headers, json=payload, stream=True) as r:
         for line in r.iter_lines():
             if line and line.startswith(b"data: "):
                 yield line[len(b"data: "):].decode()
 
-# Main button: summarize and speak
+# Main: Summarize and speak
 if st.button("🔁 Summarize"):
     if not query and not url and not uploaded_file:
         st.warning("Please enter a question, a URL, or upload a file.")
     else:
         with st.spinner("Talking to GPT..."):
             try:
-                # Load file content if present
                 if uploaded_file:
                     st.session_state.file_text = uploaded_file.read().decode("utf-8")
 
-                # Build GPT prompt
                 context = ""
                 if st.session_state.file_text:
                     context += st.session_state.file_text + "\n\n"
@@ -73,7 +66,6 @@ if st.button("🔁 Summarize"):
                     context += f"Summarize this page: {url}\n\n"
                 context += f"{tone_prompt}\n\nNow answer: {query}"
 
-                # OpenAI request setup
                 headers = {"Authorization": f"Bearer {OPENAI_API_KEY}"}
                 payload = {
                     "model": "gpt-4o",
@@ -82,28 +74,42 @@ if st.button("🔁 Summarize"):
                     "stream": True
                 }
 
-                # Streaming UI update
                 st.session_state.answer = ""
                 answer_box = st.empty()
+                logger.info("🧠 Starting GPT streaming")
 
                 for chunk in stream_openai_response(payload, headers):
-                    parsed = json.loads(chunk)
-                    delta = parsed['choices'][0]['delta'].get('content', '')
-                    st.session_state.answer += delta
-                    answer_box.markdown(st.session_state.answer)
+                    if chunk.strip() == "[DONE]":
+                        logger.info("🟢 GPT stream complete marker received")
+                        continue
+
+                    try:
+                        parsed = json.loads(chunk)
+                        delta = parsed['choices'][0]['delta'].get('content', '')
+                        st.session_state.answer += delta
+                        answer_box.markdown(st.session_state.answer)
+                    except json.JSONDecodeError as json_err:
+                        logger.warning(f"⚠️ Skipping non-JSON chunk: {chunk}")
+                        continue
+
+
+                logger.info("🧠 GPT response complete. Now generating audio...")
 
-                # Generate audio from final answer
                 audio_key = str(uuid.uuid4())
                 generate_audio(st.session_state.answer, voice_id, audio_key)
                 st.session_state.audio_key = audio_key
 
             except Exception as e:
                 st.error(f"🔥 Error: {e}")
+                logger.exception("🔥 Exception during summarize or audio generation")
 
-# Final UI display
+# Final display
 if st.session_state.answer:
     st.subheader("📜 Answer")
     st.success(st.session_state.answer)
     if st.session_state.audio_key:
         audio_path = os.path.join(AUDIO_DIR, f"{st.session_state.audio_key}.mp3")
-        st.audio(audio_path)
+        if os.path.exists(audio_path):
+            st.audio(audio_path)
+        else:
+            st.error("❗ Audio file missing. Please check logs.")

generate_audio.py

@@ -1,35 +1,50 @@
 import os
-import logging
-from elevenlabs import stream
-from elevenlabs.client import ElevenLabs
 from dotenv import load_dotenv
+from elevenlabs.client import ElevenLabs
+from logger_setup import logger
+
+# Load environment variables
 load_dotenv()
 
+# Use absolute path for output
+AUDIO_DIR = os.path.join(os.path.dirname(__file__), "audio_outputs")
 
-AUDIO_DIR = "audio_outputs"
-logger = logging.getLogger(__name__)
+# Verify API key
+api_key = os.getenv("ELEVENLABS_API_KEY")
+if not api_key:
+    logger.error("❌ ELEVENLABS_API_KEY is missing or not loaded from .env")
+    raise RuntimeError("ELEVENLABS_API_KEY missing")
 
-client = ElevenLabs(api_key=os.getenv("ELEVENLABS_API_KEY"))
+client = ElevenLabs(api_key=api_key)
 
 def generate_audio(text: str, voice_id: str, audio_key: str):
     try:
         logger.info("🎯 Starting ElevenLabs audio generation")
+        os.makedirs(AUDIO_DIR, exist_ok=True)
 
-        audio_stream = client.text_to_speech.convert_as_stream(
-            text=text,
-            voice_id=voice_id,
-            model_id="eleven_multilingual_v2"
-        )
+        try:
+            audio_stream = client.text_to_speech.convert_as_stream(
+                text=text,
+                voice_id=voice_id,
+                model_id="eleven_multilingual_v2"
+            )
+            logger.info("✅ Audio stream received from ElevenLabs")
+        except Exception as stream_err:
+            logger.error(f"❌ Failed to get audio stream: {stream_err}")
+            raise
 
-        os.makedirs(AUDIO_DIR, exist_ok=True)
         output_path = os.path.join(AUDIO_DIR, f"{audio_key}.mp3")
 
-        with open(output_path, "wb") as f:
-            for chunk in audio_stream:
-                if isinstance(chunk, bytes):
-                    f.write(chunk)
-
-        logger.info(f"✅ Audio saved to {output_path}")
+        try:
+            with open(output_path, "wb") as f:
+                for chunk in audio_stream:
+                    if isinstance(chunk, bytes):
+                        f.write(chunk)
+            logger.info(f"✅ Audio saved to {output_path}")
+        except Exception as write_err:
+            logger.error(f"❌ Failed to save audio to file: {write_err}")
+            raise
 
     except Exception as e:
-        logger.error(f"🔥 Error generating audio: {e}")
+        logger.exception("🔥 Exception in generate_audio")
+        raise

logger_setup.py

@@ -0,0 +1,16 @@
+# logger_setup.py
+import logging
+import os
+
+LOG_FILE = os.path.join(os.path.dirname(__file__), "logfile.log")
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s - %(message)s",
+    handlers=[
+        logging.FileHandler(LOG_FILE, mode='a', encoding='utf-8'),
+        logging.StreamHandler()
+    ]
+)
+
+logger = logging.getLogger("voice-agent")