Update main.py

main.py

@@ -1,6 +1,6 @@
 # main.py
 from fastapi import FastAPI, File, UploadFile, HTTPException
-from fastapi.responses import StreamingResponse
+from fastapi.responses import FileResponse
 from fastapi.middleware.cors import CORSMiddleware
 import google.generativeai as genai
 import pdfplumber
@@ -11,8 +11,6 @@ import tempfile
 import shutil
 from gtts import gTTS
 from pydub import AudioSegment
-import asyncio
-import io
 
 app = FastAPI()
 
@@ -112,8 +110,8 @@ def generate_conversation(pdf_text):
         print(f"Problem text: {cleaned_text}")
         raise ValueError(f"Failed to parse generated conversation: {str(e)}")
 
-def create_audio_from_conversation(conversation, temp_dir):
-    """Create audio file from conversation with progress tracking"""
+def create_audio_from_conversation(conversation, output_path):
+    """Create audio file from conversation"""
     # Female voice
     def generate_female_voice(text, filename):
         tts = gTTS(text=text, lang='en')
@@ -141,10 +139,12 @@ def create_audio_from_conversation(conversation, temp_dir):
         "Bob": "male"
     }
     
+    # Create temp directory
+    temp_dir = tempfile.mkdtemp()
+    
     # Combine lines
     final_podcast = AudioSegment.silent(duration=1000)  # 1 sec silence at start
     
-    total_lines = len(conversation)
     for i, line_dict in enumerate(conversation):
         for speaker, line in line_dict.items():
             voice_type = speaker_voice_map.get(speaker, "female")
@@ -156,125 +156,46 @@ def create_audio_from_conversation(conversation, temp_dir):
                 voice = generate_male_voice(line, filename)
             
             final_podcast += voice + AudioSegment.silent(duration=500)
-        
-        # Yield progress update
-        progress = int(100 * (i+1) / total_lines)
-        yield json.dumps({
-            "status": "processing",
-            "step": "generating_audio",
-            "progress": progress,
-            "message": f"Processing dialogue {i+1}/{total_lines}"
-        }) + "\n"
     
     # Export final audio
-    output_path = f"{temp_dir}/final_podcast.mp3"
     final_podcast.export(output_path, format="mp3")
     
-    # Return the path to the final audio file
-    return output_path
+    # Clean up temp files
+    shutil.rmtree(temp_dir)
 
-async def process_pdf_to_audio_stream(file_path):
-    """Process PDF to audio with streaming progress updates"""
+@app.post("/convert/")
+async def convert_pdf_to_audio(file: UploadFile = File(...)):
+    """Convert PDF to audio"""
     try:
-        # Stream initial status
-        yield json.dumps({
-            "status": "processing",
-            "step": "extracting_text",
-            "progress": 10, 
-            "message": "Extracting text from PDF..."
-        }) + "\n"
+        # Create temporary file for PDF
+        temp_pdf = tempfile.NamedTemporaryFile(delete=False, suffix=".pdf")
+        temp_pdf_path = temp_pdf.name
+        
+        # Save uploaded PDF
+        with open(temp_pdf_path, "wb") as pdf_file:
+            shutil.copyfileobj(file.file, pdf_file)
         
         # Extract text from PDF
-        pdf_text = extract_text_from_pdf(file_path)
+        pdf_text = extract_text_from_pdf(temp_pdf_path)
         if not pdf_text.strip():
-            yield json.dumps({
-                "status": "error",
-                "message": "No text extracted from PDF"
-            }) + "\n"
-            return
-        
-        # Stream progress update
-        yield json.dumps({
-            "status": "processing",
-            "step": "generating_conversation",
-            "progress": 30, 
-            "message": "Generating conversation from PDF content..."
-        }) + "\n"
+            os.unlink(temp_pdf_path)
+            raise HTTPException(status_code=400, detail="No text extracted from PDF")
         
         # Generate conversation
         conversation = generate_conversation(pdf_text)
         
-        # Stream progress update
-        yield json.dumps({
-            "status": "processing",
-            "step": "starting_audio_generation",
-            "progress": 50, 
-            "message": "Starting audio generation..."
-        }) + "\n"
-        
-        # Create temp directory for audio files
-        temp_dir = tempfile.mkdtemp()
+        # Create audio file
+        output_filename = f"temp/output_{file.filename.split('.')[0]}.mp3"
+        create_audio_from_conversation(conversation, output_filename)
         
-        # Generate audio for each line with streaming updates
-        audio_file_path = None
-        async for update in async_generator_wrapper(create_audio_from_conversation(conversation, temp_dir)):
-            yield update
-            # The last non-json output will be the file path
-            if not update.startswith("{"):
-                audio_file_path = update.strip()
-        
-        if not audio_file_path or not os.path.exists(audio_file_path):
-            yield json.dumps({
-                "status": "error",
-                "message": "Failed to generate audio file"
-            }) + "\n"
-            return
-            
-        # Read the audio file
-        with open(audio_file_path, "rb") as audio_file:
-            audio_data = audio_file.read()
-            
-        # Stream completion status with the audio data as base64
-        import base64
-        audio_base64 = base64.b64encode(audio_data).decode('utf-8')
-        yield json.dumps({
-            "status": "complete",
-            "progress": 100,
-            "message": "Audio generation complete",
-            "audio_data": audio_base64
-        }) + "\n"
-        
-        # Clean up
-        shutil.rmtree(temp_dir)
-        
-    except Exception as e:
-        yield json.dumps({
-            "status": "error",
-            "message": f"Error: {str(e)}"
-        }) + "\n"
-
-async def async_generator_wrapper(sync_generator):
-    """Convert a synchronous generator to an async generator"""
-    for item in sync_generator:
-        await asyncio.sleep(0.01)  # Small sleep to allow other tasks to run
-        yield item
-
-@app.post("/convert-stream/")
-async def convert_pdf_to_audio_stream(file: UploadFile = File(...)):
-    """Convert PDF to audio with streaming progress updates"""
-    try:
-        # Create temporary file for PDF
-        temp_pdf = tempfile.NamedTemporaryFile(delete=False, suffix=".pdf")
-        temp_pdf_path = temp_pdf.name
-        
-        # Save uploaded PDF
-        with open(temp_pdf_path, "wb") as pdf_file:
-            shutil.copyfileobj(file.file, pdf_file)
+        # Clean up PDF file
+        os.unlink(temp_pdf_path)
         
-        # Return streaming response
-        return StreamingResponse(
-            process_pdf_to_audio_stream(temp_pdf_path),
-            media_type="text/event-stream"
+        # Return audio file
+        return FileResponse(
+            path=output_filename,
+            media_type="audio/mpeg",
+            filename=f"audio_{file.filename.split('.')[0]}.mp3"
         )
         
     except Exception as e: