Upload app.py

app.py

@@ -1,4 +1,4 @@
-import gradio as gr
+import streamlit as st
 from pydub import AudioSegment
 from google import genai
 from google.genai import types
@@ -284,7 +284,7 @@ Follow this example structure:
 
         try:
             if progress:
-                progress(0.3, "Generating podcast script...")
+                progress.progress(0.3, "Generating podcast script...")
                 
             # Add timeout to the API call
             response = await asyncio.wait_for(
@@ -306,19 +306,19 @@ Follow this example structure:
                 timeout=60  # 60 seconds timeout
             )
         except asyncio.TimeoutError:
-            raise gr.Error("The script generation request timed out. Please try again later.")
+            raise Exception("The script generation request timed out. Please try again later.")
         except Exception as e:
             if "API key not valid" in str(e):
-                raise gr.Error("Invalid API key. Please provide a valid Gemini API key.")
+                raise Exception("Invalid API key. Please provide a valid Gemini API key.")
             elif "rate limit" in str(e).lower():
-                raise gr.Error("Rate limit exceeded for the API key. Please try again later or provide your own Gemini API key.")
+                raise Exception("Rate limit exceeded for the API key. Please try again later or provide your own Gemini API key.")
             else:
-                raise gr.Error(f"Failed to generate podcast script: {e}")
+                raise Exception(f"Failed to generate podcast script: {e}")
 
         print(f"Generated podcast script:\n{response.text}")
         
         if progress:
-            progress(0.4, "Script generated successfully!")
+            progress.progress(0.4, "Script generated successfully!")
             
         return json.loads(response.text)
     
@@ -327,7 +327,7 @@ Follow this example structure:
         # Check file size before reading
         file_size = os.path.getsize(file_obj.name)
         if file_size > MAX_FILE_SIZE_BYTES:
-            raise gr.Error(f"File size exceeds the {MAX_FILE_SIZE_MB}MB limit. Please upload a smaller file.")
+            raise Exception(f"File size exceeds the {MAX_FILE_SIZE_MB}MB limit. Please upload a smaller file.")
             
         async with aiofiles.open(file_obj.name, 'rb') as f:
             return await f.read()
@@ -356,7 +356,7 @@ Follow this example structure:
         except asyncio.TimeoutError:
             if os.path.exists(temp_filename):
                 os.remove(temp_filename)
-            raise gr.Error("Text-to-speech generation timed out. Please try with a shorter text.")
+            raise Exception("Text-to-speech generation timed out. Please try with a shorter text.")
         except Exception as e:
             if os.path.exists(temp_filename):
                 os.remove(temp_filename)
@@ -364,7 +364,7 @@ Follow this example structure:
 
     async def combine_audio_files(self, audio_files: List[str], progress=None) -> str:
         if progress:
-            progress(0.9, "Combining audio files...")
+            progress.progress(0.9, "Combining audio files...")
             
         combined_audio = AudioSegment.empty()
         for audio_file in audio_files:
@@ -375,14 +375,14 @@ Follow this example structure:
         combined_audio.export(output_filename, format="wav")
         
         if progress:
-            progress(1.0, "Podcast generated successfully!")
+            progress.progress(1.0, "Podcast generated successfully!")
             
         return output_filename
 
     async def generate_podcast(self, input_text: str, language: str, speaker1: str, speaker2: str, api_key: str, file_obj=None, progress=None) -> str:
         try:
             if progress:
-                progress(0.1, "Starting podcast generation...")
+                progress.progress(0.1, "Starting podcast generation...")
                 
             # Set overall timeout for the entire process
             return await asyncio.wait_for(
@@ -390,18 +390,18 @@ Follow this example structure:
                 timeout=600  # 10 minutes total timeout
             )
         except asyncio.TimeoutError:
-            raise gr.Error("The podcast generation process timed out. Please try with shorter text or try again later.")
+            raise Exception("The podcast generation process timed out. Please try with shorter text or try again later.")
         except Exception as e:
-            raise gr.Error(f"Error generating podcast: {str(e)}")
+            raise Exception(f"Error generating podcast: {str(e)}")
     
     async def _generate_podcast_internal(self, input_text: str, language: str, speaker1: str, speaker2: str, api_key: str, file_obj=None, progress=None) -> str:
         if progress:
-            progress(0.2, "Generating podcast script...")
+            progress.progress(0.2, "Generating podcast script...")
             
         podcast_json = await self.generate_script(input_text, language, api_key, file_obj, progress)
         
         if progress:
-            progress(0.5, "Converting text to speech...")
+            progress.progress(0.5, "Converting text to speech...")
         
         # Process TTS in batches to prevent overwhelming the system
         audio_files = []
@@ -410,7 +410,7 @@ Follow this example structure:
         for i, item in enumerate(podcast_json['podcast']):
             if progress:
                 current_progress = 0.5 + (0.4 * (i / total_lines))
-                progress(current_progress, f"Processing speech {i+1}/{total_lines}...")
+                progress.progress(current_progress, f"Processing speech {i+1}/{total_lines}...")
                 
             try:
                 audio_file = await self.tts_generate(item['line'], item['speaker'], speaker1, speaker2)
@@ -420,12 +420,12 @@ Follow this example structure:
                 for file in audio_files:
                     if os.path.exists(file):
                         os.remove(file)
-                raise gr.Error(f"Error generating speech for line {i+1}: {str(e)}")
+                raise Exception(f"Error generating speech for line {i+1}: {str(e)}")
         
         combined_audio = await self.combine_audio_files(audio_files, progress)
         return combined_audio
-    
-async def process_input(input_text: str, input_file, language: str, speaker1: str, speaker2: str, api_key: str = "", progress=gr.Progress()) -> str:
+
+async def process_input(input_text: str, input_file, language: str, speaker1: str, speaker2: str, api_key: str = "", progress=None) -> str:
     start_time = time.time()
 
     voice_names = {
@@ -443,12 +443,13 @@ async def process_input(input_text: str, input_file, language: str, speaker1: st
     speaker2 = voice_names[speaker2]
 
     try:
-        progress(0.05, "Processing input...")
+        if progress:
+            progress.progress(0.05, "Processing input...")
 
         if not api_key:
             api_key = os.getenv("GENAI_API_KEY")
             if not api_key:
-                raise gr.Error("No API key provided. Please provide a Gemini API key.")
+                raise Exception("No API key provided. Please provide a Gemini API key.")
 
         podcast_generator = PodcastGenerator()
         podcast = await podcast_generator.generate_podcast(input_text, language, speaker1, speaker2, api_key, input_file, progress)
@@ -461,18 +462,25 @@ async def process_input(input_text: str, input_file, language: str, speaker1: st
         # Ensure we show a user-friendly error
         error_msg = str(e)
         if "rate limit" in error_msg.lower():
-            raise gr.Error("Rate limit exceeded. Please try again later or use your own API key.")
+            raise Exception("Rate limit exceeded. Please try again later or use your own API key.")
         elif "timeout" in error_msg.lower():
-            raise gr.Error("The request timed out. This could be due to server load or the length of your input. Please try again with shorter text.")
+            raise Exception("The request timed out. This could be due to server load or the length of your input. Please try again with shorter text.")
         else:
-            raise gr.Error(f"Error: {error_msg}")
+            raise Exception(f"Error: {error_msg}")
 
-iface = gr.Interface(
-    fn=process_input,
-    inputs=[
-        gr.Textbox(label="Input Text"),
-        gr.File(label="Or Upload a PDF or TXT file"),
-        gr.Dropdown(label="Language", choices=[
+# Streamlit UI
+def main():
+    st.set_page_config(page_title="PodcastGen 🎙️", page_icon="🎙️", layout="wide")
+    
+    st.title("PodcastGen 🎙️")
+    st.write("Generate a 2-speaker podcast from text input or documents!")
+    
+    with st.sidebar:
+        st.header("Configuration")
+        api_key = st.text_input("Your Gemini API Key (Optional)", type="password", 
+                              help="In case you are getting rate limited")
+        
+        language_options = [
             "Auto Detect",
             "Afrikaans", "Albanian", "Amharic", "Arabic", "Armenian", "Azerbaijani",
             "Bahasa Indonesian", "Bangla", "Basque", "Bengali", "Bosnian", "Bulgarian",
@@ -487,20 +495,10 @@ iface = gr.Interface(
             "Slovak", "Slovene", "Somali", "Spanish", "Sundanese", "Swahili",
             "Swedish", "Tamil", "Telugu", "Thai", "Turkish", "Ukrainian", "Urdu",
             "Uzbek", "Vietnamese", "Welsh", "Zulu"
-        ],
-        value="Auto Detect"),
-        gr.Dropdown(label="Speaker 1 Voice", choices=[
-            "Andrew - English (United States)",
-            "Ava - English (United States)",
-            "Brian - English (United States)",
-            "Emma - English (United States)",
-            "Florian - German (Germany)",
-            "Seraphina - German (Germany)",
-            "Remy - French (France)",
-            "Vivienne - French (France)"
-        ],
-        value="Andrew - English (United States)"),
-        gr.Dropdown(label="Speaker 2 Voice", choices=[
+        ]
+        language = st.selectbox("Language", language_options, index=0)
+        
+        voice_options = [
             "Andrew - English (United States)",
             "Ava - English (United States)",
             "Brian - English (United States)",
@@ -509,17 +507,95 @@ iface = gr.Interface(
             "Seraphina - German (Germany)",
             "Remy - French (France)",
             "Vivienne - French (France)"
-        ],
-        value="Ava - English (United States)"),
-        gr.Textbox(label="Your Gemini API Key (Optional) - In case you are getting rate limited"),
-    ],
-    outputs=[
-        gr.Audio(label="Generated Podcast Audio")
-    ],
-    title="PodcastGen 🎙️",
-    description="Generate a 2-speaker podcast from text input or documents!",
-    allow_flagging="never",
-)
+        ]
+        speaker1 = st.selectbox("Speaker 1 Voice", voice_options, index=0)
+        speaker2 = st.selectbox("Speaker 2 Voice", voice_options, index=1)
+    
+    col1, col2 = st.columns([2, 1])
+    
+    with col1:
+        input_text = st.text_area("Input Text", height=250)
+    
+    with col2:
+        uploaded_file = st.file_uploader("Or Upload a PDF or TXT file", type=["pdf", "txt"])
+    
+    if st.button("Generate Podcast"):
+        if not input_text and not uploaded_file:
+            st.error("Please provide either input text or upload a file.")
+            return
+        
+        # Create a progress bar for the async operation
+        progress_bar = st.progress(0)
+        status_text = st.empty()
+        
+        # Create a progress wrapper for compatibility with the existing code
+        class StreamlitProgress:
+            def progress(self, value, text=None):
+                progress_bar.progress(value)
+                if text:
+                    status_text.text(text)
+        
+        try:
+            # Prepare file if uploaded
+            file_obj = None
+            if uploaded_file:
+                # Save the uploaded file to a temporary location
+                file_path = f"temp_upload_{uuid.uuid4()}{os.path.splitext(uploaded_file.name)[1]}"
+                with open(file_path, "wb") as f:
+                    f.write(uploaded_file.getbuffer())
+                
+                class FileWrapper:
+                    def __init__(self, path, name):
+                        self.name = name
+                        self.path = path
+                    
+                    @property
+                    def name(self):
+                        return self._name
+                    
+                    @name.setter
+                    def name(self, value):
+                        self._name = value
+                
+                file_obj = FileWrapper(file_path, uploaded_file.name)
+                file_obj.name = file_path  # Set the path as the name for proper file reading
+            
+            # Run the async function in a new event loop
+            progress_wrapper = StreamlitProgress()
+            audio_file = asyncio.run(process_input(
+                input_text, 
+                file_obj, 
+                language, 
+                speaker1, 
+                speaker2, 
+                api_key,
+                progress_wrapper
+            ))
+            
+            # Display the audio
+            st.subheader("Generated Podcast")
+            st.audio(audio_file, format="audio/wav")
+            
+            # Provide a download button
+            with open(audio_file, "rb") as f:
+                audio_bytes = f.read()
+            
+            st.download_button(
+                label="Download Podcast",
+                data=audio_bytes,
+                file_name="podcast.wav",
+                mime="audio/wav"
+            )
+            
+            # Clean up the temporary file
+            if file_obj:
+                try:
+                    os.remove(file_path)
+                except:
+                    pass
+            
+        except Exception as e:
+            st.error(str(e))
 
 if __name__ == "__main__":
-    iface.launch()
+    main()