Update app.py

app.py

@@ -12,9 +12,6 @@ import logging
 from dash.exceptions import PreventUpdate
 import pandas as pd
 import time
-import os
-from huggingface_hub import HfApi
-from dash import callback_context
 
 # Set up logging
 logging.basicConfig(level=logging.INFO)
@@ -23,12 +20,8 @@ logger = logging.getLogger(__name__)
 # Initialize Dash app
 app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
 
-# Initialize Gemini AI with the API key from HF secret
-gemini_api_key = os.environ.get("GEMINI_API_KEY")
-if not gemini_api_key:
-    raise ValueError("GEMINI_API_KEY not found in environment variables. Please set it as a secret in your Hugging Face Space.")
-
-genai.configure(api_key=gemini_api_key)
+# Initialize Gemini AI
+genai.configure(api_key='YOUR_GEMINI_API_KEY')
 
 def generate_podcast_script(api_key, content, duration, num_hosts):
     genai.configure(api_key=api_key)
@@ -47,8 +40,6 @@ def generate_podcast_script(api_key, content, duration, num_hosts):
         Do not use any special characters or markdown. Only include the monologue with proper punctuation.
         Ensure the content flows naturally and stays relevant to the topic.
         Limit the script length to match the requested duration of {duration}.
-        Do not put an intro our outro music as I only need the dialog
-        The dialog must have proper punctuation like apostrophes
         """
     else:
         prompt = f"""
@@ -63,8 +54,6 @@ def generate_podcast_script(api_key, content, duration, num_hosts):
         Do not use any special characters or markdown. Only include the alternating dialogue lines with proper punctuation.
         Ensure the conversation flows naturally and stays relevant to the topic.
         Limit the script length to match the requested duration of {duration}.
-        Do not put an intro our outro music as I only need the dialog
-        The dialog must have proper punctuation like apostrophes
         """
     
     response = model.generate_content(prompt)
@@ -166,6 +155,7 @@ app.layout = dbc.Container([
     
     dbc.Card([
         dbc.CardBody([
+            dbc.Input(id="api-key-input", type="password", placeholder="Enter your Gemini API Key"),
             dbc.Textarea(id="content-input", placeholder="Paste your content or upload a document", rows=5, className="my-3"),
             dcc.Upload(
                 id='document-upload',
@@ -254,66 +244,108 @@ def update_voice2_options(lang):
 
 @app.callback(
     [Output("script-output", "value"),
-     Output("script-progress", "value"),
-     Output("audio-output", "children"),
+     Output("script-progress", "value")],
+    Input("generate-btn", "n_clicks"),
+    [State("api-key-input", "value"),
+     State("content-input", "value"),
+     State("duration", "value"),
+     State("num-hosts", "value")],
+    prevent_initial_call=True
+)
+def generate_script(n_clicks, api_key, content, duration, num_hosts):
+    if n_clicks is None:
+        raise PreventUpdate
+    try:
+        for i in range(10):
+            time.sleep(0.5)  # Simulate progress
+            # Instead of yielding, we'll just pass and update at the end
+            pass
+        script = generate_podcast_script(api_key, content, duration, num_hosts)
+        return script, 100
+    except Exception as e:
+        logger.error(f"Error generating script: {str(e)}")
+        return f"Error: {str(e)}", 0
+
+@app.callback(
+    [Output("audio-output", "children"),
      Output("download-audio", "data"),
      Output("podcast-progress", "value")],
-    [Input("generate-btn", "n_clicks"),
-     Input("generate-podcast-btn", "n_clicks")],
-    [State("content-input", "value"),
-     State("duration", "value"),
-     State("num-hosts", "value"),
+    Input("generate-podcast-btn", "n_clicks"),
+    [State("api-key-input", "value"),
+     State("script-output", "value"),
      State("voice1-select", "value"),
      State("voice2-select", "value"),
-     State("script-output", "value")],
+     State("num-hosts", "value")],
     prevent_initial_call=True
 )
-def generate_script_and_podcast(generate_clicks, podcast_clicks, content, duration, num_hosts, voice1, voice2, script):
-    ctx = callback_context
-    triggered_id = ctx.triggered[0]['prop_id'].split('.')[0]
-
-    if not ctx.triggered:
+def render_and_download_podcast(n_clicks, api_key, script, voice1, voice2, num_hosts):
+    if n_clicks is None:
         raise PreventUpdate
+    try:
+        # Run the async function in a synchronous context
+        sample_rate, audio_data = asyncio.run(render_podcast(api_key, script, voice1, voice2, num_hosts))
+        
+        # Convert numpy array to WAV
+        wav_audio = AudioSegment(
+            audio_data.tobytes(), 
+            frame_rate=sample_rate, 
+            sample_width=audio_data.dtype.itemsize, 
+            channels=1
+        )
+        
+        # Convert WAV to MP3
+        buffer = io.BytesIO()
+        wav_audio.export(buffer, format="mp3")
+        buffer.seek(0)
+        mp3_bytes = buffer.getvalue()
+        
+        # Create base64 audio for playback
+        audio_base64 = base64.b64encode(mp3_bytes).decode('utf-8')
+        audio_src = f"data:audio/mp3;base64,{audio_base64}"
+        
+        return html.Audio(src=audio_src, controls=True), dcc.send_bytes(mp3_bytes, "podcast.mp3"), 100
+    except Exception as e:
+        logger.error(f"Error rendering podcast: {str(e)}")
+        return html.Div(f"Error: {str(e)}"), None, 0
 
-    if triggered_id == "generate-btn":
-        try:
-            for i in range(10):
-                time.sleep(0.5)  # Simulate progress
-            script = generate_podcast_script(content, duration, num_hosts)
-            return script, 100, dash.no_update, dash.no_update, dash.no_update
-        except Exception as e:
-            logger.error(f"Error generating script: {str(e)}")
-            return f"Error: {str(e)}", 0, dash.no_update, dash.no_update, dash.no_update
+@app.callback(
+    [Output("lang2-select", "style"),
+     Output("voice2-select", "style")],
+    Input("num-hosts", "value")
+)
+def update_second_voice_visibility(num_hosts):
+    if num_hosts == 2:
+        return {"display": "block"}, {"display": "block"}
+    else:
+        return {"display": "none"}, {"display": "none"}
 
-    elif triggered_id == "generate-podcast-btn":
+@app.callback(
+    Output("content-input", "value"),
+    Input("document-upload", "contents"),
+    State("document-upload", "filename"),
+    prevent_initial_call=True
+)
+def update_content(contents, filename):
+    if contents is not None:
+        content_type, content_string = contents.split(',')
+        decoded = base64.b64decode(content_string)
         try:
-            # Run the async function in a synchronous context
-            sample_rate, audio_data = asyncio.run(render_podcast(script, voice1, voice2, num_hosts))
-            
-            # Convert numpy array to WAV
-            wav_audio = AudioSegment(
-                audio_data.tobytes(), 
-                frame_rate=sample_rate, 
-                sample_width=audio_data.dtype.itemsize, 
-                channels=1
-            )
-            
-            # Convert WAV to MP3
-            buffer = io.BytesIO()
-            wav_audio.export(buffer, format="mp3")
-            buffer.seek(0)
-            mp3_bytes = buffer.getvalue()
-            
-            # Create base64 audio for playback
-            audio_base64 = base64.b64encode(mp3_bytes).decode('utf-8')
-            audio_src = f"data:audio/mp3;base64,{audio_base64}"
-            
-            return dash.no_update, dash.no_update, html.Audio(src=audio_src, controls=True), dcc.send_bytes(mp3_bytes, "podcast.mp3"), 100
+            if 'csv' in filename:
+                # Assume that the user uploaded a CSV file
+                df = pd.read_csv(io.StringIO(decoded.decode('utf-8')))
+                return df.to_string()
+            elif 'xls' in filename:
+                # Assume that the user uploaded an excel file
+                df = pd.read_excel(io.BytesIO(decoded))
+                return df.to_string()
+            elif 'txt' in filename or 'md' in filename:
+                # Assume that the user uploaded a text or markdown file
+                return decoded.decode('utf-8')
+            else:
+                return 'Unsupported file type. Please upload a CSV, Excel, text, or markdown file.'
         except Exception as e:
-            logger.error(f"Error rendering podcast: {str(e)}")
-            return dash.no_update, dash.no_update, html.Div(f"Error: {str(e)}"), None, 0
-
-    return dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update
+            logger.error(f"Error processing uploaded file: {str(e)}")
+            return f'There was an error processing this file: {str(e)}'
 
 # Run the app
 if __name__ == '__main__':