Update app.py

app.py

@@ -83,16 +83,66 @@ def user_input(user_question):
     print(response)
     st.write("Reply: ", response["output_text"])
 
-
+import streamlit.components.v1 as components
+
+def voice_recorder(key=None):
+    component_html = """
+    <script src="https://cdn.jsdelivr.net/npm/@ffmpeg/ffmpeg"></script>
+    <button id="recordButton">Record</button>
+    <button id="stopButton" disabled>Stop</button>
+    <audio id="audioPlayback" controls></audio>
+    <script>
+    let recordButton = document.getElementById("recordButton");
+    let stopButton = document.getElementById("stopButton");
+    let audioPlayback = document.getElementById("audioPlayback");
+
+    let recorder;
+    let audioData;
+
+    recordButton.onclick = () => {
+        navigator.mediaDevices.getUserMedia({ audio: true })
+        .then(stream => {
+            const mediaRecorder = new MediaRecorder(stream);
+            mediaRecorder.start();
+            const audioChunks = [];
+            mediaRecorder.addEventListener("dataavailable", event => {
+                audioChunks.push(event.data);
+            });
+            recorder = mediaRecorder;
+            stopButton.disabled = false;
+            recordButton.disabled = true;
+        });
+    };
+
+    stopButton.onclick = () => {
+        recorder.stop();
+        recorder.addEventListener("stop", () => {
+            const audioBlob = new Blob(audioChunks);
+            const audioUrl = URL.createObjectURL(audioBlob);
+            audioPlayback.src = audioUrl;
+            const reader = new FileReader();
+            reader.readAsDataURL(audioBlob);
+            reader.onload = () => {
+                window.parent.postMessage({ audioData: reader.result }, '*');
+            };
+        });
+        stopButton.disabled = true;
+        recordButton.disabled = false;
+    };
+    </script>
+    """
+    audio_base64 = components.html(component_html, height=150, key=key)
+    return audio_base64
+    
 def main():
     st.set_page_config("Chat PDF")
     st.header("Chat with PDF using Gemini💁")
 
-    # Audio input
-    audio_file = st.file_uploader("Upload an audio file or record your question", type=['wav', 'mp3', 'ogg'])
-    if audio_file and st.button("Transcribe and Process"):
-        with st.spinner("Transcribing audio..."):
-            user_question = transcribe_audio(audio_file)
+    # Replace the existing audio file uploader with the custom voice recorder component
+    audio_data = voice_recorder("audio_recorder")
+    if audio_data and st.button("Process Recorded Audio"):
+        with st.spinner("Processing audio..."):
+            user_question = transcribe_audio(audio_data)
             st.write(f"Transcribed Question: {user_question}")
             user_input(user_question)