Update app.py

app.py

@@ -7,41 +7,16 @@ import torch
 import faiss
 import numpy as np
 from gtts import gTTS
+import io
+from PIL import Image
+from flask import Flask, request, jsonify
+from werkzeug.utils import secure_filename
 
-# Function to convert audio file to text
-def audio_to_text(audio_file):
-    recognizer = sr.Recognizer()
-    with sr.AudioFile(audio_file) as source:
-        audio = recognizer.record(source)
-    try:
-        text = recognizer.recognize_google(audio)
-        return text
-    except sr.UnknownValueError:
-        return "Sorry, I did not understand the audio"
-    except sr.RequestError:
-        return "Sorry, there was a problem with the request"
-
-# Function to extract text from a PDF file
-def extract_text_from_pdf(pdf_file):
-    text = ""
-    pdf_document = fitz.open(pdf_file)
-    for page_num in range(len(pdf_document)):
-        page = pdf_document.load_page(page_num)
-        text += page.get_text()
-    return text
-
-# Function to embed text using a transformer model
-def embed_text(texts, model, tokenizer):
-    inputs = tokenizer(texts, return_tensors='pt', truncation=True, padding=True)
-    with torch.no_grad():
-        embeddings = model(**inputs).last_hidden_state.mean(dim=1).numpy()
-    return embeddings
+# Initialize Streamlit components
+st.title("Parenting Guide App")
 
-# Function to convert text to speech
-def text_to_speech(text, output_file):
-    tts = gTTS(text=text, lang='en')
-    tts.save(output_file)
-    return output_file
+# HTML component for the recording interface
+st.components.v1.html(open("audio_recorder.html").read(), height=600)
 
 # Initialize model and tokenizer
 tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
@@ -52,7 +27,7 @@ dimension = 768  # Size of BERT embeddings
 index = faiss.IndexFlatL2(dimension)
 
 # Folder path containing PDFs
-pdf_folder_path = "pdfsforRAG"
+pdf_folder_path = "pdfs"
 
 # Read all PDF files from the specified folder
 pdf_paths = [os.path.join(pdf_folder_path, f) for f in os.listdir(pdf_folder_path) if f.endswith('.pdf')]
@@ -66,26 +41,28 @@ for path in pdf_paths:
 embeddings = embed_text(texts, model, tokenizer)
 index.add(embeddings)
 
-# Streamlit application
-st.title("Parenting Guide App")
+# Define Flask app to handle audio uploads
+app = Flask(__name__)
+
+@app.route('/upload-audio', methods=['POST'])
+def upload_audio():
+    audio_file = request.files['audio']
+    if audio_file:
+        audio_file.save('temp_audio.wav')
+        text = audio_to_text('temp_audio.wav')
 
-audio_file = st.file_uploader("Upload an audio file", type=["wav", "mp3"])
+        # Find relevant advice
+        query_embedding = embed_text([text], model, tokenizer)
+        D, I = index.search(query_embedding, k=1)  # Search for the most similar advice
+        closest_text = texts[I[0][0]]
 
-if audio_file:
-    st.write("Processing...")
+        # Convert advice to speech
+        output_file = "advice.mp3"
+        output_path = text_to_speech(closest_text, output_file)
 
-    # Convert audio to text
-    text = audio_to_text(audio_file)
-    st.write("Voice command:", text)
+        return jsonify({'status': 'success', 'audio': output_file})
 
-    # Find relevant advice
-    query_embedding = embed_text([text], model, tokenizer)
-    D, I = index.search(query_embedding, k=1)  # Search for the most similar advice
-    closest_text = texts[I[0][0]]
-    
-    st.write("Advice:", closest_text)
+    return jsonify({'status': 'error', 'message': 'No audio file received'})
 
-    # Convert advice to speech
-    output_file = "advice.mp3"
-    output_path = text_to_speech(closest_text, output_file)
-    st.audio(output_path)
+if __name__ == '__main__':
+    app.run(debug=True)