Update app.py

app.py

@@ -5,16 +5,21 @@ import pdfplumber
 from docx import Document
 import openpyxl
 from werkzeug.utils import secure_filename
+from threading import Thread
 from model_utils import extract_mcqs_with_model
 
 app = Flask(__name__)
 
-# Use /tmp/uploads to avoid permission errors on HF Spaces
+# Use /tmp/uploads to avoid permission errors
 app.config['UPLOAD_FOLDER'] = '/tmp/uploads'
 os.makedirs(app.config['UPLOAD_FOLDER'], exist_ok=True)
 
 quiz_data_store = {}
+processing_results = {}
 
+# -------------------------------
+# Text Extraction Functions
+# -------------------------------
 def extract_text_from_pdf(filepath):
     with pdfplumber.open(filepath) as pdf:
         return "\n".join([p.extract_text() for p in pdf.pages if p.extract_text()])
@@ -32,6 +37,31 @@ def extract_text_from_excel(filepath):
             text += " ".join([str(cell) for cell in row if cell is not None]) + "\n"
     return text
 
+# -------------------------------
+# Background Thread Function
+# -------------------------------
+def background_process(file_path, quiz_id, ext):
+    try:
+        if ext == 'pdf':
+            text = extract_text_from_pdf(file_path)
+        elif ext == 'docx':
+            text = extract_text_from_docx(file_path)
+        elif ext in ['xls', 'xlsx']:
+            text = extract_text_from_excel(file_path)
+        else:
+            processing_results[quiz_id] = {'error': 'Unsupported file type'}
+            return
+
+        mcqs = extract_mcqs_with_model(text)
+        processing_results[quiz_id] = {'mcqs': mcqs}
+        quiz_data_store[quiz_id] = mcqs
+
+    except Exception as e:
+        processing_results[quiz_id] = {'error': str(e)}
+
+# -------------------------------
+# Non-blocking Upload Route
+# -------------------------------
 @app.route('/upload', methods=['POST'])
 def upload_file():
     if 'file' not in request.files:
@@ -45,21 +75,29 @@ def upload_file():
     save_path = os.path.join(app.config['UPLOAD_FOLDER'], uid + '_' + filename)
     file.save(save_path)
 
-    if ext == 'pdf':
-        text = extract_text_from_pdf(save_path)
-    elif ext == 'docx':
-        text = extract_text_from_docx(save_path)
-    elif ext in ['xls', 'xlsx']:
-        text = extract_text_from_excel(save_path)
-    else:
-        return jsonify({'error': 'Unsupported file type'}), 400
-
-    mcqs = extract_mcqs_with_model(text)
     quiz_id = str(uuid.uuid4())
-    quiz_data_store[quiz_id] = mcqs
+    thread = Thread(target=background_process, args=(save_path, quiz_id, ext))
+    thread.start()
+
+    return jsonify({'quiz_id': quiz_id, 'status': 'processing'})
+
+# -------------------------------
+# Status Check Route
+# -------------------------------
+@app.route('/status/<quiz_id>', methods=['GET'])
+def check_status(quiz_id):
+    if quiz_id not in processing_results:
+        return jsonify({'status': 'processing'}), 202
+
+    result = processing_results[quiz_id]
+    if 'error' in result:
+        return jsonify({'status': 'failed', 'error': result['error']}), 500
 
-    return jsonify({'quiz_id': quiz_id, 'mcqs': mcqs})
+    return jsonify({'status': 'completed', 'mcqs': result['mcqs']}), 200
 
+# -------------------------------
+# Quiz Submission
+# -------------------------------
 @app.route('/submit', methods=['POST'])
 def submit_quiz():
     data = request.json
@@ -80,9 +118,15 @@ def submit_quiz():
 
     return jsonify({'score': correct, 'total': total, 'accuracy': accuracy})
 
+# -------------------------------
+# Root Route
+# -------------------------------
 @app.route('/')
 def home():
     return "MCQ Extraction Flask API is running!"
 
+# -------------------------------
+# Start the App
+# -------------------------------
 if __name__ == '__main__':
     app.run(host='0.0.0.0', port=7860)