Update app.py

app.py

@@ -67,7 +67,8 @@ try:
     )
     from evaluate import (
         handle_evaluation_request, handle_upload_reference,
-        init_reference_audio, calculate_similarity
+        init_reference_audio, calculate_similarity, preprocess_all_references,
+        get_preprocessing_status  # Import the new function
     )
 
     logger.info("✅ All required libraries imported successfully")
@@ -248,7 +249,7 @@ def home():
     return jsonify({
         "message": "Speech API is running", 
         "status": "active",
-        "version": "1.1",
+        "version": "1.2",  # Updated version to reflect reference preprocessing
         "environment": "Hugging Face Spaces"
     })
 
@@ -272,6 +273,9 @@ def health_check():
         "translation_cache_size": len(translation_cache)
     }
     
+    # Add reference preprocessing status
+    health_status["reference_preprocessing"] = get_preprocessing_status()
+    
     return jsonify(health_status)
 
 # ASR with optimizations
@@ -409,6 +413,12 @@ def evaluate_pronunciation():
     user_output_dir = g.user_output_dir if hasattr(g, 'user_output_dir') else OUTPUT_DIR
     return handle_evaluation_request(request, REFERENCE_AUDIO_DIR, user_output_dir, SAMPLE_RATE)
 
+# New endpoint to check preprocessing status
+@app.route("/reference_preprocessing_status", methods=["GET"])
+def reference_preprocessing_status():
+    """Get the current status of reference audio preprocessing"""
+    return jsonify(get_preprocessing_status())
+
 @app.route("/check_references", methods=["GET"])
 def check_references():
     """Optimized endpoint to check if reference files exist"""
@@ -433,7 +443,8 @@ def check_references():
         "directory_exists": os.path.exists(REFERENCE_AUDIO_DIR),
         "total_patterns": len(ref_patterns),
         "existing_patterns": 0,
-        "total_files": 0
+        "total_files": 0,
+        "preprocessing_status": get_preprocessing_status()  # Add preprocessing status
     }
     
     for pattern in ref_patterns:
@@ -479,7 +490,8 @@ def check_references_detailed():
 
     return jsonify({
         "reference_audio_dir": REFERENCE_AUDIO_DIR,
-        "patterns": results
+        "patterns": results,
+        "preprocessing_status": get_preprocessing_status()  # Add preprocessing status
     })
 
 @app.route("/upload_reference", methods=["POST"])
@@ -487,6 +499,28 @@ def check_references_detailed():
 def upload_reference_audio():
     return handle_upload_reference(request, REFERENCE_AUDIO_DIR, SAMPLE_RATE)
 
+# Add an endpoint to manually trigger reference preprocessing
+@app.route("/preprocess_references", methods=["POST"])
+def manual_preprocess_references():
+    """Manually trigger reference audio preprocessing"""
+    # Only allow from local or with API key
+    if not (request.remote_addr == '127.0.0.1' or 
+            request.headers.get('X-Admin-Key') == os.environ.get('ADMIN_KEY', 'admin-secret')):
+        return jsonify({"error": "Unauthorized"}), 403
+        
+    # Start preprocessing in a background thread to avoid blocking
+    def preprocess_worker():
+        preprocess_all_references(REFERENCE_AUDIO_DIR, SAMPLE_RATE)
+        
+    preprocessing_thread = threading.Thread(target=preprocess_worker)
+    preprocessing_thread.daemon = True
+    preprocessing_thread.start()
+    
+    return jsonify({
+        "message": "Reference preprocessing started in background",
+        "current_status": get_preprocessing_status()
+    })
+
 # Add a cleanup endpoint
 @app.route("/cleanup", methods=["POST"])
 def cleanup_files():
@@ -548,5 +582,10 @@ if __name__ == "__main__":
     for lang, model_status in status['tts_models'].items():
         logger.info(f"📊 TTS model {lang}: {'✅' if model_status == 'loaded' else '❌'}")
     
+    # Log reference preprocessing status
+    preproc_status = get_preprocessing_status()
+    logger.info(f"📊 Reference preprocessing: {'✅ Complete' if preproc_status['complete'] else '🔄 In progress'}")
+    logger.info(f"📊 Preprocessed files: {preproc_status['preprocessed_files']}")
+    
     # Use threaded=True for better performance
     app.run(host="0.0.0.0", port=7860, debug=False, threaded=True)