Update app.py

app.py

@@ -42,7 +42,7 @@ try:
     import tempfile
     import torchaudio
     import soundfile as sf
-    from flask import Flask, request, jsonify, send_file
+    from flask import Flask, request, jsonify, send_file, g
     from flask_cors import CORS
     from transformers import Wav2Vec2ForCTC, AutoProcessor, VitsModel, AutoTokenizer
     from transformers import MarianMTModel, MarianTokenizer
@@ -533,21 +533,27 @@ def evaluate_pronunciation():
             return jsonify({"error": "No audio file uploaded"}), 400
 
         audio_file = request.files["audio"]
-        reference_word = request.form.get("reference_word", "").strip()
-        language = request.form.get("language", "tagalog").lower()  # Default to tagalog for ASR
+        reference_locator = request.form.get("reference_locator", "").strip()
+        language = request.form.get("language", "kapampangan").lower()
 
-        # Check if reference word is valid
-        reference_patterns = [
-            "mayap_a_abak", "mayap_a_ugtu", "mayap_a_gatpanapun", 
-            "mayap_a_bengi", "komusta_ka"
-        ]
-        
-        if not reference_word or reference_word not in reference_patterns:
-            logger.warning(f"⚠️ Invalid reference word: {reference_word}")
-            return jsonify({"error": f"Invalid reference word. Available: {reference_patterns}"}), 400
+        # Validate reference locator
+        if not reference_locator:
+            logger.warning("⚠️ No reference locator provided")
+            return jsonify({"error": "Reference locator is required"}), 400
+
+        # Construct full reference directory path
+        reference_dir = os.path.join(REFERENCE_AUDIO_DIR, reference_locator)
+        if not os.path.exists(reference_dir):
+            logger.warning(f"⚠️ Reference directory not found: {reference_dir}")
+            return jsonify({"error": f"Reference audio directory not found: {reference_locator}"}), 404
+
+        reference_files = glob.glob(os.path.join(reference_dir, "*.wav"))
+        if not reference_files:
+            logger.warning(f"⚠️ No reference audio files found in {reference_dir}")
+            return jsonify({"error": f"No reference audio found for {reference_locator}"}), 404
 
         lang_code = LANGUAGE_CODES.get(language, language)
-        logger.info(f"🔄 Evaluating pronunciation of '{reference_word}' in {language}")
+        logger.info(f"🔄 Evaluating pronunciation for reference: {reference_locator}")
 
         # Save the uploaded file temporarily
         with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
@@ -571,19 +577,6 @@ def evaluate_pronunciation():
             logger.error(f"❌ Audio processing failed: {str(e)}")
             return jsonify({"error": f"Audio processing failed: {str(e)}"}), 500
 
-        # Find reference audio files
-        reference_dir = os.path.join(REFERENCE_AUDIO_DIR, reference_word)
-        if not os.path.exists(reference_dir):
-            logger.warning(f"⚠️ Reference directory not found: {reference_dir}")
-            return jsonify({"error": f"Reference audio for {reference_word} not found"}), 404
-            
-        reference_files = glob.glob(os.path.join(reference_dir, "*.wav"))
-        if not reference_files:
-            logger.warning(f"⚠️ No reference audio files found in {reference_dir}")
-            return jsonify({"error": f"No reference audio found for {reference_word}"}), 404
-            
-        logger.info(f"📊 Found {len(reference_files)} reference files for '{reference_word}'")
-
         # Transcribe user audio
         try:
             # Process audio for ASR
@@ -663,7 +656,7 @@ def evaluate_pronunciation():
             "score": best_score,
             "feedback": feedback,
             "transcription": user_transcription,
-            "reference_word": reference_word,
+            "reference_locator": reference_locator,
             "details": results
         })