Update app_main.py

app_main.py

@@ -17,7 +17,7 @@ import tempfile
 import torch
 from langchain_groq import ChatGroq
 from langgraph.prebuilt import create_react_agent
-import logging
+import logging, time
 
 # Configure logging
 logging.basicConfig(
@@ -126,9 +126,12 @@ def extract_images_from_pdf(pdf_path, output_json_path):
                 f"❌ Failed to extract images from PDF: {str(e)}")
 
         try:
+            start_time = time.perf_counter()
             with open(output_json_path, "w") as f:
                 json.dump([element.to_dict()
                           for element in elements], f, indent=4)
+            elapsed = time.perf_counter() - start_time
+            logger.info(f"✅ extracted.json write in {elapsed:.2f} seconds")
         except Exception as e:
             raise RuntimeError(f"❌ Failed to write extracted.json: {str(e)}")
 
@@ -172,6 +175,7 @@ def extract_images_from_pdf(pdf_path, output_json_path):
             start_count = 1
 
         sprite_count = start_count
+        start_time = time.perf_counter()
         for i, element in enumerate(file_elements):
             if "image_base64" in element["metadata"]:
                 try:
@@ -246,7 +250,9 @@ def extract_images_from_pdf(pdf_path, output_json_path):
                     sprite_count += 1
                 except Exception as e:
                     print(f"⚠️ Error processing Sprite {i+1}: {str(e)}")
-
+        elapsed = time.perf_counter() - start_time
+        logger.info(f"✅ extracted_sprites.json write in {elapsed:.2f} seconds")
+        
         # Save manipulated JSON
         with open(final_json_path, "w") as sprite_file:
             json.dump(manipulated_json, sprite_file, indent=4)
@@ -294,12 +300,15 @@ def similarity_matching(input_json_path: str) -> str:
         sprites_data = json.load(f)
 
     sprite_ids, texts, sprite_base64 = [], [], []
+    start_time = time.perf_counter()
     for sid, sprite in sprites_data.items():
         sprite_ids.append(sid)
         texts.append(
             "This is " + sprite.get("description", sprite.get("name", "")))
         sprite_base64.append(sprite["base64"])
-
+    elapsed = time.perf_counter() - start_time
+    logger.info(f"✅ Append Sprite's Name and Description in {elapsed:.2f} seconds")
+    
     # ============================== #
     #     INITIALIZE CLIP EMBEDDER   #
     # ============================== #
@@ -340,13 +349,16 @@ def similarity_matching(input_json_path: str) -> str:
     # ============================== #
     temp_dir = tempfile.mkdtemp()
     sprite_image_paths = []
+    start_time = time.perf_counter()
     for idx, b64 in enumerate(sprite_base64):
         image_data = base64.b64decode(b64.split(",")[-1])
         img = Image.open(BytesIO(image_data)).convert("RGB")
         temp_path = os.path.join(temp_dir, f"sprite_{idx}.png")
         img.save(temp_path)
         sprite_image_paths.append(temp_path)
-
+    elapsed = time.perf_counter() - start_time
+    logger.info(f"✅ Decoded Sprite Base64 in {elapsed:.2f} seconds")
+    
     # ============================== #
     #      EMBED SPRITE IMAGES       #
     # ============================== #
@@ -367,6 +379,7 @@ def similarity_matching(input_json_path: str) -> str:
     # ============= Match and copy ================
     project_data, backdrop_data = [], []
     copied_folders = set()
+    start_time = time.perf_counter()
     for sprite_idx, matched_idx in enumerate(most_similar_indices):
         matched_entry = embedding_json[matched_idx]
         # matched_image_path = os.path.normpath(folder_image_paths[matched_idx])
@@ -413,11 +426,13 @@ def similarity_matching(input_json_path: str) -> str:
             "agent": "OpenAI ScratchVision Agent"
         }
     }
-
+    start_time = time.perf_counter()
     for sprite in project_data:
         if not sprite.get("isStage", False):
             final_project["targets"].append(sprite)
-
+    elapsed = time.perf_counter() - start_time
+    logger.info(f"✅ Append sprite 'targets' in {elapsed:.2f} seconds")
+    
     if backdrop_data:
         all_costumes, sounds = [], []
         for idx, bd in enumerate(backdrop_data):