dataset_prep.py

import os
import shutil

dataset_path = "/home/codernotme/Projects/Github/katariaoptics/dataset"
standard_shapes = ["heart", "oblong", "oval", "round", "square"]

def normalize_dataset():
    if not os.path.exists(dataset_path):
        print(f"Dataset path {dataset_path} does not exist.")
        return

    # Create standard folders if they don't exist
    for shape in standard_shapes:
        target_dir = os.path.join(dataset_path, shape)
        os.makedirs(target_dir, exist_ok=True)

    # Walk through the dataset directory
    for item in os.listdir(dataset_path):
        item_path = os.path.join(dataset_path, item)
        
        if os.path.isdir(item_path):
            lower_name = item.lower()
            
            # If it matches a standard shape but has different case (e.g., "Heart")
            if lower_name in standard_shapes and item != lower_name:
                target_dir = os.path.join(dataset_path, lower_name)
                print(f"Merging {item} into {lower_name}...")
                
                for file_name in os.listdir(item_path):
                    src_file = os.path.join(item_path, file_name)
                    dst_file = os.path.join(target_dir, file_name)
                    
                    # Handle duplicates by renaming
                    if os.path.exists(dst_file):
                        base, ext = os.path.splitext(file_name)
                        dst_file = os.path.join(target_dir, f"{base}_1{ext}")
                        
                    shutil.move(src_file, dst_file)
                
                # Remove the empty directory
                os.rmdir(item_path)
                print(f"Removed {item}")

if __name__ == "__main__":
    normalize_dataset()