import os
import re
import base64

INPUT_CSS = 'SD/css/imgs.css'
OUTPUT_DIR = 'Imgs/'

os.makedirs(OUTPUT_DIR, exist_ok=True)


with open(INPUT_CSS, 'r') as f:
    css = f.read()
print('First 500 chars of CSS:')
print(css[:500])


# Improved regex to match the actual CSS structure in imgs.css
# Handles: img.className { ... content: url(data:image/TYPE;base64,BASE64DATA) ... }

# Regex for CSS image blocks with single braces and correct escaping
pattern = re.compile(
    r'img\.([a-zA-Z0-9_]+)\s*\{[^}]*?content\s*:\s*url\(data:image/(png|bmp|jpeg|jpg);base64,([A-Za-z0-9+/=\s]+)\)[^}]*?\}',
    re.DOTALL)


matches = list(pattern.finditer(css))
print(f"Found {len(matches)} images in CSS.")
if matches:
    print('First match:')
    print(matches[0].group(0))

for match in matches:
    class_name, img_type, b64data = match.groups()
    ext = 'jpg' if img_type == 'jpeg' else img_type
    out_path = os.path.join(OUTPUT_DIR, f"{class_name}.{ext}")
    try:
        clean_b64 = ''.join(b64data.split())
        with open(out_path, 'wb') as imgf:
            imgf.write(base64.b64decode(clean_b64))
        print(f"Extracted {out_path} (class: {class_name}, type: {img_type})")
    except Exception as e:
        print(f"Failed to extract {out_path}: {e}")

print("Done.")