import os
import re
import shutil

def clean_md_file(input_path, output_path):
    with open(input_path, 'r', encoding='utf-8') as f:
        content = f.read()

    # Remove YAML frontmatter (with surrounding ---)
    content = re.sub(r'^---\n(.*?\n)*?---\n', '', content, flags=re.DOTALL)

    # Remove all Hugo shortcodes like {{< ... >}} or {{< /... >}}
    content = re.sub(r'{{<[^>]*>}}', '', content)

    # Remove comments shortcode and everything after
    content = re.sub(r'{{<\s*comments\s*>}}.*', '', content, flags=re.DOTALL)

    # Remove trailing shell prompt lines (optional)
    content = re.sub(r'^hacker@selfhost1:.*$', '', content, flags=re.MULTILINE)

    # Collapse multiple blank lines to max two
    content = re.sub(r'\n{3,}', '\n\n', content)

    # Strip leading/trailing whitespace
    content = content.strip()

    # Write cleaned content
    with open(output_path, 'w', encoding='utf-8') as f:
        f.write(content)

    print(f"Processed: {input_path} -> {output_path}")

def batch_clean_md_files():
    input_dir = '/var/www/html/professional-site/content/post'
    assets_dir = '/var/www/html/professional-site/assets/summaries'
    static_dir = '/var/www/html/professional-site/static/summaries'

    os.makedirs(assets_dir, exist_ok=True)
    os.makedirs(static_dir, exist_ok=True)

    for filename in os.listdir(input_dir):
        if filename.endswith('.md'):
            input_path = os.path.join(input_dir, filename)
            output_filename = filename.replace('.md', '.txt')

            output_path_assets = os.path.join(assets_dir, output_filename)
            output_path_static = os.path.join(static_dir, output_filename)

            clean_md_file(input_path, output_path_assets)

            # Also copy the cleaned file to static folder
            shutil.copyfile(output_path_assets, output_path_static)
            print(f"Copied summary to static: {output_path_static}")

batch_clean_md_files()