Delete obsidian notes with all the referenced images

delete_md_with_images.py

"""
Script to delete all images referenced in Markdown files within a specified notes path.
It scans the Markdown files for Obsidian image syntax and deletes all notes with the referenced images.
"""

import os
import re
import argparse

def delete_by_path(file_path):
    """
    Delete file from the file system by its path.
    """
    try:
        if os.path.exists(file_path):
            os.remove(file_path)
            print(f"Deleted: {file_path}")
        else:
            print(f"File {file_path} not found.")
    except OSError as error:
        print(f"Error deleting {file_path}: {error}")


def get_image_files_dict(folder_path):
    """
    Receives a folder path containing images and returns a dictionary
    with image file names (with extension) as keys and full file paths as values.
    Only image files with common extensions are considered.
    """
    image_extensions = {'.jpg', '.jpeg', '.png', '.gif', '.bmp', '.tiff', '.webp'}
    image_files_dict = {}

    for file_name in os.listdir(folder_path):
        file_path = os.path.join(folder_path, file_name)

        if os.path.isfile(file_path):
            file_ext = os.path.splitext(file_name)[1].lower()
            if file_ext in image_extensions:
                image_files_dict[file_name] = file_path

    return image_files_dict


def delete_notes_with_images(folder_path, image_files_dict):
    """
    Traverses the folder and subdirectories to find markdown files. 
    For each markdown file, searches for images with the syntax ![[image.png|title]].
    If an image is found and exists in image_files_dict, it deletes the image file.
    """
    # Regex to match image syntax ![[image.png|title]] in markdown files
    image_pattern = re.compile(r"!\[\[(.+?)\|.+?\]\]")
   
    for root, _, files in os.walk(folder_path):
        for file_name in files:
            if file_name.endswith('.md'):
                file_path = os.path.join(root, file_name)
                print(file_path)
                with open(file_path, 'r', encoding='utf-8') as md_file:
                    content = md_file.read()
                matches = image_pattern.findall(content)
                for image_name in matches:
                    if image_name in image_files_dict:
                        image_path = image_files_dict[image_name]
                        delete_by_path(image_path)
                delete_by_path(file_path)


def main():
    """
    Main function to parse command-line arguments and trigger the deletion process.
    """
    parser = argparse.ArgumentParser(description="Delete all images referenced in Markdown notes.")
    parser.add_argument("notes_path", help="Path to the folder containing markdown notes.")
    parser.add_argument("files_root", help="Path to the root folder containing image files.")
    args = parser.parse_args()
    images = get_image_files_dict(args.files_root)
    delete_notes_with_images(args.notes_path, images)


if __name__ == "__main__":
    main()