Skip to content

Instantly share code, notes, and snippets.

@gudata
Created January 21, 2023 12:39
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
Star You must be signed in to star a gist
Save gudata/594a526a3f5ad44d35ccd9b228bf0762 to your computer and use it in GitHub Desktop.
Conver the tomboy notes xml files to markdown format
import os
import xml.etree.ElementTree as ET
import re
directory = "./"
output_directory = "./markdown"
def read_file_as_string(file_path):
with open(file_path, "r") as file:
return file.read()
def process_file(filename):
xml_file = os.path.join(directory, filename)
file_contents = read_file_as_string(xml_file)
match = re.search('<title>(.*?)</title>', file_contents)
if match:
original_filename = match.group(1)
filename = original_filename.replace('http://', '')
filename = re.sub(r'[^a-zA-Z а-яА-Я]', '', filename)
filename = re.sub(r'\s+', ' ', original_filename)
filename = filename.replace('/', '-')
filename += ".md"
markdown = file_contents.replace('<?xml version="1.0" encoding="utf-8"?>', "")
markdown = re.sub("<note[^>]*>", "", markdown)
markdown = re.sub("\s*<title>", "# ", markdown)
markdown = re.sub("</title>", "", markdown)
markdown = re.sub('<text xml:space="preserve">', "", markdown)
markdown = re.sub("</text>", "", markdown)
markdown = re.sub("\s*<text[^>]*>", "", markdown)
markdown = re.sub("<link:internal>", "", markdown)
markdown = re.sub("</link:internal>", "", markdown)
markdown = re.sub("<link:url>", "", markdown)
markdown = re.sub("</link:url>", "", markdown)
markdown = re.sub("<link:broken>", "", markdown)
markdown = re.sub("</link:broken>", "", markdown)
markdown = re.sub("<list>", "", markdown)
markdown = re.sub("</list>", "", markdown)
markdown = re.sub("<list-item[^>]*>", "- ", markdown)
markdown = re.sub("</list-item>", "", markdown)
markdown = re.sub("</note-content>.*", "", markdown, flags=re.DOTALL)
markdown = re.sub("<bold>", "**", markdown)
markdown = re.sub("</bold>", "**", markdown)
markdown = re.sub("<underline>", "_", markdown)
markdown = re.sub("</underline>", "_", markdown)
markdown = re.sub("&#x2028;", "\n", markdown)
markdown = re.sub("&gt;", ">", markdown)
markdown = re.sub("&lt;", "<", markdown)
markdown = re.sub("&amp;", "&", markdown)
with open(os.path.join(output_directory, filename.replace(".xml", ".md")), "w") as f:
f.write(markdown)
for filename in os.listdir(directory):
if filename.endswith(".note"):
process_file(filename)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment