A script to convert google keep notes to a bunch markdown files aiming mainly to nextcloud notes

convert-keep-notes-to-md.py

#!/usr/bin/env python3

# from pdb import set_trace as bkp
from sys import argv, exit
from os import listdir, path, mkdir, utime
from json import loads as to_dict
from datetime import datetime as date

if '-h' in argv or '--help' in argv:
  print('''
  The commands will print those informations on notes:
  -a all
  -p pinned
  -c note color
  -l labels (it will print only when there's more than one)
  -le last edit date
  -s people who the note is shared with
  
  None will be printed by default
  ''')
  exit()
  
is_printing = {
  'pinned': '-a' in argv or '-p' in argv,
  'color': '-a' in argv or '-c' in argv,
  'labels': '-a' in argv or '-l' in argv,
  'last-edit': '-a' in argv or '-c' in argv,
  'shared': '-a' in argv or '-c' in argv,
  'none': len(argv) == 1
}

keep_dir = './Keep'
labels_file = 'Labels.txt'
nextcloud_dir = './nextcloud-notes'

if not path.exists(keep_dir):
  print('"Keep" folder not found. Place this file on the same folder as "Keep" backup folder')
  exit()

# get labels to create folders with the labels names
with open(path.join(keep_dir, labels_file), 'a+') as f:
  f.seek(0)
  labels = f.read().split('\n')
if len(labels): labels.pop()
labels.append('trash')
labels.append('archived')

# read files and separate the note jsons from attachments
files = listdir(keep_dir)

notes = list(filter(lambda f: '.json' in f, files))
attachments = list(filter(lambda f: '.json' not in f and '.html' not in f and f != labels_file, files))

print('{} notes found\n\n'.format(len(notes)))

def fix_filename(filename):
  new_name = filename
  for char in ['<', '>', ':', '"', '/', '\\', '|', '?', '*']:
    new_name = new_name.replace(char, '-')
  return new_name

# create nextcloud and label folders
if not path.exists(nextcloud_dir):
  mkdir(nextcloud_dir)
  print('created nextcloud folder')

for label in labels:
  label_dir = path.join(nextcloud_dir, fix_filename(label))
  if not path.exists(label_dir):
    mkdir(label_dir)
    print('created "{}" folder inside nextcloud folder'.format(label_dir))
  
# create notes
for note_file in notes:
  # if note_file == 'file-to-debug.json': bkp()
  print('processing note "{}"'.format(note_file))
  note = to_dict(open(path.join(keep_dir, note_file), encoding='utf-8').read())
  text = ''
  note_labels = []
  
  # simplify labels array
  if 'labels' in note:
    for l in note['labels']:
      note_labels.append(l['name'])
  
  # create note body
  if note['title']: text += '# ' + note['title'] + '\n'
  if is_printing['pinned'] and note['isPinned']: text += '**PINNED**\n\n'
  elif note['title']: text += '\n'

  #
  # create note content
  #
  if 'textContent' in note: text += note['textContent']
  elif 'listContent' in note:
    # print MD list
    for item in note['listContent']:
      text += '- [{}] {}\n'.format('x' if item['isChecked'] else ' ', item['text'])
      
  # attachments
  if 'attachments' in note:
    for a in note['attachments']:
      text += '![]({})\n'.format(a['filePath'])

  is_space_added = False
  # add space before last information if it isn't added yet
  def add_space():
    global is_space_added
    global text
    if not is_space_added:
      text += '\n'
      is_space_added = True
    
  # print color
  if is_printing['color'] and ('color' in note and note['color'] != 'DEFAULT'):
    add_space()
    text += '\ncolor: ' + note['color']
    
  # print label when it has more than one
  if is_printing['labels'] and len(note_labels) > 1:
    add_space()
    text += '\nlabels: ' + str(note_labels)[1:len(str(note_labels))-1].replace('\'', '')
    
  # print last edit time
  if is_printing['last-edit']:
    add_space()
    timestamp = int(int(note['userEditedTimestampUsec']) / 1000 / 1000)
    date_str = date.utcfromtimestamp(timestamp).strftime('%d %B %Y %H:%M:%S')
    text += '\nlast edit: {}'.format(date_str)
    
  # print if is shared
  if is_printing['shared'] and 'sharees' in note:
    add_space()
    text += 'shared with: '
    for u in note['sharees']:
      text += '{} {}'.format(u['email'], 'is the owner\n' if u['isOwner'] else '')
  #
  # end note content creation
  #
  
  # decide path
  path_name = ''
  if len(note_labels): path_name = fix_filename(note_labels[0])
  if note['isArchived']: path_name = 'archived'
  elif note['isTrashed']: path_name = 'trash'
    
  # create note
  fname = '{}.{}'.format(note_file[0:len(note_file)-5], 'md')
  with open(path.join(nextcloud_dir, path_name, fname), 'w', encoding='utf-8') as f:
    # use the original modification time for the new file
    original_mod_time = path.getmtime(path.join(keep_dir, note_file))
    utime(path.join(nextcloud_dir, path_name, fname), (original_mod_time, original_mod_time))
    f.write(text)
  
  print('note created inside "{}" folder\n'.format(path_name or 'nextcloud'))
  
print('\nAll notes converted and placed into {} folder'.format(nextcloud_dir))

This is very useful! Thanks a lot.

I would suggest a small improvement, in my case I didn't have any label, and so the file "Labels.txt" didn't exist, which resulted in "FileNotFoundError". I think line 41 should be replaced by:

with open(path.join(keep_dir, labels_file), 'a+') as f:
  f.seek(0)
  labels = f.read().split('\n')

Or alternatively, replace line 41 and 42 with:

try:
  labels = open(path.join(keep_dir, labels_file)).read().split('\n')
  labels.pop()
except:
  labels = []

I didn't thoroughly test it, but it seems to work fine like that.

Hello Mageek627 thanks for the suggestions, I'm very happy that this was useful for someone!

I will pick your first suggestion, I'm not using the script in the moment too so i will change it and commit for the next one who will use it, if there's a problem i hope the people notify me.

Thanks man!

Just wanted to say thanks for providing this. Seems to have worked perfectly to convert, now I just need to import the resulting files into nextcloud.

Many thanks for this, very useful. I hit errors to do with Unicode character mapping e.g.:

UnicodeEncodeError: 'charmap' codec can't encode character '\u200b' in position 1: character maps to <undefined>

So I changed line 149 from:

f = open(path.join(nextcloud_dir, path_name, fname), 'w')

to:

f = open(path.join(nextcloud_dir, path_name, fname), 'w', encoding="utf-8")

I know, hardcoding character for an environment is wrong but I figured this would be okay for a one-off conversion.

Just wanted to say thanks for providing this. Seems to have worked perfectly to convert, now I just need to import the resulting files into nextcloud.

I'm very glad that this had helped you! :)

Many thanks for this, very useful. I hit errors to do with Unicode character mapping e.g.:

UnicodeEncodeError: 'charmap' codec can't encode character '\u200b' in position 1: character maps to <undefined>

So I changed line 149 from:

f = open(path.join(nextcloud_dir, path_name, fname), 'w')

to:

f = open(path.join(nextcloud_dir, path_name, fname), 'w', encoding="utf-8")

I know, hardcoding character for an environment is wrong but I figured this would be okay for a one-off conversion.

thanks for the suggestion, i'll implement here, i'm very glad that this helped you!

Thanks for this useful script.

This small change sets the file modification time so it matches the original Google file. I find this helpful as then the Notes app can sort by date.

Line 5:

from os import listdir, path, mkdir, utime

Insert at line 153:

# use the original modification time for the new file
original_mod_time = path.getmtime(path.join(keep_dir, note_file))
utime(path.join(nextcloud_dir, path_name, fname), (original_mod_time, original_mod_time))

Thanks for this useful script.

This small change sets the file modification time so it matches the original Google file. I find this helpful as then the Notes app can sort by date.

Line 5:

from os import listdir, path, mkdir, utime

Insert at line 153:

# use the original modification time for the new file original_mod_time = path.getmtime(path.join(keep_dir, note_file)) utime(path.join(nextcloud_dir, path_name, fname), (original_mod_time, original_mod_time))

You're welcome, thanks for the suggestion i'll make it right now

The utime call needs to follow the f.close() statement. Otherwise, the close call will update the modification time to the current time. :)

The utime call needs to follow the f.close() statement. Otherwise, the close call will update the modification time to the current time. :)

man, there's a lot of time that I don't put my hands on this, can you send me an example for me to update the script?

does anyone have an idea how to extend this ingenious script so that the german umlauts (ä, ö, ü, ß) are converted correctly?

EDIT: ok it was easier than I thought

Line 77:

From: note = to_dict(open(path.join(keep_dir, note_file)).read())
to this: note = to_dict(open(path.join(keep_dir, note_file), encoding='utf-8').read())

updated line 77 as @Expello mentioned

Hello,

thanks for the script, it saved my day :-)

I think there is a small mistake, line 27 'last-edit': '-a' in argv or '-c' in argv, should be 'last-edit': '-a' in argv or '-le' in argv,

And related to this, to keep even better control of the imported notes, could the updated timestamp of the new files be changed to that 'last edit' date?

I did it with a small script after importing the notes, but I think it'd be natural to do it during the import time.

Thanks!