Minecraft backups: prune the backups to just one file per play session (i.e. find groups of consecutive backups and delete all but the newest)

reduce-to-daily.py

from datetime import datetime, timedelta
from dateutil import tz
from functools import reduce
import os
from pprint import pprint
import re
import sys

if sys.version_info < (3,7):
    sys.exit('Python 3.7 guarantees ordered dicts, which this script relies on')

dir = r'Z:\srv\minecraft\backups\Enigmatica6Expert\backups'
min_time_between_backups = timedelta(hours=3)
min_time_ago_to_cull = timedelta(days=14)
#date_regex = re.compile(r'^Backup--world--(?P<year>[0-9]+)-(?P<month>[0-9]+)-(?P<day>[0-9]+)--(?P<hour>[0-9]+)-(?P<minute>[0-9]+).zip$')
date_regex = re.compile(r'^(?P<year>[0-9]+)-(?P<month>[0-9]+)-(?P<day>[0-9]+)-(?P<hour>[0-9]+)-(?P<minute>[0-9]+)-(?P<second>[0-9]+).zip$')

names = os.listdir(dir)
names.sort() # it seems to already be sorted, but just in case...

def parse_date(filename):
    match = date_regex.match(filename)
    if match is None:
        raise ValueError(f'Invalid filename: {filename}')
    return datetime(int(match['year']), int(match['month']), int(match['day']), int(match['hour']), int(match['minute']), tzinfo=tz.tzutc())

dates = {name: parse_date(name) for name in names}

def sufficiently_spaced(datetime1, datetime2):
    delta = abs(datetime1 - datetime2)
    return delta >= min_time_between_backups

now = datetime.now(tz=tz.tzlocal())

def add_preserve_entry(map, entry):
    name, date = entry

    # if first record, default to preserving it
    if len(map) == 0:
        map[name] = True
        return map

    # get the previous entry
    prevname = list(map.keys())[-1]
    prevdate = dates[prevname]

    # preserve the current entry
    map[name] = True

    # only delete it if it's more than min_time_ago
    old_enough = (now - prevdate) > min_time_ago_to_cull

    # if the previous one was within a short delta, don't keep it
    if old_enough and not sufficiently_spaced(prevdate, date):
        map[prevname] = False

    return map

preserve = reduce(add_preserve_entry, dates.items(), {})
#pprint(preserve)
#exit(0)

files_to_delete = [key for key, value in preserve.items() if value is False]
#pprint(files_to_delete)

files_to_keep = [key for key, value in preserve.items() if value is True]
#print("Keeping files:")
#pprint(files_to_keep)

sizes = {file: os.path.getsize(f'{dir}\{file}') for file in files_to_delete}

total_bytes = sum(sizes.values())

bytes_so_far = 0
for file in files_to_delete:
    os.remove(f'{dir}\{file}')
    bytes_so_far += sizes[file]
    progress = bytes_so_far / total_bytes

    print(f'{file} {progress:.0%}')