Dpkg Time Machine

dpkg-since.py

#!/usr/bin/env python3

# file: dpkg-since.py
# description: lists installed packages since the specified date

# changelog:
# * Feb 10 2015 - init release
# * Mar 24 2015 - new: packages sorted by installation time/date
#                 fix: ignored TIME argument
# * Apr 9 2015  - fix: exclude duplicates
# * Apr 27 2015 - new: recognize dates like '1 hour ago', 'yesterday'
#                      requires 'dateparser'
#               - new: use new line or other splitter character
# * Jul 27 2019 - new: upgraded to python 3
# * Feb 03 2020 - fix: clean up
# * Mar 22 2020 - change: removed 'dateparser' dependency, performance fixes

import sys
import getopt
from os.path import basename
from datetime import datetime
from glob import glob
import gzip
import re
import subprocess

def uniquify(seq):
    seen = set()
    seen_add = seen.add
    return [x for x in seq if x not in seen and not seen_add(x)]

def print_usage():
    basename_ = basename(sys.argv[0])
    print('')
    print('Usage:\n\t' + basename_ + ' [OPTIONS] DATE')
    print('')
    print('Options:')
    print('\t-p, --print-date      - prints the recognized date as the last line')
    print("\t-n, --new-line        - use new line as a splitter character")
    print("\t-s, --splitter 'char' - set a custom splitter character")
    print('\t-h, --help            - display this help and exit')
    print('')
    print('Examples:')
    print('\t' + basename_ + ' -p yesterday')
    print('\t' + basename_ + ' -n 1 week ago')
    print('\t' + basename_ + " -s ', ' 2015-02-14 16:40:35")

def main():
    print_date = False
    splitter = " "

    try:
        opts, args = getopt.getopt(sys.argv[1:], "hps:n", ["help", "print-date", "splitter=", "new-line"])
    except getopt.GetoptError as err:
        print(str(err))
        print_usage()
        sys.exit(2)
    for o, a in opts:
        if o in ("-h", "--help"):
            print_usage()
            sys.exit()
        elif o in ("-p", "--print-date"):
            print_date = True
        elif o in ("-s", "--splitter"):
            splitter = a
            if splitter == "\\n":
                splitter = "\n"
        elif o in ("-n", "--new-line"):
            splitter = "\n"
        else:
            assert False, "invalid option"
    if (not args):
        print('missing date arguments')
        print_usage()
        sys.exit(1)

    date_pattern = re.compile(r'(\d{4})-(\d{2})-(\d{2}) (\d{2}):(\d{2}):(\d{2})')

    since_date_str = subprocess.check_output(['date', '-d', ' '.join(args), '+"%Y-%m-%d %H:%M:%S"']).decode()[1:-2]
    since_date     = datetime(*map(int, date_pattern.match(since_date_str).groups()))

    packages = []

    for log in sorted(glob('/var/log/dpkg.log*'), reverse=True):
        if log.endswith('.gz'):
            with gzip.open(log, 'rt') as f:
                lines = f.read().splitlines()
        else:
            with open(log, 'r') as f:
                lines = f.read().splitlines()

        for line in lines:
            cols = line.split()
            line_date = datetime(*map(int, date_pattern.match(' '.join(cols[:2])).groups()))

            if line_date >= since_date:
                # keep only installed ones
                if cols[2] == 'install':
                    packages.append(cols[3])

    print(splitter.join(uniquify(packages)))

    if print_date:
        print("\n" + since_date.ctime() + "")

if __name__ == '__main__':
    main()