Use the Wayback Machine API to search old websites.

wayback.py

#!/usr/bin/env python3
# -*- coding: utf8 -*-

from argparse import ArgumentTypeError, ArgumentParser
from collections import Counter
from datetime import datetime, timezone
from pathlib import Path
from re import search

from archivecdx import Listing as ArchiveListing


def main():
    def input_dt(s):
        try:
            dt = datetime.fromisoformat(s)
        except ValueError:
            raise ArgumentTypeError(f'invalid date: {s}')
        if dt.tzinfo:
            return dt
        else:
            return dt.astimezone()
    parser = ArgumentParser(
        description='Resurrect websites from the Wayback Machine.')
    parser.add_argument('url', help='URL parameter to search')
    parser.add_argument('-matchType', default='exact',
        choices=['exact', 'prefix', 'host', 'domain'],
        help='select scope for search')
    parser.add_argument('-from', dest='tfrom', type=input_dt,
        help='start search from this date')
    parser.add_argument('-to', dest='tto', type=input_dt,
        help='end search at this date')
    parser.add_argument('-filter', action='append',
        help='apply a search filter; add multiple for AND')
    parser.add_argument('-collapse', action='append',
        help='apply a collapse filter; add multiple for AND')
    args = parser.parse_args()

    def wayback_time(dt): return dt.astimezone(timezone.utc).strftime('%Y%m%d%H%M%S')
    download_listing(ArchiveListing(
        args.url,
        matchType=args.matchType,
        filter=args.filter if args.filter else [],
        collapse=args.collapse if args.collapse else [],
        _from=wayback_time(args.tfrom) if args.tfrom else None,
        to=wayback_time(args.tto) if args.tto else None))


def download_listing(listing):
    def urlkey_path(urlkey):
        parts = Path(urlkey).parts
        def domain(part):
            m = search(r'(.+?)(?::(\d+))?\)$', part)
            if not m:
                raise ValueError(f'invalid domain: {part}')
            domainl = m.group(1).split(',')
            domainl.reverse()
            domain = '.'.join(domainl)
            port = m.group(2)
            return f'{domain}:{port}' if port else domain
        return Path(*((domain(parts[0]),) + parts[1:]))
    def strip_illegals(path):
        def strip(s):
            return s.replace(':', '_')
        return Path(*(strip(p) for p in path.parts))
    def row_path(row, isdupe):
        if isdupe[row.urlkey]:
            parts = urlkey_path(row.urlkey).parts
            fsplit = parts[-1].split('.')
            if len(fsplit) > 1:
                fname = (f'{fsplit[0]}_{row.timestamp}_{row.digest}'
                         f".{'.'.join(fsplit[1:])}")
            else:
                fname = f'{fsplit[0]}_{row.timestamp}_{row.digest}'
            return strip_illegals(Path(*(parts[:-1] + (fname,))))
        else:
            return strip_illegals(urlkey_path(row.urlkey))

    urlkey_dupes = Counter(row.urlkey for row in listing)
    urlkey_isdupe = {urlkey: True if urlkey_dupes[urlkey] > 1 else False
                     for urlkey in urlkey_dupes}
    for row in listing:
        out_path = row_path(row, urlkey_isdupe)
        print(f'https://web.archive.org/web/{row.timestamp}id_/{row.original}')
        print(f'\tout={out_path}')


if __name__ == '__main__':
    main()