enhancement: select crawl by exact timestamp

[laurieburchell]

Querying the index brings back a status, timestamp, url triple, e.g.:

$ cdxt --cc --crawl CC-MAIN-2025-43 iter 'commoncrawl.org/get-started'  

status 200, timestamp 20251014220259, url https://www.commoncrawl.org/get-started
status 200, timestamp 20251016192109, url https://commoncrawl.org/get-started

It would be good to have direct method to bring back a particular record based on the timestamp alone. I'm aware you can do something like cdxt --cc --crawl CC-MAIN-2025-43 --from 20251016192109 --limit 1 warc 'commoncrawl.org/get-started' but a direct --timestamp flag or similar would be useful, given the presentation of the index records.

[wumpus]

I'm in favor of this, the iter command returns a timestamp and it ought to be directly usable by the user.

You'll note that whirlwind set both --from and --to as a way to get an exact timestamp. That's me being too clever for my own good 😅

[wumpus]

Given that timestamps in the archive world are allowed to be shorter than an exact second, the safe way to implement this is literally to turn timestamp into "from" and "to" and warn in the docs that more than one record might be returned. The reason is that "from" rounds down and "to" rounds up.

@malteos I'd put this into the main branch.