Ask HN: How would you improve this bash oneliner for deleting tweets?

[jamiehall]

Many people use tweet deletion services, which periodically remove everything from their Twitter timeline; I wondered if it could be done from a Bash command line.

I wrote up my experiences as an explainer for nontechnical people: https://jamiehall.cc/2020/03/10/delete-all-your-tweets-with-...

TL;DR, here is the oneliner I've been using:

  $ twurl "/1.1/statuses/user_timeline.json?screen_name=YOUR_TWITTER_HANDLE&count=200
    &max_id=$(
      twurl '/1.1/statuses/user_timeline.json?screen_name=YOUR_TWITTER_HANDLE&count=200&include_rts=1'
     | jq -c -r '.[] | .id_str' | head -10 | tail -1)
    &include_rts=1"
   | jq -c -r '.[] | .id_str' 
   | parallel -j 10 -a - twurl -X POST /1.1/statuses/destroy/{1}.json
   > /dev/null

[Edit: I've put line breaks in there to make it more legible.]

I'm curious if it's possible to do better. In particular: could this be more elegant? Is it possible to do it using common built-ins, instead of twurl and jq?

Any suggestions or improvements would be very welcome!

[rhacker]

I feel like this is normally the content you see in Stack Overflow

[ollien]

Probably best for the code review StackExchange forum, actually.

[rococode]

Might be pretty fun for the code golf stack exchange too!

[toomuchtodo]

You might consider putting this in a GitHub Gist, sharing the link for it, and accept comments and improvements within the Gist comments functionality.

This would also allow others to fork your code to improve upon or keep a copy for themselves.

[jamiehall]

Great idea, thanks! https://gist.github.com/nerdcha/3e811ef28d04561f73337e46267a...

[toomuchtodo]

Did you get the feedback you were looking for?

[benburleson]

(Security risks aside,)

That can even allow people to curl and pipe to run in the terminal direct from github.

[Pirate-of-SV]

    jq -c -r '.[] | .id_str'
    # Can be rewritten to
    jq -r '.[].id_str'

    jq -c -r '.[] | .id_str' | head -10 | tail -1
    # Can be rewritten to
    jq -r '.[9].id_str'

[jamiehall]

Ah, nice. Thanks! :D

[t0astbread]

What are the rate limits on the /statuses/destroy endpoint? I've checked the docs and the docs say it is limited but the actual limit is not specified.

Other than that, thanks for writing this! I've been thinking about a tool like this and this might come in handy. The code looks fine to me although I would probably spin this out into a script and add some logging, as others have already pointed out.

[viraptor]

I don't think there's a good reason to calculate the max_id every time. If you want to delete all tweets, you could skip it.

[a-wu]

I've tried this before in a Python script but quickly got rate limited by the 300 tweet update limit [0]. Does this get around that?

[0] https://developer.twitter.com/en/docs/basics/rate-limits

[darepublic]

It's hard to imagine that it would

[jamiehall]

Thank you to everybody who commented! The suggestions to use something more sensible than a oneliner (and to post on Stack instead) are well taken. A big motivation of this project was "just because", for shits and giggles, so I hope it's amused you for a minute or two. Cheers!

[jlelse]

Not answering your question, but I found a way to mass delete tweets without creating an app: https://jlelse.blog/posts/mass-delete-tweets/

[mirimir]

Not a Twitter user here, but would this delete retweets by others of your tweets?

Those might include tweets that you'd most want to delete. As well as ones that you'd most want to retain, for that matter.

[jamiehall]

Yes, it takes care of that; if someone has retweeted you with a comment, the thing they're commenting on is replaced with "This tweet is no longer available." (Obviously, it's no defence against people screenshotting your tweet, manually copy-pasting what you said, etc etc.)

[ksherlock]

You could improve it by not tweeting in the first place.

[cosmiccatnap]

Make it delete the account instead.

[staktrace]

You could improve it by making it not a one-liner! Usually making it a multiple-liner can improve readability. And unless you're typing it into the command line directly everytime you run it (as opposed to putting it in a file that you invoke) it doesn't really matter how many lines it has.

[Normal_gaussian]

It's not just readability.

One day the code breaks. For a one liner I have to pick it apart, add logging / diagnostics, fix it, and reassemble. The two most time consuming and error prone parts of that are dealing with the 'one liner'

[wespiser_2018]

I came here to say this. In addition, searching through bash history is a terrible "save and fetch" interface