Property-based tests for sets, ordsets and gb_sets

[Maria-12648430]

This PR provides property-based tests for the sets module.

It does not test sets:new/0,1, since there isn't much that could be property-tested there.

[github-actions]

CT Test Results

       2 files       90 suites   40m 3s ⏱️
1 920 tests 1 657 ✔️ 263 💤 0 ❌
2 221 runs  1 942 ✔️ 279 💤 0 ❌

Results for commit cc63f8c.

♻️ This comment has been updated with latest results.

To speed up review, make sure that you have read Contributing to Erlang/OTP and that all checks pass.

See the TESTING and DEVELOPMENT HowTo guides for details about how to run test locally.

Artifacts

• Complete CT logs (Download Logs)
• HTML Documentation (Download HTML Docs)
• Windows Installer

// Erlang/OTP Github Action Bot

[Maria-12648430]

If #7232 and/or #7183 get accepted and merged earlier, this PR should be updated to include the functions introduced there.

[juhlig]

If #7232 and/or #7183 get accepted and merged earlier, this PR should be updated to include the functions introduced there.

The equality check functions from #7232 would be pretty handy in this PR I guess 🙂

[Maria-12648430]

The last commit refactors a lot of what has been done before, going for a model-like approach. It also tests all functions on all set implementations (the ones they have in common, that is).

[Maria-12648430]

The last commit does some cleanup (ordering/grouping of functions, comments) and adds another property to test sequences of modifying operations.

[Maria-12648430]

On a side note, this suite makes heavy use of the any() generator (explicit and implicit via the list() generator), which contains the atom() generator. The atom() generator produces atoms at random, which can become a problem re the atom limit.

[Maria-12648430]

I marked the PR as WIP, as it still needs some touching up :)

[Maria-12648430]

Last commit adds tests for the new functions map/2, filtermap/2 and is_equal/2.

[Maria-12648430]

As it stands, this suite alone generates ~800,000 new atoms. Waiting for #7364 which should remedy this.

[Maria-12648430]

The last commit uses the atomlimit-safe generators of ct_proper_ext, provided by #7364.

[bjorng]

Please squash your commits.

[Maria-12648430]

Please squash your commits.

Done 😉

[bjorng]

Thanks! Added to our daily builds.

[bjorng]

I will now go on vacation and probably be back in about four weeks. @jhogberg will take care of merging this pull request if no problems turn up in our daily builds.

[Maria-12648430]

I will now go on vacation

Enjoy 👋

@jhogberg will take care of merging this pull request if no problems turn up in our daily builds.

And hi John 🤝

[Maria-12648430]

Rebased on current master to make use of #7495.

[Maria-12648430]

@jhogberg @bjorng I wonder, what is the status of this PR? I just now noticed that the testing label was removed shortly after I pushed the last update 🤔

On another note, I realized that the tests in this suite cover only the functions that the three set implementations have in common, but gb_sets has some additional functions that the other two don't have, like singleton, delete, smallest, take_largest, iterator and some more. Should I add them to this suite (like in sets_SUITE) or put up another property test suite for the additional functions of gb_sets), what do you think?

[jhogberg]

@jhogberg @bjorng I wonder, what is the status of this PR? I just now noticed that the testing label was removed shortly after I pushed the last update 🤔

The status is that it fell between the cracks. 😅

I'll merge this in the next few days if the tests pass. :)

On another note, I realized that the tests in this suite cover only the functions that the three set implementations have in common, but gb_sets has some additional functions that the other two don't have, like singleton, delete, smallest, take_largest, iterator and some more. Should I add them to this suite (like in sets_SUITE) or put up another property test suite for the additional functions of gb_sets), what do you think?

Feel free to do the latter in a separate PR.

[Maria-12648430]

The status is that it fell between the cracks. 😅

Ah, no problem 😉

Feel free to do the latter in a separate PR.

Will do 👍

[Maria-12648430]

@bjorng so, with @jhogberg going on vacation, it seems that this PR has come back to haunt you 😁

I noticed that I forgot the copyright header and added it in the last commit.

[bjorng]

Thanks for your pull request.