Hey Kotesters! I struggle a lot with the different...

# kotest

[Olaf Gottschalk]

Hey Kotesters!
I struggle a lot with the different concepts of the two functions:

shouldHaveSingleElement(element)
shouldHaveSingleElement { block... }

Every time I want to use them, I have to go back to documentation and/or code to see for myself that the first one, which takes an element actually does assert that the collection contains just a single element and the latter, which takes a predicate, does not do that, instead it only checks the collection to contain one single element that passes the predicate, but could potentially contain many more elements.

The documentation piece I found here looks like this:
collection.shouldHaveSingleElement(element) Asserts that the collection only contains a single element and that that element is the given one.
collection.shouldHaveSingleElement { block } Asserts that the collection contains a single element by a given predicate.

This is very error prone as most of the time my assumption is I need a check that the collection contains a single element and the next step is only to either be able to provide the element for equals test OR a predicate to check some aspects of the element - but never I expect different behaviour in the main aspect of the function.

Can somebody explain to my why we have this and whether it really is intentionally? Maybe I don't see it...
Thanks!

[Klitos Kyriacou]

May I confuse you even further 😈 with:
collection.shouldBeSingleton()
collection.shouldBeSingleton { block }

[Olaf Gottschalk]

@Klitos Kyriacou Please, don't! 😉 Reason I did not even see shouldBeSingleton right next to shouldHaveSingleElement is that it lacks its infix modifier! I am so used to using functions with only a predicate lambda using infix style, I only type a space and look at possible completions... So, why does shouldHaveSingleElement have infix, but shouldBeSingleton does not? Why does shouldBeEmpty work on nullable collections/iterables whereas these functions require an additional shouldNotBeNull check? ...so many questions...

[Olaf Gottschalk]

I created the necessary changes for the infix modifier. But changing the behaviour is something that should be discussed...

[Alex Kuznetsov]

which name(s) would be less confusing than shouldHaveSingleElement?

[Olaf Gottschalk]

It's not about the name. It's about the fact that all functions named identical should behave identically. Name is fine, but seeing that there is shouldBeSingleton as well, maybe shouldHaveSingleElement should NOT at all be about singleton check, but only about assuring that one of the elements matches.

[Alex Kuznetsov]

I'd suggest renaming
shouldHaveSingleElement { block... }
to
shouldHaveSingleElementMatching { block... }
and deprecate the old one.

[Olaf Gottschalk]

Yes, I agree. I can propose a PR, I have already started to rework this...

[Alex Kuznetsov]

regarding "maybe shouldHaveSingleElement should NOT at all be about singleton check, but only about assuring that one of the elements matches" - do we need that function at all?

shouldHaveSingleElementMatching { it == ... }

should do it, right?

[Alex Kuznetsov]

makes sense, feel free to cut a PR

[Alex Kuznetsov]

and maybe collection.shouldBeSingleton { block } needs some attention as well? in a separate PR