It seems like the Kotest intellij plugin's test browser ignores the context blocks in DescribeSpec.....

# kotest

[dave08]

It seems like the Kotest intellij plugin's test browser ignores the context blocks in DescribeSpec...

[wasyl]

Also it ignores context blocks when determining if tests have the same name

[dave08]

Yup, that's the first thing I noticed... but it's probably related.

[sam]

The context thing is fixed in the latest update (pending approval with Jetbrains)

[sam]

@wasyl in 4.5 duplicated names get the (1) (2) treatment like data tests

[wasyl]

Just for clarity, the duplicated test warning is a false positive as the plugin doesn’t understand context blocks. So it should also be fixed after the update 🙂

[dave08]

So the IDE won't complain anymore for real duplicates in 4.5?

[sam]

correct @dave08. Do you think that's an issue? Duplicated test names seemed an edge case to me - we errored because junit would just "hide them" (which is awful)

[sam]

In 4.5, I made it append (1) or (2) and so on, so they have a unique name

[sam]

That's required for data tests where you re-use parameters, but I could add an option to error on duplicated names as before if we think that's valuable

[dave08]

But sometimes a test is copy-pasted, the inners are changed and the name isn't 😛... so the IDE complaining is good in a sense...

[sam]

Yeah, but it just means your test output would be "test I copied" and "test I copied (1)"

[sam]

4.5 isn't released yet so I can still add the option in if we want

[dave08]

Maybe if the (1) was a prefix, it would be more noticeable...

[sam]

yes I like that as well

[dave08]

And possibly an IDE warning?

A setting might be a plus though, because some (like me), never use duplicate names and prefer the IDE and test crashing to tell them...

The only occasion where I came across doubles was with table() tests that the only difference was the number of commas in the name...

[sam]

We'll do this before 4.5 final https://github.com/kotest/kotest/issues/2229