https://git.io/cursive-issues

https://github.com/cursive-ide/cursive/issues/2844

is that the same place?

Yep

Aren’t the tests in the case form supposed to be constants?

I think you should be using the cond form in this case.

> Takes an expression, and a set of clauses. > Each clause can take the form of either: > test-constant result-expr

@UDCGPTV9R https://clojuredocs.org/clojure.core/case#example-542692ccc026201cdc326c89

Or perhaps a more focused one https://clojuredocs.org/clojure.core/case#example-61fc0a53e4b0b1e3652d75a6

Oh, so this is trying to match on a form. I see.

no the semantics of case is that the matching forms can be a constant object to match or a list of constant objects to match. In my case I provide a singleton list of constant object(s). The code works correctly, and works as documented.

the code works as intended.

an important distinction between cond and case` is that cond evaluates the forms to obtain a boolean, while case does not evaluate the forms. so the static analizer should not consider a list in this position as a function call, but rather as list of constants.

Interesting! The https://github.com/clojure/clojure/blob/master/src/clj/clojure/core.clj#L6779 say: "The test-constants are not evaluated. They must be compile-time literals, and need not be quoted" So this works:

(let [x 'foo]
    (case x
      (foo) :yo))

(let [x 'foo]
    (case x
      ('foo) :yo))

@U052XLL3A That’s because reader macros are expanded, but the resulting form is then not evaluated. So you’re trying to match on ((quote foo)) there. If you pass '(quote foo) as x that will work.

But yes, this bug is a result of the fact that Cursive doesn’t understand that some parts of some macros (`case`, but also binding forms in e.g. let) aren’t evaluated. I’ll try to fix that.