field using types turned into array when declared(params)

[braktar]

It seems that the recent works on declared change the interaction between 'declared' and the 'types' field

[grape-bot]

	1 Warning
⚠️	Unless you’re refactoring existing code, please update CHANGELOG.md.

Here's an example of a CHANGELOG.md entry:

* [#2112](https://github.com/ruby-grape/grape/pull/2112): Field using type turned into array when declared(params) - [@braktar](https://github.com/braktar).

Generated by 🚫 danger

[dblock]

what was the behavior in 1.3.x, 1.4.x?

either way looks like a bug, another one... /cc: @tlconnor

[dblock]

Related #2103

[braktar]

Before the integration of #2103 the spec passed successfully

[braktar]

Here is a fix which avoid the misinterpretation of the type

[braktar]

That's strange that the type of empty_typed_arr turned from Array[String] to [String] when arrived into handle_passed_param

[dblock]

Let's extend tests with Hash and Set along with Array?

[tlconnor]

You are right @dblock that working with Strings here is not ideal. I don't know enough to know if there is a better alternative.

[braktar]

Let's extend tests with Hash and Set along with Array?

I don't really see what should exhibit these tests, as the only ambiguity is between types : [x, y] and type: Array[x, y] which both give type: [x, y] while processed.

It seems to require a deeper investigation

[dblock]

Let's extend tests with Hash and Set along with Array?

I don't really see what should exhibit these tests, as the only ambiguity is between types : [x, y] and type: Array[x, y] which both give type: [x, y] while processed.

It seems to require a deeper investigation

I mean what happens if requires :bar, type: Hash do ...?

[dblock]

@stanhu isn't this the same?

[stanhu]

Yes, it's the same. I just picked the commit in #2117.