[elm] fix code generation for nullable types

[KMahoney]

PR checklist

• Read the contribution guidelines.
• Ran the shell script under ./bin/ to update Petstore sample so that CIs can verify the change. (For instance, only need to run ./bin/{LANG}-petstore.sh, ./bin/openapi3/{LANG}-petstore.sh if updating the {LANG} (e.g. php, ruby, python, etc) code generator or {LANG} client's mustache templates). Windows batch files can be found in .\bin\windows\. If contributing template-only or documentation-only changes which will change sample output, be sure to build the project first.
• Filed the PR against the correct branch: master, 4.1.x, 5.0.x. Default: master.
• Copied the technical committee to review the pull request if your PR is targeting a particular programming language.

Description of the PR

The code generation for nullable types put a parenthesis in the wrong place and added a default value on to the encoder. This fixes both issues.
@eriktim

[auto-labeler]

👍 Thanks for opening this issue!
🏷 I have applied any labels matching special text in your issue.

The team will review the labels and make any necessary changes.

[eriktim]

Thanks for the PR @KMahoney. Can you please provide a small example OpenAPI spec showing the issue.

[KMahoney]

Sure: https://gist.github.com/KMahoney/937b9b7910ebb8fdb7288bcb8ec6b7e3
On master, produces:

decoder : Decoder Example
decoder =
    Decode.succeed Example
        |> required "example" (Decode.nullable Decode.int) Nothing



encode : Example -> Encode.Value
encode model =
    Encode.object
        [ ( "example", Maybe.withDefault Encode.null (Maybe.map Encode.int) model.example )

        ]

[eriktim]

Thanks for the example. Your fix indeed repairs this. However, it breaks the case when the field is not required but nullable. Can you try surrounding the {{{defaultValue}}} you removed with {{^required}}..{{/required}}?

P.S. I am working on a better test spec file, so I've added this case as well.

[KMahoney]

@eriktim Ok, I think that does it. I updated the gist with a more comprehensive test.

[eriktim]

@KMahoney looks good now. Thanks for fixing this!

@wing328 can you poke the CircleCI build?

[wing328]

CircleCI failure not related to this PR.

[wing328]

@KMahoney thanks for the PR, which has been included in v4.2.0 release: https://twitter.com/oas_generator/status/1189824932345069569