Computed Constant feature request

[David Alayachew]

Hello Per and Maurizio,

Thank you both for your response!

Your Map<K, ComputedConstant<LargeList<MyResource>>> sounds like exactly
what I am looking for Per! I understand if it does not come out in this JEP
specifically. But that is a 1-to-1 match with what I am looking for!

And what a smart idea Maurizio. Perhaps I could break apart my
implementation, and then make use of the existing ComputedConstant -> List
feature to load each element from the LargeList lazily. I believe that is
what you were leading to Maurizio?

Thank you both for your time and help!
David Alayachew

On Mon, Aug 21, 2023 at 6:31 AM Maurizio Cimadamore <
maurizio.cimadamore at oracle.com> wrote:

> In addition to what Per says, you might also consider modelling your
> "large list of resources" as a lazy list? While you still have the problem
> of mapping enum values to ordinals, you would at least retain the list-like
> nature of the bag of constants.
>
> Maurizio
> On 20/08/2023 00:36, David Alayachew wrote:
>
> Hello Leyden Dev Team,
>
> I am very excited for the ComputedConstant JEP that has been recently
> announced. There are lots of places in my current projects that could use a
> feature like this.
>
> Today, for example, I want to lazy load a gigantic list of resources
> (>100). However, these resources are not fetched via an index, they are
> fetched via a key.
>
> My current data type that I am storing them into is essentially a
> Map<MyEnum, LargeList<MyResource>>. I would like to be able to swap out
> this map for a ComputedConstant that can give me my LargeList<MyResource>
> when I provide it an instance of MyEnum.
>
> I could technically create a mapper that takes in MyEnum, turns it into
> its ordinal, and then does something like this.
>
> ComputedConstant
>     .of
>     (
>         MyEnum.values().length,
>         i -> fetchLargeListOfMyResource(MyEnum.values()[i])
>     )
>     ;
>
> But I'd like to avoid that if at all possible. It feels like I would end
> up right back into that error-prone territory from before.
>
> Is this feature possible and worth doing as part of this JEP?
>
> Thank you for your time and help!
> David Alayachew
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/leyden-dev/attachments/20230821/6ba7ec18/attachment.htm>