Associative array

[Jonathan M Davis]

On Friday, October 13, 2023 5:32:29 AM MDT Imperatorn via Digitalmars-d wrote:
> Hello there.
>
> I'm wondering if anyone would be opposed to us/me making
> associative array having resize/reserve public?
>
> Currently there exists private members in druntime etc for this
> but they are not exposed.
>
> Basically I would like to be able to reserve n buckets before
> populating.
>
> For reference, see std::unordered_map, which is kvp, which has
> rehashing etc:
>
> ```cpp
>    template<typename _Key, typename _Tp,
>      typename _Hash = hash<_Key>,
>      typename _Pred = equal_to<_Key>,
>      typename _Alloc = allocator<std::pair<const _Key, _Tp>>>
>      class unordered_map
>      {
>        typedef __umap_hashtable<_Key, _Tp, _Hash, _Pred, _Alloc>
> _Hashtable;
>        _Hashtable _M_h;
> ```
>
> https://github.com/gcc-mirror/gcc/blob/master/libstdc%2B%2B-v3/include/bits/
> unordered_map.h
>
> https://github.com/gcc-mirror/gcc/blob/master/libstdc%2B%2B-v3/include/bits/
> hashtable.h

My main concern here would be that we not add calls which relate to how AAs
are currently implemented but which may not relate to how they're
implemented in the future if/when we make changes, and I'm inclined to think
that anything to do with buckets is an implementation detail, making
reserving not something that users should really be doing. It _might_ make
sense to have a general reserve function that says that you're planning to
insert X number of elements, allowing the AA to pre-allocate some stuff
based on that information, but I don't know how much sense that really makes
given that what that would mean for a given AA implementation could differ
quite a bit.

If rehashing is what you want, we already have that:

https://dlang.org/spec/hash-map.html#properties

Ultimately though, I think that we want to be careful about how much we
expose of the AA, since that's going to restrict what we can do to improve
it in the future, and if someone wants to do much a tweaking, IMHO, a
user-defined type would make more sense than the language-provided AAs.

Steven would likely have a much better idea of how much sense this idea
makes given that he's been working on the AA stuff and is familiar with some
of the changes that have been made to it over time.

- Jonathan M Davis