[slot_map] Can you get a key using an iterator? about sg14

[Quuxplusone]

@p-groarke: If you have "many features" you'd like to add, I'd recommend forking the repo, adding all the features in a branch, and then submitting the whole thing as a big patch so we can see how all the new things work together. Bugfixes = incremental, design work = holistic. :) (#135 is certainly related, if we're doing design work.)

FWIW, I haven't seen any comments on this repo from @Masstronaut (Allan Deutsch) for a long time. I myself am the original author of this repo's "slot_map.h", which I did just by slavishly following Allan's P0661R0. Allan has his own slot_map at https://github.com/Masstronaut/slot_array .

I see there is also https://github.com/twiggler/slotmap . I don't know anything about it except that it has a lot of heavy dependencies: std::random_device, Boost.Integer, Boost.Hana,...

[p-groarke]

OK I believe that is provided by slot_iter_from_value_iter, it is simply marked private.

[Quuxplusone]

This, like #131, sounds like a legitimate design issue. @Masstronaut, are you still interested in slot_map? Are these features good ideas, or did you consider them and then exclude them from the slot_map proposal for reasons that it would be worthwhile to recap here?

@p-groarke: If the functionality you think-you-need is there-but-private, then this issue should still be open, right? Either we should make the functionality public, or we should explain why you don't really need it, or we should explain why if you do need it you shouldn't be using slot_map.

My kneejerk reaction is that we should make it public, which means bikeshedding the name/API. Like,

slot_map<K, V> sm;
std::vector<V>::iterator it = sm.begin();
K k = sm.key(it);  // new functionality -- is this a good API?
assert(sm.find(k) == it);

[p-groarke]

@Quuxplusone Fair point, until we get an answer from the author and windows CI, my hands are a little tied. I have many features I'd like to add. I'm happy to send a PR for this one if the original author is on board / doesn't have objections.

[p-groarke]

@Quuxplusone Sounds good to me. Will continue to gain some experience with the container in the meantime. I personally really like the name you proposed sm.key(it). It is short and to-the-point. I have some general questions, would you rather I open another issue or email you with them?