Navigating tradeoffs in bidirectional filter design

[monkyyy]

On Friday, 10 January 2025 at 02:47:24 UTC, Salih Dincer wrote:
> On Thursday, 9 January 2025 at 21:56:59 UTC, monkyyy wrote:
>> Im aware phoboes takes ....more complexity to implement filter 
>> and there should be *some* extra complexity to make a 
>> bidirectional filter, but when I ussally look into phoboes 
>> 1000 line functions I usually make different tradeoffs. What 
>> are the actual considerations here?
>
> I don't really understand, what needs to be considerations? 
> Isn't the interface below sufficient?
>
> ```d
> auto filter(alias F,R)(R r)
> {
>   struct Filter
>   {
>     R r;
>     auto back()     => r.back;
>     auto front()    => r.front;
>     auto popFront() => r = r.findNext!F;
>     auto popBack()  => r = r.findPrevious!F;
>     auto empty()    => r.empty;
>     auto save()     => r.save;
>   }
>   return Filter(r.find!F);
> }
> ```
>
> SDB at 79

Well:
1) `[1,2,3,4,5].Filter!(a=>a==3).back` should equal 3, in yours 
it will equal 5
2) your no longer supporting unidirectional ranges

And that only what I notice without any testing; I feel worse 
about being eager for back then front