|
Boost : |
Subject: Re: [boost] Is there any interest in type-safe container of bool flags with noexcept guarantees?
From: Roman Orlov (compmaniak_at_[hidden])
Date: 2017-03-23 20:10:53
On 23.03.2017 01:40, Andrey Semashev via Boost wrote:
> Usually, I solve this problem with std::bitset with an enum providing
> names for the flags.
>
> enum animal_traits
> {
> eats_meat, eats_grass, has_tail, _count
> };
> typedef std::bitset< animal_traits::_count > animal;
>
> animal a1; // all traits are false
> a1[eats_grass] = 1;
> a1[has_tail] = 1;
>
> I think, this provides the necessary expressiveness and performance. I
> guess, your proposal provides better type safety than this solution, but
> I'm not sure the improvement is significant enough to outweigh the more
> complicated syntax. For example, one could write a slightly more
> complicated version of the above:
>
> enum animal_traits
> {
> eats_meat, eats_grass, has_tail, _count
> };
>
> template< typename Enum, unsigned int Count >
> class typed_bitset :
> public std::bitset< Count >
> {
> public:
> bool operator[] (Enum idx) const
> {
> return std::bitset< Count >::operator[](idx);
> }
> };
>
> typedef typed_bitset< animal_traits, animal_traits::_count > animal;
>
> animal a1; // all traits are false
> a1[eats_grass] = 1;
> a1[has_tail] = 1;
> a1[10] = 1; // error
>
> I suspect it would also be faster to compile.
>
Using std::bitset in this way is not a good solution.
C++ standard doesn't claim that bitset's backing storage should be as
compact as possible. In stdlibc++ bits are stored in array of
unsigned long items. So on my Ubuntu x64
sizeof(typed_bitset< animal_traits, animal_traits::_count >) == 8
That's not acceptable for only 3 bits.
Another problem is that member functions for bit manipulations are not
marked as *noexcept*. Using weak enum-based parameterization you can
be sure in proper use bits. But the compiler doesn't know it.
That's why I've come to making my own container. I need compact storage
and noexcept guarantee at compile time.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk