get rid of the ::value in include_summand<propto, T1, ..., TN>::value

[mcol]

Description

This was suggested by @bob-carpenter in #1520 (comment):

It would also be nice to get rid of the ::value using the same trick as the standard library uses for enable_if_t with some parameter pack magic for good measure.

template <bool propto, typename ...Ts>
using include_summand_b = include_summand<propto, Ts...>::value;

Then the usage could go from include_summand<propto, T1, ..., TN>::value to
include_summand_v<propto, T1, ..., TN>.

We settled on using _v because the result is a value, rather than the _t for a type in enable_if_t.

[mcol]

I could not get @bob-carpenter's suggestion to compile, but the following works well and existing tests pass:

template <bool propto, typename... T_pack>
constexpr bool include_summand_b = include_summand<propto, T_pack...>::value;

@SteveBronder Is this the correct approach? If so I can post a PR.

[rok-cesnovar]

Yes this is it. Thank you!

[SteveBronder]

Yep!

[bob-carpenter]

Thanks---I didn't try it so was sure to mess up the ... in the parameter packs. Now looking at other types like enabl_if, they use _v for "value", so we should probably stick to that convention.

…

On Dec 16, 2019, at 4:32 AM, Marco Colombo ***@***.***> wrote: I could not get @bob-carpenter's suggestion to compile, but the following works well and existing tests pass: template <bool propto, typename... T_pack> constexpr bool include_summand_b = include_summand<propto, T_pack...>::value; @SteveBronder Is this the correct approach? If so I can post a PR. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[mcol]

This has to wait until we move our Windows compiler to RTools 4.0, as RTools 3.5 has a bug in the constexpr implementation that doesn't allow to compile the expression we need.