Skip to content

Roger issues 21.12 #10

@camio

Description

@camio

21.12.2 [reflect.synopsis]

The ObjectSequence operations are written with the template parameter named
'S' but all the others use 'T'

s/S>/T>/g in this part:

template struct get_size;
template <size_t I, ObjectSequence S> struct get_element;
template <template <class...> class Tpl, ObjectSequence S>
struct unpack_sequence;

template
constexpr auto get_size_v = get_size::value;
template <size_t I, ObjectSequence S>
using get_element_t = typename get_element<I, S>::type;
template <template <class...> class Tpl, ObjectSequence S>
constexpr auto unpack_sequence_t = unpack_sequence<Tpl, S>::type;

» DJS: TODO

21.12.4 [reflect.ops]

"x_t will even satisfy Class."

s/even //
(I don't think a replacement for 'even' is required.)

AN: Class is a refinement. Without "even" this seems to contradict the
specification: "get_type_t is specified to be a Type, but here it's a
Class, surprise".
RO: How about changing 'even satisfy' to 'also satisfies' ?
DJS: TODO

21.12.4.7 [reflect.ops.record]

The order of (9.1), (9.2), (9.3) does not match the order at the head of p9
The order of (12.1), (12.2), (12.3) does not match the order at the head of
10

» DJS: TODO

21.12.4.16 [reflect.ops.expr.fcttypeconv]

template struct get_constructor;
1 All specializations of get_converting_callable

I presume get_converting_callable was an earlier name for this?

» DJS: TODO

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions