On Wed, Aug 13, 2025 at 9:40 PM Patrick Palka <ppa...@redhat.com> wrote:
> On Wed, 13 Aug 2025, Patrick Palka wrote:
>
> > Thanks for the patch! Looks good to me for the most part.
> >
> > On Fri, 8 Aug 2025, Yihan Wang wrote:
> >
> > > libstdc++-v3/ChangeLog:
> > >
> > > * include/std/expected:
> >
> > This ChangeLog entry should be filled in, e.g.
> >
> > * include/std/expected (expected::expected(_Up&&)): Add
> > missing constraint as per LWG 4222.
> >
> > > * testsuite/20_util/expected/lwg4222.cc: New test.
> > >
> > > Signed-off-by: Yihan Wang <yronglin...@gmail.com>
> > > ---
> > > libstdc++-v3/include/std/expected | 1 +
> > > .../testsuite/20_util/expected/lwg4222.cc | 15 +++++++++++++++
> > > 2 files changed, 16 insertions(+)
> > > create mode 100644 libstdc++-v3/testsuite/20_util/expected/lwg4222.cc
> > >
> > > diff --git a/libstdc++-v3/include/std/expected
> b/libstdc++-v3/include/std/expected
> > > index 60f1565f15b..2b200ea0589 100644
> > > --- a/libstdc++-v3/include/std/expected
> > > +++ b/libstdc++-v3/include/std/expected
> > > @@ -474,6 +474,7 @@ namespace __expected
> > > template<typename _Up = remove_cv_t<_Tp>>
> > > requires (!is_same_v<remove_cvref_t<_Up>, expected>)
> > > && (!is_same_v<remove_cvref_t<_Up>, in_place_t>)
> > > + && (!is_same_v<remove_cvref_t<_Up>, unexpect_t>)
> >
> > Seems this line is overly indented causing it to not be aligned with the
> rest.
> >
> > > && is_constructible_v<_Tp, _Up>
> > > && (!__expected::__is_unexpected<remove_cvref_t<_Up>>)
> > > && __expected::__not_constructing_bool_from_expected<_Tp, _Up>
> > > diff --git a/libstdc++-v3/testsuite/20_util/expected/lwg4222.cc
> b/libstdc++-v3/testsuite/20_util/expected/lwg4222.cc
> > > new file mode 100644
> > > index 00000000000..a260cfef3dd
> > > --- /dev/null
> > > +++ b/libstdc++-v3/testsuite/20_util/expected/lwg4222.cc
> > > @@ -0,0 +1,15 @@
> > > +// { dg-do compile { target c++23 } }
> > > +
> > > +// LWG 4222. 'expected' constructor from a single value missing a
> constraint
> > > +
> > > +#include <expected>
> > > +#include <type_traits>
> > > +
> > > +struct T {
> > > + explicit T(auto) {}
> > > +};
> > > +struct E {
> > > + E(int) {}
> > > +};
> > > +
> > > +static_assert(!std::is_constructible_v<std::expected<T, E>,
> std::unexpect_t>);
>
> Maybe we should also test constructing from const unexpect_t and
> unexpect_t& and variations thereof.
>
I agree. I would like to also add a test using inplace constructor to
construct
wrapped type from expected. Something like:
struct V {
explicit V(std:unexpect_t) {}
};
std::expected<V, int> e1(std::inplace, std::unexcept);
VERIFY( e1.has_value() );
std::expected<int, V> e2(std::unexcept, std::unexcept);
VERIFY( !e2.has_value() );
> > > --
> > > 2.39.5
> > >
> > >
> >
>
>
