[INFRA] Update the ranges-v3 library to 0.11.0

[marehr]

Fixes seqan/product_backlog#170

Fixes seqan/product_backlog#48

[rrahn]

Thanks 👍

[marehr]

Locally works for:

• gcc-8 -std=c++17 -fconcepts
• gcc-9 -std=c++17 -fconcepts
• gcc-10.1 -std=c++17 -fconcepts
• gcc-10.1 -std=c++2a
• gcc-10.2 -std=c++2a

[eseiler]

Something triggers the visibility bug... https://travis-ci.org/github/seqan/seqan3/jobs/715804696

[marehr]

Something triggers the visibility bug... https://travis-ci.org/github/seqan/seqan3/jobs/715804696

My guess is that the ranges library now uses the short hand syntax 🐌

[marehr]

Yep

#include <seqan3/std/ranges> // submodules/range-v3/include/range/v3/view/split.hpp is the main culprit right now.

  class A{ int i{5}; };
 template <typename t> META_CONCEPT private_bug = requires(t a){a.i;};
 static_assert(!private_bug<A>, "See https://github.com/seqan/seqan3/issues/1317");

Does trigger it.

[eseiler]

So, do we fix it upstream (and maybe hope for an 11.1 release or just use some commit)?
Btw, I couldn't find the syntax that causes the bug...

[marehr]

I could get it reduced to this:

#include <range/v3/iterator/traits.hpp>
#include <range/v3/detail/prologue.hpp>
namespace ranges {
template(typename I)(
    concept(readable_)(I),
    same_as<iter_reference_t<I const>, iter_reference_t<I>> AND
        same_as<iter_rvalue_reference_t<I const>, iter_rvalue_reference_t<I>>
            AND common_reference_with<iter_reference_t<I> &&, iter_value_t<I> &>
                AND common_reference_with<iter_reference_t<I> &&,
                                          iter_rvalue_reference_t<I> &&>
                    AND common_reference_with<iter_rvalue_reference_t<I> &&,
                                              iter_value_t<I> const &>);
template <typename I>
CPP_concept indirectly_readable = CPP_concept_ref(ranges::readable_,
                                                  uncvref_t<I>);
template <typename I1, typename I2>
CPP_requires(indirectly_swappable_,
             requires(I1 const i1, I2 const i2)(ranges::iter_swap(i1, i2),
                                                ranges::iter_swap(i1, i1),
                                                ranges::iter_swap(i2, i2),
                                                ranges::iter_swap(i2, i1)));
template <typename I1, typename I2 = I1>
CPP_concept indirectly_swappable = indirectly_readable<I1> &&
    indirectly_readable<I2> &&CPP_requires_ref(ranges::indirectly_swappable_,
                                               I1, I2);
} // namespace ranges
namespace ranges {
template <typename Rng>
using iterator_t = decltype(begin(std::declval<Rng &>()));
template <typename V, typename Pattern> struct split_view;
namespace detail {
template <typename JoinView, bool Const> struct split_inner_iterator;
template <typename V, typename Pattern, bool Const>
struct split_inner_iterator<split_view<V, Pattern>, Const> {
  using Base = meta::const_if_c<Const, V>;
  CPP_broken_friend_member friend constexpr auto
  iter_swap(split_inner_iterator const &x, split_inner_iterator const &y)
      -> CPP_broken_friend_ret(void)(
          requires indirectly_swappable<iterator_t<Base>>) {}
};
} // namespace detail
} // namespace ranges
class A {
  int i{5};
};
template <typename t> META_CONCEPT private_bug = requires(t a) { a.i; };
static_assert(!private_bug<A>,
              "See https://github.com/seqan/seqan3/issues/1317");

[marehr]

I created a bug fix upstream ericniebler/range-v3#1547.

I'll wait this week, if we have no reaction upstream, I'll temporarily disable the header test.

[eseiler]

Upstream PR was merged

[codecov]

Codecov Report

Merging #2014 into release-3.0.2 will not change coverage.
The diff coverage is n/a.

@@              Coverage Diff               @@
##           release-3.0.2    #2014   +/-   ##
==============================================
  Coverage          97.89%   97.89%           
==============================================
  Files                263      263           
  Lines               9879     9879           
==============================================
  Hits                9671     9671           
  Misses               208      208           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ada8b38...03f1625. Read the comment docs.