On OpenBSD string_view is both in boost and std

[omoerbeek]

Short description

Checklist

I have:

• read the CONTRIBUTING.md document
• compiled this code
• tested this code
• included documentation (including possible behaviour changes)
• documented the code
• added or modified regression test(s)
• added or modified unit test(s)
• checked that this code was merged to master

[rgacogne]

Well, we really would like to use std::string_view when available.. How does lmdb-safe.hh compiles on OpenBSD? Because the string_view detection is a copy-paste from that code..

[omoerbeek]

I normally don't build lmdb support on OpenBSD. But indeed, it gives similar errors.

[RalfvdEnden]

FreeBSD suffers from the same issue. After adding && !defined(__FreeBSD) to the #if line in the patch it builds fine.

[Habbie]

NetBSD 9 does not need this.

[omoerbeek]

lmdb-safe.cc build on OpenBSD now, but warns:

  CXX      ext/lmdb-safe/lmdb-safe.o
ext/lmdb-safe/lmdb-safe.cc:338:10: warning: moving a temporary object prevents copy elision [-Wpessimizing-move]
  return std::move(getRWTransaction());
         ^
ext/lmdb-safe/lmdb-safe.cc:338:10: note: remove std::move call here
  return std::move(getRWTransaction());
         ^~~~~~~~~~                  ~
1 warning generated.

[rgacogne]

lmdb-safe.cc build on OpenBSD now, but warns:
[...]

clang++ does warn on Arch as well. AFAICT the std::move() should go as it prevents RVO.

[omoerbeek]

OK, IMO we need a separate PR for that.

[rgacogne]

I'm OK with the fix for now, but that clearly indicates that we have an issue on some systems where we use C++17's features even though we only require C++11. I hope it will not come back to haunt us later :-/