[cursecheck] fix build for gcc-13.3.1

[myk002]

current code (with a vector being constructed inline in a for loop initializer) produces the following error (for me, at least; seems to pass fine in CI)

FAILED: plugins/CMakeFiles/cursecheck.dir/cursecheck.cpp.o 
/usr/bin/ccache /usr/lib/ccache/bin/g++ -DDFHACK64 -DHAVE_CUCHAR -DLINUX_BUILD -DLUA_BUILD_AS_DLL -DPROTOBUF_USE_DLLS -D_LINUX -Dcursecheck_EXPORTS -I/home/myk/src/dfhack/depends/SDL2/SDL2-2.26.2/include -I/home/myk/src/dfhack/depends/protobuf -I/home/myk/src/dfhack/depends/lua/include -I/home/myk/src/dfhack/depends/md5 -I/home/myk/src/dfhack/depends/tinyxml -I/home/myk/src/dfhack/depends/lodepng -I/home/myk/src/dfhack/depends/clsocket/src -I/home/myk/src/dfhack/depends/xlsxio/include -I/home/myk/src/dfhack/library/include -I/home/myk/src/dfhack/library/proto -I/home/myk/src/dfhack/plugins/proto -I/home/myk/src/dfhack/library/depends/xgetopt -fvisibility=hidden -mtune=generic -Wall -Werror -Wl,--disable-new-dtags -Wno-unknown-pragmas -m64 -mno-avx -Wl,-z,defs -O2 -g  -std=c++20 -fPIC -MD -MT plugins/CMakeFiles/cursecheck.dir/cursecheck.cpp.o -MF plugins/CMakeFiles/cursecheck.dir/cursecheck.cpp.o.d -o plugins/CMakeFiles/cursecheck.dir/cursecheck.cpp.o -c /home/myk/src/dfhack/plugins/cursecheck.cpp
In file included from /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/x86_64-pc-linux-gnu/bits/c++allocator.h:33,
                 from /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/allocator.h:46,
                 from /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/list:63,
                 from /home/myk/src/dfhack/library/include/ColorText.h:28,
                 from /home/myk/src/dfhack/library/include/Console.h:27,
                 from /home/myk/src/dfhack/plugins/cursecheck.cpp:19:
In member function ‘void std::__new_allocator<_Tp>::deallocate(_Tp*, size_type) [with _Tp = df::unit*]’,
    inlined from ‘constexpr void std::allocator< <template-parameter-1-1> >::deallocate(_Tp*, std::size_t) [with _Tp = df::unit*]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/allocator.h:210:35,
    inlined from ‘static constexpr void std::allocator_traits<std::allocator<_Up> >::deallocate(allocator_type&, pointer, size_type) [with _Tp = df::unit*]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/alloc_traits.h:517:23,
    inlined from ‘constexpr void std::_Vector_base<_Tp, _Alloc>::_M_deallocate(pointer, std::size_t) [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:390:19,
    inlined from ‘constexpr std::_Vector_base<_Tp, _Alloc>::~_Vector_base() [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:369:15,
    inlined from ‘constexpr std::vector<_Tp, _Alloc>::~vector() [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:738:7,
    inlined from ‘DFHack::command_result cursecheck(DFHack::color_ostream&, std::vector<std::__cxx11::basic_string<char> >&)’ at /home/myk/src/dfhack/plugins/cursecheck.cpp:209:5:
/usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/new_allocator.h:172:33: error: ‘void operator delete(void*, std::size_t)’ called on pointer ‘<unknown>’ with nonzero offset 8 [-Werror=free-nonheap-object]
  172 |         _GLIBCXX_OPERATOR_DELETE(_GLIBCXX_SIZED_DEALLOC(__p, __n));
      |                                 ^
In member function ‘_Tp* std::__new_allocator<_Tp>::allocate(size_type, const void*) [with _Tp = df::unit*]’,
    inlined from ‘constexpr _Tp* std::allocator< <template-parameter-1-1> >::allocate(std::size_t) [with _Tp = df::unit*]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/allocator.h:198:40,
    inlined from ‘static constexpr _Tp* std::allocator_traits<std::allocator<_Up> >::allocate(allocator_type&, size_type) [with _Tp = df::unit*]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/alloc_traits.h:482:28,
    inlined from ‘constexpr std::_Vector_base<_Tp, _Alloc>::pointer std::_Vector_base<_Tp, _Alloc>::_M_allocate(std::size_t) [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:381:33,
    inlined from ‘constexpr std::_Vector_base<_Tp, _Alloc>::pointer std::_Vector_base<_Tp, _Alloc>::_M_allocate(std::size_t) [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:378:7,
    inlined from ‘constexpr void std::_Vector_base<_Tp, _Alloc>::_M_create_storage(std::size_t) [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:398:44,
    inlined from ‘constexpr std::_Vector_base<_Tp, _Alloc>::_Vector_base(std::size_t, const allocator_type&) [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:335:26,
    inlined from ‘constexpr std::vector<_Tp, _Alloc>::vector(const std::vector<_Tp, _Alloc>&) [with _Tp = df::unit*; _Alloc = std::allocator<df::unit*>]’ at /usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/stl_vector.h:603:61,
    inlined from ‘DFHack::command_result cursecheck(DFHack::color_ostream&, std::vector<std::__cxx11::basic_string<char> >&)’ at /home/myk/src/dfhack/plugins/cursecheck.cpp:160:81:
/usr/lib/gcc/x86_64-pc-linux-gnu/13/include/g++-v13/bits/new_allocator.h:151:55: note: returned from ‘void* operator new(std::size_t)’
  151 |         return static_cast<_Tp*>(_GLIBCXX_OPERATOR_NEW(__n * sizeof(_Tp)));
      |                                                       ^

[lethosor]

I would have just made a temporary vector for the loop here. Small thing, though.

[myk002]

my first version did just that, but it looked messy (because of the required null check, but it's not like what I ended up with was entirely "clean" either), but maybe I can clean it up more

[myk002]

#5150 is arguably cleaner

[ab9rf]

while i'm not going to change the code we have now, this would arguably have been a good use for a std::span

[myk002]

for my own education, how would that have worked? a ternary expression must have the same type in both of its branches. How would we wrap units.all to have the same type as the inline span?

[ab9rf]

i was thinking that we knew where in the all vector the selected unit is, but we don't, so there's no value in having either the entire units vector or just part of it, which is what std::span would have done for us here

we want here an object that holds a reference to a vector which is either owning or nonowning, and the object knows which. this is because in the single-unit case, we need a temporary vector which will be destructed when we're done with it, but in the all-units case we do not want to destruct the vector to be iterated. there are ways to do this with shared_ptr but they're not less oblique than the solution used