Re: [lmi] [lmi-commits] master 8f7ad46 2/3: Resolve a unit-test failure

[Greg Chicares]

On 2020-08-16 23:47, Vadim Zeitlin wrote:
> On Sun, 16 Aug 2020 19:34:52 -0400 (EDT) Greg Chicares 
> <gchicares@sbcglobal.net> wrote:
> 
> GC>     Commentary explains why std::filesystem::remove() deletes a
> GC>     directory that the msw C runtime's remove() does not. C99 (N1256) says
> GC>     [7.19.4.1] "If the file is open, the behavior of the remove function 
> is
> GC>     implementation-defined", but this directory doesn't seem to be "open",
> GC>     and apparently the msw implementation simply refuses to remove any
> GC>     directory.
> 
>  I'm too lazy to check the standard (and MSVC doesn't follow nor claim to
> support this one anyhow), but I'd be surprised if it mandated that remove()
> work for the directories.

Yes, I think you're right. I had been reading the C++17 filesystem
documentation first, and then the POSIX documentation. Later, when
I read the C standard, I assumed that it used 'file' as a generic
term meaning "file or directory"; but that's not justified, and
that standard never mentions directories at all.