|
Boost Testing : |
From: Christopher Cambly (ccambly_at_[hidden])
Date: 2007-09-27 17:33:21
> This is the third case so far of two POSIX macros being defined as the
> same value. The others being EOPNOTSUPP / ENOTSUP and EAGAIN /
> EWOULDBLOCK. The recipe we are using to cope is:
>
> # if ENOTEMPTY != EEXIST
> case ENOTEMPTY: return make_error_condition( directory_not_empty );
> # endif // ENOTEMPTY != EEXIST
>
> The advantage being that it works for any other system that happens to
> have done the same thing.
>
> I've committed this change to SVN.
Great, thanks!
>
> > As an aside, one
> > of our developers mentioned that Boost.Filesystem has a similar
problem
> > with the macros ENOTEMPTY and EEXIST not being unique, however I not
> > verified this yet.
>
> Was this in operations.cpp? The code I see there will still work if the
> two macros define the same value, unless I'm missing something.
>
Likely, my guess would be that it was much earlier version operations.cpp
that he was refering too.
Chris Cambly
XL C++ Compiler Development