20 Feb
2005
20 Feb
'05
8:22 a.m.
On Sat, Feb 19, 2005 at 12:41:57PM +0100, Hartmut Kaiser <HartmutKaiser@t-online.de> wrote:
You are right! I haven't realised, that the path::operator/=() implicitly calls the path() constructor when used with a string on the right sight:
fs::path p; std::string s; ...
p /= s; // uses the default name checker here
This problem is fixed now (at least for the Windows environment). Thanks! Thats the most common problem with boost::fs, whats the rationale behind op/= not using the path checker of the existing path object?
Regards Andreas