Re: [boost] [algorithm][hex] function hex_char_to_int in unnamed namespace in header file

Subject: Re: [boost] [algorithm][hex] function hex_char_to_int in unnamed namespace in header file
From: Andrey Semashev (andrey.semashev_at_[hidden])
Date: 2012-12-03 15:10:58

On December 3, 2012 11:58:54 PM Marshall Clow <mclow.lists_at_[hidden]> wrote:
> On Dec 3, 2012, at 9:43 AM, Felipe Magno de Almeida 
> <felipe.m.almeida_at_[hidden]> wrote:
>
> > 
> http://boost.2283326.n4.nabble.com/Boost-inspection-notification-2007-06-09-RC-1-34-0-X-tt2628426.html#a2628478
> >
> > AFAIK, an unnamed namespace creates different entities for each
> > translation on which it is defined (by #inclusion,
> > for example).
>
> Yes.
>
> > If any entity defined in a unnamed namespace is used in
> > an external linkage definition, such as a
> > inline function or a template definition and this definition is used
> > in multiple TUs it causes ODR violations, because
> > it is defined differently (uses different entities) in each TU.
>
> And No. Because the call to it (in each TU) will call the entity in the 
> specific unnamed namespace that was part of that translation unit.

Exactly, and that's what makes ODR violated. Imagine an inline function 
in a header, the function depends on the entity in an unnamed 
namespace. The function will be defined differently in each TU that 
includes the header, and that's UB.