On 06/12/2010 04:09 AM, Joel de Guzman wrote:
On 6/12/10 5:13 AM, Andrey Semashev wrote:
On 06/11/2010 12:35 PM, Hartmut Kaiser wrote:
There isn't anything like that. The reason is that usually you want to specify a character set in addition to the character type to use. So the solution above is nothing I would like to see in Spirit, even if it might be sufficient for your particular case. OTOH, if it is sufficient for you why don't you just add it inside your namespace and be done?
Well, I did so. It just so happens that I find myself replicating that code in different places.
I'm not very familiar with the standard specs on the new character types but isn't there a strong relationship between the character type and encoding? Can't we be sure that e.g. char16_t is UTF-16, char is the standard narrow and wchar_t is the standard wide encoding?
I don't think so. I use char/UTF-8 a lot now; I don't want char to be tied to standard narrow.
Ok, what about specifying encoding tag then? template< typename EncodingTagT > struct encoding_specific { ... }; I could create my traits to deduce the default encoding by the character type for my case.