Alexey Parshin wrote: > Currently, CVariant is treating strings as 0-terminated. At least, I > didn't > enforce it to work with strings as with a buffer. We can alter this. I don't know. It isn't immediately obvious what is correct behavior in this case. Perhaps we may just need asBuffer() function for CVariant, and an extra data type (Buffer, BLOB, whatever you want to call it). > As for > the wide strings - I'm still thinking about it. Coala was working on > unicode > version of SPTK, and as soon as it's working - the next logical step > would > be to switch everything to wstring, right? Not sure. Wstrings take up more space/take longer to process. This should be optional behavior. Either we'll supply two different libraries, or do something funny with templates, and supply two full sets of functions that involve text manipulation. It should be relatively easy in reality, provided we drop support for ancient compilers that can't handle templates correctly. > > 2006/8/22, Ilya A. Volynets-Evenbakh <ilya@total-knowledge.com>: >> >> I don't know if it's bug or feature, but if you assign std::sting that >> contains NUL character >> in the middle to CVariant, CVariant seems to truncate the value at NUL. >> One danger here >> is wide string handling - is it going to work with it? >> >> -- >> Ilya A. Volynets-Evenbakh >> Total Knowledge. CTO >> http://www.total-knowledge.com >> >> > > -- Ilya A. Volynets-Evenbakh Total Knowledge. CTO http://www.total-knowledge.com
List hosted by Total Knowledge