Abdelrazak Younes wrote:
> Georg Baum wrote:
>> Abdelrazak Younes wrote:
>>
>>> Exactly and that was my argument but Lars insisted that the vector
>>> solution is better. He even said that docstring would probably better
>>> use vector instead.
>>
>> I must have overread that.
>
> Just for reference, here are his exact words:
>
> Abdelrazak Younes writes:
>
> | Jean-Marc Lasgouttes wrote:
> | >>>>>> "Lars" == Lars Gullik Bjønnes writes:
> | > Lars> So imho if docstring should change to anything as of now it is
> | > a
> | > Lars> std::vector<boost::uint32_t>
> | > Is it a threat? ;)
> |
> | No, just Lars reinventing basic_string with vector ;-)
> |
> | If you look at the STL code basic_string is very similar to vector
> | anyway... with a few string oriented specialisations. But wait, isn't
> | that what we are looking for? :-)
>
> No, it is what I am trying to avoid.
So Lars tries to avoid "with a few string oriented specialisations", or I'm
wrong?
But could we use such a string in the stream templates as string parameter?
Anyway, is there a consensus about the 32bit string?
And am I right that the current plan to complete the Unicode support is:
1. find a 32bit per character string class: docstring
2. find docstring streams
3. replace all std::string and streams with the 32bit versions
4. connect to the rest of the world by:
- input: utf8->ucs4 (files, ascii code points are part of utf8)
utf16->ucs4 (Qt)
- output: ucs4->utf8 (files)
ucs4->utf16 (Qt)
Peter
