Re: [boost] Should we add two simplecharacter-to-Unicode converters?

From: Robert Ramey (ramey_at_[hidden])
Date: 2005-08-22 10:30:09

The "Dataflow Iterators" section of the serialization library contains 
iterators for this purpose.  Its been part of boost since last November.

Robert Ramey

Daryle Walker wrote:
> On 8/19/05 9:15 AM, "Vladimir Prus" <ghost_at_[hidden]> wrote:
>
>> Daryle Walker wrote:
>>
>>> Nothing fancy, just something like:
>>>
>>>     int_fast32_t   char_to_Unicode( char c );
>>>     int_fast32_t  wchar_to_Unicode( wchar_t c );
>>>
>>> that converts a native character to a Unicode value.
>>
>> Maybe, but it's hard to comment as you haven't even explained what
>> those function will do. What's a "native character" and what a
>> "Unicode value" and how the conversion will be done? If the first
>> function does conversion from local 8 bit encoding to unicode then:
>
> "Native characters" are the character set a particular platform uses.
> Before the Unicode era, a platform (could) assume that all text files
> used the platform's character set.  (e.g. MacRoman for pre-X Macs,
> Cp-1251 for Windows, Latin-1 for UNIX)  My functions assume a
> one-to-one mapping from a native character to a Unicode code-point,
> because Phase 1 of C++ translation (see section 2.1 of the standard)
> assumes that.
>
>> - do you have a working implementation?
>
> No.  I'm just requesting for comments.
>
>> - isn't dealing with individual characters too slow?
>
> Probably.  Maybe we could add an iterator-copying version.