Re: [boost] Report from Berlin C++ Standards Committee meeting

From: Beman Dawes (bdawes_at_[hidden])
Date: 2006-04-10 07:03:49

"Aristid Breitkreuz" <aribrei_at_[hidden]> wrote in message 
news:1144663521.11938.4.camel_at_localhost.localdomain...
> Am Montag, den 10.04.2006, 10:49 +0100 schrieb Phil Nash:
>> Maxim Yegorushkin wrote:
>
>> As to the error handling - I haven't personally given it much thought -
>> but I'm not sure that can be easily parameterised without disturbing the
>> interface.
>
> Another nice feature would be to check if the output was actually
> generated from the complete input. Imagine the following scenario: You
> have
>  lexical_cast<double>("1,2")
> and as it happens, it returns 1 but happily. This is why I used strtod
> directly in an application of mine. You could add parameters with
> Boost.Parameters, by the way. So I would call:
>  lexical_cast<double>("1,2", ignore_tail = false)
> and get a exception. I will probably have to wait for Boost 1.36
> though :D.

The LWG is also moving forward with a separate set of conversion functions 
that are lower level, consume the input string (if any), and in general are 
to be used where lexical_cast is not the right tool.

There will be a paper from Pete Becker available in two or three weeks.

--Beman