Re: [Bino-list] Bug: translation of Usage text using UTF-8

[Martin Lambers]

Hi Alexey!

On 17/04/11 20:21, Alexey Osipov wrote:
> In process of making russian translation for bino, I've found a bug,
> which makes translated Usage text seem not quite good.
> 
> In msg::msg_txt you use 'for' loop to cycle through std::string,
> representing a text (the Usage text in particular), and count symbols to
> determine the line length. But if text is in UTF-8 (or other multi-byte
> encoding), the counting is doing wrong - two-byte character counts for
> two characters. So the total allowed length of the string becomes less
> than default 80 columns and hence translated string is wrapped to the
> next line.
> 
> See attached screenshot.
> 
> You should use std::wstring or something to count multi-byte characters
> as one character.

You are right. This should fix it:
http://git.savannah.gnu.org/cgit/bino.git/commit/?id=66993854b13c0039d98ce3e7b3e2ec8856236c98

On systems that have wcwidth(), this should also work with locales where
one character does not necessarily mean one output column.

Does this work for you?

Martin