Hi,
Christian Weisgerber wrote on Fri, Oct 23, 2015 at 03:52:31PM +0200:
> Ted Unangst:
>> I'm very scared to try counting chars vs bytes upfront in such code.
Actually, that turns out to be simple.
> I think that's insufficient to cover rs's functionality. -z will
> overestimate the required
Ingo Schwarze wrote:
> I think that way we can actually start committing such patches and
> improve our userland.
>
> Two final notes:
>
> 1. It turns out each of the three programs needs exactly one
> multibyte-character helper function in utf8.c, and each helper
> function uses
Ted Unangst:
> I'm very scared to try counting chars vs bytes upfront in such code. However,
> the code that prints spaces to pad the output is much simpler.
I think that's insufficient to cover rs's functionality. -z will
overestimate the required widths and... yes, -j is completely broken.