Re: new unicode width test failure

[Daiki Ueno]

Pádraig Brady <address@hidden> writes:

> I've not looked into this at all yet.
>
> cheers,
> Pádraig.
>
> $ rpm -q gcc glibc
> gcc-5.1.1-1.fc22.x86_64
> glibc-2.21-5.fc22.x86_64
>
> FAIL: uniwidth/test-uc_width2.sh
> ================================
>
> 58,59c58,59
> < 085C..08E2    A
> < 08E3..0902    0
> ---
>> 085C..08E3    A
>> 08E4..0902    0

The diff looks like a reverse of the latest update, i.e., the old
uniwidth code is used.  How did you run the test?

I cannot reproduce it with:

  $ ./gnulib-tool --test --dir t uniwidth/width

nor:

  $ ./gnulib-tool --test --dir t uniwidth/width libunistring-optional

Regards,
-- 
Daiki Ueno