Re: setlocale.3: Wording issues

["Alejandro Colomar (man-pages)" <alx.manpages@gmail.com>]

Hi Michael,

On 1/8/21 2:17 PM, Michael Kerrisk (man-pages) wrote:
> [CC += Bruno]
> 
> Hi Alex,
> 
> On 1/8/21 2:05 PM, Alejandro Colomar (man-pages) wrote:
>> Hi Michael,
>>
>> On 1/8/21 1:45 PM, Michael Kerrisk (man-pages) wrote:
>>> On 1/8/21 12:41 PM, Alejandro Colomar (man-pages) wrote:
>>>> Hi Michael,
>>>>
>>>> On 1/8/21 10:26 AM, Michael Kerrisk (man-pages) wrote:
>>>>> Hi Alex,
>>>>>
>>>>> On 1/7/21 7:32 PM, Alejandro Colomar (man-pages) wrote:
>>>>>> Hi Michael,
>>>>>>
>>>>>> I don't understand what this paragraph means, I think it needs some wfix.
>>>>>>
>>>>>> Around setlocale.3:179:
>>>>>> [
>>>>>>        On  startup  of  the main program, the portable "C" locale is
>>>>>>        selected as default.  A program may be made portable  to  all
>>>>>>        locales by calling:
>>>>>>
>>>>>>            setlocale(LC_ALL, "");
>>>>>>
>>>>>>        after  program  initialization,  by using the values returned
>>>>>>        from a localeconv(3) call for  locale-dependent  information,
>>>>>>        by  using the multibyte and wide character functions for text
>>>>>>        processing if MB_CUR_MAX > 1, and by  using  strcoll(3),  wc‐
>>>>>>        scoll(3) or strxfrm(3), wcsxfrm(3) to compare strings.
>>>>>>
>>>>>> <<<Especially these last 2 lines
>>>>>>
>>>>>> ]
>>>>>
>>>>> I see what you mean. I had to read that a few times to parse it.
>>>>> It looks like the text was added in 1999. I think the following
>>>>> clarifies and preserves the meaning:
>>>>>
>>>>> [[
>>>>>        On startup of the main program, the portable "C" locale is select‐
>>>>>        ed  as  default.  A program may be made portable to all locales by
>>>>>        calling:
>>>>>
>>>>>            setlocale(LC_ALL, "");
>>>>>
>>>>>        after program initialization, and then:
>>>>>
>>>>>        (a) using the values returned from a localeconv(3)  call  for  lo‐
>>>>>            cale-dependent information;
>>>>>
>>>>>        (c) using the multibyte and wide character functions for text pro‐
>>>>>            cessing if MB_CUR_MAX > 1; and
>>>>>
>>>>>        (c) using strcoll(3), wcscoll(3) or strxfrm(3), wcsxfrm(3) to com‐
>>>>>            pare strings.
>>>>> ]]
>>>>>
>>>>> What do you think?
>>>>
>>>> Much better.
>>>>
>>>> But I still don't get why [A, B or  C, D].  What does it mean?
>>>
>>> I don't read it that way. I see it as: [A and (B and/or C and/or D].
>>> Do you see what I mean?
>>
>> About the three lines (a) (b) (c) as a whole, yes, I understand that part.
>> About [strcoll(3), wcscoll(3) or strxfrm(3), wcsxfrm(3)], nope, I still
>> can't understand the punctuation there.
>> So I can't parse the inner contents of (c).
> 
> Ah sorry. Now I see what you mean. I'm not sure of this, but I suspect
> that it's supposed to mean:
> 
>     Use [strcoll(3)+strxfrm(3)] for non-wide-char strings and 
>     [wcscoll(3)+wcsxfrm(3)] for wide char strings.
> 
> Maybe Bruno can help.

Yes, that was my first thought (after a few trials), but I preferred to
check with you :)

I don't know how to express it.
Maybe [A and B, or C and D]?

Thanks,

Alex

> 
> Thanks,
> 
> Michael
> 

-- 
Alejandro Colomar
Linux man-pages comaintainer; https://www.kernel.org/doc/man-pages/
http://www.alejandro-colomar.es/