Mailing-List: contact dev-help@stdcxx.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@stdcxx.apache.org
Received-SPF: pass (athena.apache.org: local policy)
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: multipart/mixed;
	boundary="----_=_NextPart_001_01C89042.D0585937"
Subject: RE: [Stdcxx Wiki] Update of "LocaleLookup" by TravisVitek
Date: Thu, 27 Mar 2008 13:42:30 -0600
Message-ID: 
 <CFFDD219128FD94FB4F92B99F52D0A49465B27@exchmail01.Blue.Roguewave.Com>
Thread-Topic: [Stdcxx Wiki] Update of "LocaleLookup" by TravisVitek
Thread-Index: AciQGnanm0UoNo5WTgOfqN+Pkiq+KQAKDs/8
References: <20080326220304.410.29885@eos.apache.org>
 <47EBB480.100@roguewave.com>
From: "Travis Vitek" <Travis.Vitek@roguewave.com>
To: <dev@stdcxx.apache.org>

------_=_NextPart_001_01C89042.D0585937
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

>
>Look at you: you've become a regular wiki formatting artist! ;-)
>

Yup. I'm getting used to clicking through the Wiki formatting guide.

>> + || Test || Criteria ||
>
>Great breakdown. I assume the Criteria are the current conditions.
>We should also try to come up with the ideal conditions that we'll
>want to implement using the new API.

Yeah, that is the hard part.

>> + || 22.LOCALE.CODECVT.MT.CPP || *1,+ ||
>
>Any particular reason these names are all caps? :)
>

That is just the format of the output that I got when gathering the =
names
of all tests that use `rw_locales()'

>> + || 22.LOCALE.CODECVT.OUT.CPP || *2 ||
>
>IIRC, this test tries to exercise all values of MB_CUR_MAX, not
>just the largest.
>

Nope. It calls a function named `find_mb_locale()' which looks through =
the
list of installed locales for the one that has the largest return from =
the
function `get_mb_chars()'. The comment from `find_mb_locale()' says...

   finds the multibyte locale with the largest MB_CUR_MAX value and
   fills consecutive elemenets of the `mb_chars' array with multibyte
   characters between 1 and MB_CUR_MAX bytes long for such a locale

>> + || 22.LOCALE.CONS.MT.CPP || *1,+ ||
>> + || 22.LOCALE.CTYPE.CPP || *2 ||
>> + || 22.LOCALE.CTYPE.IS.CPP || *2 ||
>> + || 22.LOCALE.CTYPE.MT.CPP || *1,+ ||
>> + || 22.LOCALE.CTYPE.NARROW.CPP || *2 ||
>> + || 22.LOCALE.CTYPE.SCAN.CPP || *2 ||
>> + || 22.LOCALE.CTYPE.TOLOWER.CPP || *2 ||
>> + || 22.LOCALE.CTYPE.TOUPPER.CPP || *2 ||
>
>I thought the ctype tests were being run in all installed locales,
>just like the numpunct one? Which is what we want to move away from.
>IMO, exercising a small set (less than a dozen) of known locales and
>encodings should be plenty.
>

Yes, the non-mt ctype tests iterate over each locale for which the =
function
call `setlocale (LC_CTYPE, name)' succeeds. The mt ctype tests all limit
the number of tested locales to 32.

>> + || 22.LOCALE.TIME.GET.CPP || *5,6 ||
>
>I think in the time tests we look for a locale using a specific
>language (e.g., Danish, English, German). The encoding doesn't
>really matter, unless we throw in some Asian language as well.
>Grepping for the locale using a regular expression was the best
>we could do using the old interface. I suppose we'll still be
>using a regular expression with the new API, unless you have
>support for just language names (e.g., "da, de, en").
>

Yes. Using the grammar from the Wiki page, you can get the name of a =
single
Danish language locale for any country, codeset or MB_CUR_LEN value =
using
the following call...

  const char* name =3D rw_locale_query (LC_TIME, "da-*-*-*", 1);

If, as an enhancement, we wanted to allow the user to provide just the
language name, or language and country, we could probably do that.

Travis

------_=_NextPart_001_01C89042.D0585937--