Обсуждение: pgsql: Arrange to set the LC_XXX environment variables to match our

Log Message:
-----------
Arrange to set the LC_XXX environment variables to match our locale
setup.  This protects against undesired changes in locale behavior
if someone carelessly does setlocale(LC_ALL, "") (and we know who
you are, perl guys).

Modified Files:
--------------
    pgsql/src/backend/access/transam:
        xlog.c (r1.223 -> r1.224)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/access/transam/xlog.c.diff?r1=1.223&r2=1.224)
    pgsql/src/backend/main:
        main.c (r1.97 -> r1.98)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/main/main.c.diff?r1=1.97&r2=1.98)
    pgsql/src/backend/utils/adt:
        pg_locale.c (r1.32 -> r1.33)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/utils/adt/pg_locale.c.diff?r1=1.32&r2=1.33)
    pgsql/src/include/utils:
        pg_locale.h (r1.20 -> r1.21)
        (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/include/utils/pg_locale.h.diff?r1=1.20&r2=1.21)

This patch appears to have killed win32 builds that *don't* have NLS
enabled:

gcc -O2 -Wall -Wmissing-prototypes -Wpointer-arith -Winline
-Wdeclaration-after-
statement -Wendif-labels -fno-strict-aliasing -g
-I../../../../src/include -I./s
rc/include/port/win32 -DEXEC_BACKEND  -I/f/krb5/inc/krb5
"-I../../../../src/incl
ude/port/win32" -DBUILDING_DLL  -c -o pg_locale.o pg_locale.c
pg_locale.c: In function `pg_perm_setlocale':
pg_locale.c:109: error: `LC_MESSAGES' undeclared (first use in this
function)
pg_locale.c:109: error: (Each undeclared identifier is reported only
once
pg_locale.c:109: error: for each function it appears in.)
make: *** [pg_locale.o] Error 1


With NLS enabled, it builds.

Which brings a point - with NLS being so troublesome on win32, perhaps
one of the machines building on the buildfarm could be reconfigured to
not build with NLS?


//Magnus

> -----Original Message-----
> From: pgsql-committers-owner@postgresql.org
> [mailto:pgsql-committers-owner@postgresql.org] On Behalf Of Tom Lane
> Sent: Thursday, December 29, 2005 12:23 AM
> To: pgsql-committers@postgresql.org
> Subject: pgsql: Arrange to set the LC_XXX environment
> variables to match our
>
> Log Message:
> -----------
> Arrange to set the LC_XXX environment variables to match our
> locale setup.  This protects against undesired changes in
> locale behavior if someone carelessly does setlocale(LC_ALL,
> "") (and we know who you are, perl guys).
>
> Modified Files:
> --------------
>     pgsql/src/backend/access/transam:
>         xlog.c (r1.223 -> r1.224)
>
> (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/
> access/transam/xlog.c.diff?r1=1.223&r2=1.224)
>     pgsql/src/backend/main:
>         main.c (r1.97 -> r1.98)
>
> (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/
> main/main.c.diff?r1=1.97&r2=1.98)
>     pgsql/src/backend/utils/adt:
>         pg_locale.c (r1.32 -> r1.33)
>
> (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/
> utils/adt/pg_locale.c.diff?r1=1.32&r2=1.33)
>     pgsql/src/include/utils:
>         pg_locale.h (r1.20 -> r1.21)
>
> (http://developer.postgresql.org/cvsweb.cgi/pgsql/src/include/
> utils/pg_locale.h.diff?r1=1.20&r2=1.21)
>
> ---------------------------(end of
> broadcast)---------------------------
> TIP 1: if posting/reading through Usenet, please send an appropriate
>        subscribe-nomail command to majordomo@postgresql.org
> so that your
>        message can get through to the mailing list cleanly
>

On 31/12/05 14:35, "Magnus Hagander" <mha@sollentuna.net> wrote:

> Which brings a point - with NLS being so troublesome on win32, perhaps
> one of the machines building on the buildfarm could be reconfigured to
> not build with NLS?

Not a bad idea. I'd suggest it's not Snake though, as that box is kept in
sync compiler/config/dependency-wise with the machine that the distributions
are built on.

Regards, Dave.

"Magnus Hagander" <mha@sollentuna.net> writes:
> This patch appears to have killed win32 builds that *don't* have NLS
> enabled:

> pg_locale.c:109: error: `LC_MESSAGES' undeclared (first use in this
> function)

Silly me for assuming all WIN32 platforms are alike ;-)

What's the appropriate behavior in this case --- just treat it as a
no-op, as we do on Unix when LC_MESSAGES isn't defined?

            regards, tom lane

> > This patch appears to have killed win32 builds that *don't* have NLS
> > enabled:
>
> > pg_locale.c:109: error: `LC_MESSAGES' undeclared (first use in this
> > function)
>
> Silly me for assuming all WIN32 platforms are alike ;-)

Yes, indeed!
(LC_MESSAGES comes from libintl. Windows never supports it.)


> What's the appropriate behavior in this case --- just treat
> it as a no-op, as we do on Unix when LC_MESSAGES isn't defined?

Yes, I think that's correct.

//Magnus

"Magnus Hagander" <mha@sollentuna.net> writes:
>>> This patch appears to have killed win32 builds that *don't* have NLS
>>> enabled:

>> What's the appropriate behavior in this case --- just treat
>> it as a no-op, as we do on Unix when LC_MESSAGES isn't defined?

> Yes, I think that's correct.

OK, patch committed.

            regards, tom lane