This has refreshed my fading memory. The patch seems like the best
solution. Is there any objection to applying it?

(Petr, I assume you intended to send this to a mailing list also)

cheers

andrew

-------- Original Message --------
Subject: Re: regression failures on WIndows in machines with some
non-English locales
Date: Tue, 01 Nov 2005 07:55:59 +0100
From: Petr Jelinek <pjmodos@seznam.cz>
To: Tom Lane <tgl@sss.pgh.pa.us>, Andrew Dunstan <andrew@dunslane.net>
References: <4366C2AE.70500@dunslane.net> <21289.1130816310@sss.pgh.pa.us>



BTW Tom you should know it's not really failure it's just different
sorting of output in rules test (due to "ch" being one letter in my locale).

When I suggested --no-locale for regression test I ment something like
whats in attachement - this will not change current functionality just
adds option to set NO_LOCALE=something in make check commandline which
would enable me to setup buildfarm member (after patching buildfarm
client, I think it should be option which is off by default there too).

I want this aproach because forcing --no-locale for everybody could mean
that we won't find some errors - good example of this is "force C
locale for temp regression installations" thread
(http://archives.postgresql.org/pgsql...8/msg00352.php)
where we discussed it, you both might remmeber it.

--
Regards
Petr Jelinek (PJMODOS)




Index: GNUmakefile
================================================== =================
RCS file: /projects/cvsroot/pgsql/src/test/regress/GNUmakefile,v
retrieving revision 1.52
diff -c -r1.52 GNUmakefile
*** GNUmakefile 28 Jul 2005 04:32:32 -0000 1.52
--- GNUmakefile 1 Nov 2005 06:34:49 -0000
***************
*** 28,33 ****
--- 28,38 ----
MAXCONNOPT += --max-connections=$(MAX_CONNECTIONS)
endif

+ # locale
+ NOLOCALE :=
+ ifdef NO_LOCALE
+ NOLOCALE += --no-locale
+ endif

##
## Prepare for tests
***************
*** 132,148 ****
check: all
-rm -rf ./testtablespace
mkdir ./testtablespace
! $(SHELL) ./pg_regress --temp-install --top-builddir=$(top_builddir) --temp-port=$(TEMP_PORT) --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT)

installcheck: all
-rm -rf ./testtablespace
mkdir ./testtablespace
! $(SHELL) ./pg_regress --schedule=$(srcdir)/serial_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql

installcheck-parallel: all
-rm -rf ./testtablespace
mkdir ./testtablespace
! $(SHELL) ./pg_regress --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT)


# old interfaces follow...
--- 137,153 ----
check: all
-rm -rf ./testtablespace
mkdir ./testtablespace
! $(SHELL) ./pg_regress --temp-install --top-builddir=$(top_builddir) --temp-port=$(TEMP_PORT) --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT) $(NOLOCALE)

installcheck: all
-rm -rf ./testtablespace
mkdir ./testtablespace
! $(SHELL) ./pg_regress --schedule=$(srcdir)/serial_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(NOLOCALE)

installcheck-parallel: all
-rm -rf ./testtablespace
mkdir ./testtablespace
! $(SHELL) ./pg_regress --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT) $(NOLOCALE)


# old interfaces follow...
***************
*** 152,161 ****
runtest-parallel: installcheck-parallel

bigtest:
! $(SHELL) ./pg_regress --schedule=$(srcdir)/serial_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql numeric_big

bigcheck:
! $(SHELL) ./pg_regress --temp-install --top-builddir=$(top_builddir) --temp-port=$(TEMP_PORT) --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT) numeric_big


##
--- 157,166 ----
runtest-parallel: installcheck-parallel

bigtest:
! $(SHELL) ./pg_regress --schedule=$(srcdir)/serial_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql numeric_big $(NOLOCALE)

bigcheck:
! $(SHELL) ./pg_regress --temp-install --top-builddir=$(top_builddir) --temp-port=$(TEMP_PORT) --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT) numeric_big $(NOLOCALE)


##
Index: pg_regress.sh
================================================== =================
RCS file: /projects/cvsroot/pgsql/src/test/regress/pg_regress.sh,v
retrieving revision 1.60
diff -c -r1.60 pg_regress.sh
*** pg_regress.sh 5 Sep 2005 23:50:49 -0000 1.60
--- pg_regress.sh 1 Nov 2005 06:34:51 -0000
***************
*** 24,29 ****
--- 24,30 ----
--schedule=FILE use test ordering schedule from FILE
(may be used multiple times to concatenate)
--temp-install[=DIR] create a temporary installation (in DIR)
+ --no-locale use C locale

Options for \`temp-install' mode:
--top-builddir=DIR (relative) path to top level build directory
***************
*** 99,104 ****
--- 100,106 ----
unset mode
unset schedule
unset debug
+ unset nolocale
unset top_builddir
unset temp_install
unset multibyte
***************
*** 142,147 ****
--- 144,152 ----
--multibyte=*)
multibyte=`expr "x$1" : "x--multibyte=\(.*\)"`
shift;;
+ --no-locale)
+ nolocale=yes
+ shift;;
--temp-install)
temp_install=./tmp_check
shift;;
***************
*** 403,409 ****
fi

message "initializing database system"
! [ "$debug" = yes ] && initdb_options='--debug'
"$bindir/initdb" -D "$PGDATA" -L "$datadir" --noclean $initdb_options >"$LOGDIR/initdb.log" 2>&1

if [ $? -ne 0 ]
--- 408,415 ----
fi

message "initializing database system"
! [ "$debug" = yes ] && initdb_options="--debug"
! [ "$nolocale" = yes ] && initdb_options="$initdb_options --no-locale"
"$bindir/initdb" -D "$PGDATA" -L "$datadir" --noclean $initdb_options >"$LOGDIR/initdb.log" 2>&1

if [ $? -ne 0 ]



---------------------------(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

Andrew Dunstan <andrew@dunslane.net> writes:
> This has refreshed my fading memory. The patch seems like the best
> solution. Is there any objection to applying it?

Putting the switch at the end seems certain to fail on some platforms
(some versions of getopt are fussier than others).

> bigtest:
> ! $(SHELL) ./pg_regress --schedule=$(srcdir)/serial_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql numeric_big $(NOLOCALE)

> bigcheck:
> ! $(SHELL) ./pg_regress --temp-install --top-builddir=$(top_builddir) --temp-port=$(TEMP_PORT) --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT) numeric_big $(NOLOCALE)

Put it with the other switches, and I won't object.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

Tom Lane wrote:

>
>
>
>> bigtest:
>>! $(SHELL) ./pg_regress --schedule=$(srcdir)/serial_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql numeric_big $(NOLOCALE)
>>
>>
>
>
>
>> bigcheck:
>>! $(SHELL) ./pg_regress --temp-install --top-builddir=$(top_builddir) --temp-port=$(TEMP_PORT) --schedule=$(srcdir)/parallel_schedule --multibyte=$(MULTIBYTE) --load-language=plpgsql $(MAXCONNOPT) numeric_big $(NOLOCALE)
>>
>>
>
>Put it with the other switches, and I won't object.
>
>
>
>

applied with this fixed to 8.0 and HEAD.

cheers

andrew

---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

http://archives.postgresql.org