On Fri, 2023-02-17 at 15:07 -0800, Jeff Davis wrote:
> 2. Update the pg_database entry for template0. This has less
> potential
> for surprise in case people are actually using template0 for a
> template.
New patches attached.
0001: default autoconf to build with ICU (meson already uses 'auto')
0002: update template0 in new cluster (as described above)
0003: default initdb to use ICU
Updating template0, as in 0002, seems straightforward and unsurprising,
since only template0 is preserved and it was only initialized for the
purposes of upgrading. Also, template0 is not sensitive to locale
settings, and doesn't even have the datcollversion set. The patch
updates encoding, datlocprovider, datcollate, datctype, and
daticulocale on the new cluster. No doc update, because there are some
initdb settings (like checksums) which still need to be compatible
between the old and the new cluster.
--
Jeff Davis
PostgreSQL Contributor Team - AWS
From 34fcb5ebfd0463a659170ea0d7599bd37d6c3e76 Mon Sep 17 00:00:00 2001
From: Jeff Davis <j...@j-davis.com>
Date: Fri, 10 Feb 2023 12:08:11 -0800
Subject: [PATCH v4 1/3] Build ICU support by default.
Discussion: https://postgr.es/m/510d284759f6e943ce15096167760b2edcb2e700.ca...@j-davis.com
---
.cirrus.yml | 1 +
configure | 36 ++++++----------
configure.ac | 8 +++-
doc/src/sgml/installation.sgml | 76 +++++++++++++++++++---------------
4 files changed, 61 insertions(+), 60 deletions(-)
diff --git a/.cirrus.yml b/.cirrus.yml
index f212978752..34450a9c7b 100644
--- a/.cirrus.yml
+++ b/.cirrus.yml
@@ -751,6 +751,7 @@ task:
time ./configure \
--host=x86_64-w64-mingw32 \
--enable-cassert \
+ --without-icu \
CC="ccache x86_64-w64-mingw32-gcc" \
CXX="ccache x86_64-w64-mingw32-g++"
make -s -j${BUILD_JOBS} clean
diff --git a/configure b/configure
index e35769ea73..436c2446e8 100755
--- a/configure
+++ b/configure
@@ -1558,7 +1558,7 @@ Optional Packages:
set WAL block size in kB [8]
--with-CC=CMD set compiler (deprecated)
--with-llvm build with LLVM based JIT support
- --with-icu build with ICU support
+ --without-icu build without ICU support
--with-tcl build Tcl modules (PL/Tcl)
--with-tclconfig=DIR tclConfig.sh is in DIR
--with-perl build Perl modules (PL/Perl)
@@ -8401,7 +8401,9 @@ $as_echo "#define USE_ICU 1" >>confdefs.h
esac
else
- with_icu=no
+ with_icu=yes
+
+$as_echo "#define USE_ICU 1" >>confdefs.h
fi
@@ -8470,31 +8472,17 @@ fi
# Put the nasty error message in config.log where it belongs
echo "$ICU_PKG_ERRORS" >&5
- as_fn_error $? "Package requirements (icu-uc icu-i18n) were not met:
-
-$ICU_PKG_ERRORS
-
-Consider adjusting the PKG_CONFIG_PATH environment variable if you
-installed software in a non-standard prefix.
-
-Alternatively, you may set the environment variables ICU_CFLAGS
-and ICU_LIBS to avoid the need to call pkg-config.
-See the pkg-config man page for more details." "$LINENO" 5
+ as_fn_error $? "ICU library not found
+If you have ICU already installed, see config.log for details on the
+failure. It is possible the compiler isn't looking in the proper directory.
+Use --without-icu to disable ICU support." "$LINENO" 5
elif test $pkg_failed = untried; then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; }
- { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5
-$as_echo "$as_me: error: in \`$ac_pwd':" >&2;}
-as_fn_error $? "The pkg-config script could not be found or is too old. Make sure it
-is in your PATH or set the PKG_CONFIG environment variable to the full
-path to pkg-config.
-
-Alternatively, you may set the environment variables ICU_CFLAGS
-and ICU_LIBS to avoid the need to call pkg-config.
-See the pkg-config man page for more details.
-
-To get pkg-config, see <http://pkg-config.freedesktop.org/>.
-See \`config.log' for more details" "$LINENO" 5; }
+ as_fn_error $? "ICU library not found
+If you have ICU already installed, see config.log for details on the
+failure. It is possible the compiler isn't looking in the proper directory.
+Use --without-icu to disable ICU support." "$LINENO" 5
else
ICU_CFLAGS=$pkg_cv_ICU_CFLAGS
ICU_LIBS=$pkg_cv_ICU_LIBS
diff --git a/configure.ac b/configure.ac
index af23c15cb2..215e32120f 100644
--- a/configure.ac
+++ b/configure.ac
@@ -853,13 +853,17 @@ AC_SUBST(enable_thread_safety)
# ICU
#
AC_MSG_CHECKING([whether to build with ICU support])
-PGAC_ARG_BOOL(with, icu, no, [build with ICU support],
+PGAC_ARG_BOOL(with, icu, yes, [build without ICU support],
[AC_DEFINE([USE_ICU], 1, [Define to build with ICU support. (--with-icu)])])
AC_MSG_RESULT([$with_icu])
AC_SUBST(with_icu)
if test "$with_icu" = yes; then
- PKG_CHECK_MODULES(ICU, icu-uc icu-i18n)
+ PKG_CHECK_MODULES(ICU, icu-uc icu-i18n, [],
+ [AC_MSG_ERROR([ICU library not found
+If you have ICU already installed, see config.log for details on the
+failure. It is possible the compiler isn't looking in the proper directory.
+Use --without-icu to disable ICU support.])])
fi
#
diff --git a/doc/src/sgml/installation.sgml b/doc/src/sgml/installation.sgml
index 0ed35d99e9..4081820eca 100644
--- a/doc/src/sgml/installation.sgml
+++ b/doc/src/sgml/installation.sgml
@@ -146,6 +146,35 @@ documentation. See standalone-profile.xsl for details.
<application>pg_restore</application>.
</para>
</listitem>
+
+ <listitem>
+ <para>
+ The ICU locale provider (see <xref linkend="locale-providers"/>) is used by default. If you don't want to use it then you must specify the <option>--without-icu</option> option to <filename>configure</filename>. Using this option disables support for ICU collation features (see <xref linkend="collation"/>).
+ </para>
+ <para>
+ ICU support requires the <productname>ICU4C</productname> package to be
+ installed. The minimum required version of
+ <productname>ICU4C</productname> is currently 4.2.
+ </para>
+
+ <para>
+ By default,
+ <productname>pkg-config</productname><indexterm><primary>pkg-config</primary></indexterm>
+ will be used to find the required compilation options. This is
+ supported for <productname>ICU4C</productname> version 4.6 and later.
+ For older versions, or if <productname>pkg-config</productname> is not
+ available, the variables <envar>ICU_CFLAGS</envar> and
+ <envar>ICU_LIBS</envar> can be specified to
+ <filename>configure</filename>, like in this example:
+<programlisting>
+./configure ... ICU_CFLAGS='-I/some/where/include' ICU_LIBS='-L/some/where/lib -licui18n -licuuc -licudata'
+</programlisting>
+ (If <productname>ICU4C</productname> is in the default search path
+ for the compiler, then you still need to specify nonempty strings in
+ order to avoid use of <productname>pkg-config</productname>, for
+ example, <literal>ICU_CFLAGS=' '</literal>.)
+ </para>
+ </listitem>
</itemizedlist>
</para>
@@ -926,40 +955,6 @@ build-postgresql:
</listitem>
</varlistentry>
- <varlistentry id="configure-option-with-icu">
- <term><option>--with-icu</option></term>
- <listitem>
- <para>
- Build with support for
- the <productname>ICU</productname><indexterm><primary>ICU</primary></indexterm>
- library, enabling use of ICU collation
- features<phrase condition="standalone-ignore"> (see
- <xref linkend="collation"/>)</phrase>.
- This requires the <productname>ICU4C</productname> package
- to be installed. The minimum required version
- of <productname>ICU4C</productname> is currently 4.2.
- </para>
-
- <para>
- By default,
- <productname>pkg-config</productname><indexterm><primary>pkg-config</primary></indexterm>
- will be used to find the required compilation options. This is
- supported for <productname>ICU4C</productname> version 4.6 and later.
- For older versions, or if <productname>pkg-config</productname> is
- not available, the variables <envar>ICU_CFLAGS</envar>
- and <envar>ICU_LIBS</envar> can be specified
- to <filename>configure</filename>, like in this example:
-<programlisting>
-./configure ... --with-icu ICU_CFLAGS='-I/some/where/include' ICU_LIBS='-L/some/where/lib -licui18n -licuuc -licudata'
-</programlisting>
- (If <productname>ICU4C</productname> is in the default search path
- for the compiler, then you still need to specify nonempty strings in
- order to avoid use of <productname>pkg-config</productname>, for
- example, <literal>ICU_CFLAGS=' '</literal>.)
- </para>
- </listitem>
- </varlistentry>
-
<varlistentry id="configure-with-llvm">
<term><option>--with-llvm</option></term>
<listitem>
@@ -1231,6 +1226,19 @@ build-postgresql:
<variablelist>
+ <varlistentry id="configure-option-without-icu">
+ <term><option>--without-icu</option></term>
+ <listitem>
+ <para>
+ Build without support for the
+ <productname>ICU</productname><indexterm><primary>ICU</primary></indexterm>
+ library, disabling the use of ICU collation features<phrase
+ condition="standalone-ignore"> (see <xref
+ linkend="collation"/>)</phrase>.
+ </para>
+ </listitem>
+ </varlistentry>
+
<varlistentry id="configure-option-without-readline">
<term><option>--without-readline</option></term>
<listitem>
--
2.34.1
From d14def6ba7ad2bffa12e2e6aa7bac7946bdfb4d5 Mon Sep 17 00:00:00 2001
From: Jeff Davis <j...@j-davis.com>
Date: Fri, 24 Feb 2023 08:55:11 -0800
Subject: [PATCH v4 2/3] pg_upgrade: copy locale and encoding information to
new cluster.
Previously, pg_upgrade checked that the old and new clusters were
compatible, including the locale and encoding. But the new cluster was
just created, and only template0 from the new cluster will be
preserved (template1 and postgres are both recreated during the
upgrade process).
Because template0 is not sensitive to locale or encoding, just update
the pg_database entry to be the same as template0 from the old
cluster.
This commit makes it easier to change the default initdb locale or
encoding settings without causing needless incompatibilities.
---
src/bin/pg_upgrade/check.c | 160 --------------------------------
src/bin/pg_upgrade/info.c | 50 ++++++++++
src/bin/pg_upgrade/pg_upgrade.c | 62 +++++++++++++
src/bin/pg_upgrade/pg_upgrade.h | 1 +
4 files changed, 113 insertions(+), 160 deletions(-)
diff --git a/src/bin/pg_upgrade/check.c b/src/bin/pg_upgrade/check.c
index 7cf68dc9af..b71b00be37 100644
--- a/src/bin/pg_upgrade/check.c
+++ b/src/bin/pg_upgrade/check.c
@@ -16,9 +16,6 @@
#include "pg_upgrade.h"
static void check_new_cluster_is_empty(void);
-static void check_databases_are_compatible(void);
-static void check_locale_and_encoding(DbInfo *olddb, DbInfo *newdb);
-static bool equivalent_locale(int category, const char *loca, const char *locb);
static void check_is_install_user(ClusterInfo *cluster);
static void check_proper_datallowconn(ClusterInfo *cluster);
static void check_for_prepared_transactions(ClusterInfo *cluster);
@@ -33,7 +30,6 @@ static void check_for_jsonb_9_4_usage(ClusterInfo *cluster);
static void check_for_pg_role_prefix(ClusterInfo *cluster);
static void check_for_new_tablespace_dir(ClusterInfo *new_cluster);
static void check_for_user_defined_encoding_conversions(ClusterInfo *cluster);
-static char *get_canonical_locale_name(int category, const char *locale);
/*
@@ -194,7 +190,6 @@ check_new_cluster(void)
get_db_and_rel_infos(&new_cluster);
check_new_cluster_is_empty();
- check_databases_are_compatible();
check_loadable_libraries();
@@ -349,94 +344,6 @@ check_cluster_compatibility(bool live_check)
}
-/*
- * check_locale_and_encoding()
- *
- * Check that locale and encoding of a database in the old and new clusters
- * are compatible.
- */
-static void
-check_locale_and_encoding(DbInfo *olddb, DbInfo *newdb)
-{
- if (olddb->db_encoding != newdb->db_encoding)
- pg_fatal("encodings for database \"%s\" do not match: old \"%s\", new \"%s\"",
- olddb->db_name,
- pg_encoding_to_char(olddb->db_encoding),
- pg_encoding_to_char(newdb->db_encoding));
- if (!equivalent_locale(LC_COLLATE, olddb->db_collate, newdb->db_collate))
- pg_fatal("lc_collate values for database \"%s\" do not match: old \"%s\", new \"%s\"",
- olddb->db_name, olddb->db_collate, newdb->db_collate);
- if (!equivalent_locale(LC_CTYPE, olddb->db_ctype, newdb->db_ctype))
- pg_fatal("lc_ctype values for database \"%s\" do not match: old \"%s\", new \"%s\"",
- olddb->db_name, olddb->db_ctype, newdb->db_ctype);
- if (olddb->db_collprovider != newdb->db_collprovider)
- pg_fatal("locale providers for database \"%s\" do not match: old \"%s\", new \"%s\"",
- olddb->db_name,
- collprovider_name(olddb->db_collprovider),
- collprovider_name(newdb->db_collprovider));
- if ((olddb->db_iculocale == NULL && newdb->db_iculocale != NULL) ||
- (olddb->db_iculocale != NULL && newdb->db_iculocale == NULL) ||
- (olddb->db_iculocale != NULL && newdb->db_iculocale != NULL && strcmp(olddb->db_iculocale, newdb->db_iculocale) != 0))
- pg_fatal("ICU locale values for database \"%s\" do not match: old \"%s\", new \"%s\"",
- olddb->db_name,
- olddb->db_iculocale ? olddb->db_iculocale : "(null)",
- newdb->db_iculocale ? newdb->db_iculocale : "(null)");
-}
-
-/*
- * equivalent_locale()
- *
- * Best effort locale-name comparison. Return false if we are not 100% sure
- * the locales are equivalent.
- *
- * Note: The encoding parts of the names are ignored. This function is
- * currently used to compare locale names stored in pg_database, and
- * pg_database contains a separate encoding field. That's compared directly
- * in check_locale_and_encoding().
- */
-static bool
-equivalent_locale(int category, const char *loca, const char *locb)
-{
- const char *chara;
- const char *charb;
- char *canona;
- char *canonb;
- int lena;
- int lenb;
-
- /*
- * If the names are equal, the locales are equivalent. Checking this first
- * avoids calling setlocale() in the common case that the names are equal.
- * That's a good thing, if setlocale() is buggy, for example.
- */
- if (pg_strcasecmp(loca, locb) == 0)
- return true;
-
- /*
- * Not identical. Canonicalize both names, remove the encoding parts, and
- * try again.
- */
- canona = get_canonical_locale_name(category, loca);
- chara = strrchr(canona, '.');
- lena = chara ? (chara - canona) : strlen(canona);
-
- canonb = get_canonical_locale_name(category, locb);
- charb = strrchr(canonb, '.');
- lenb = charb ? (charb - canonb) : strlen(canonb);
-
- if (lena == lenb && pg_strncasecmp(canona, canonb, lena) == 0)
- {
- pg_free(canona);
- pg_free(canonb);
- return true;
- }
-
- pg_free(canona);
- pg_free(canonb);
- return false;
-}
-
-
static void
check_new_cluster_is_empty(void)
{
@@ -460,35 +367,6 @@ check_new_cluster_is_empty(void)
}
}
-/*
- * Check that every database that already exists in the new cluster is
- * compatible with the corresponding database in the old one.
- */
-static void
-check_databases_are_compatible(void)
-{
- int newdbnum;
- int olddbnum;
- DbInfo *newdbinfo;
- DbInfo *olddbinfo;
-
- for (newdbnum = 0; newdbnum < new_cluster.dbarr.ndbs; newdbnum++)
- {
- newdbinfo = &new_cluster.dbarr.dbs[newdbnum];
-
- /* Find the corresponding database in the old cluster */
- for (olddbnum = 0; olddbnum < old_cluster.dbarr.ndbs; olddbnum++)
- {
- olddbinfo = &old_cluster.dbarr.dbs[olddbnum];
- if (strcmp(newdbinfo->db_name, olddbinfo->db_name) == 0)
- {
- check_locale_and_encoding(olddbinfo, newdbinfo);
- break;
- }
- }
- }
-}
-
/*
* A previous run of pg_upgrade might have failed and the new cluster
* directory recreated, but they might have forgotten to remove
@@ -1524,41 +1402,3 @@ check_for_user_defined_encoding_conversions(ClusterInfo *cluster)
else
check_ok();
}
-
-
-/*
- * get_canonical_locale_name
- *
- * Send the locale name to the system, and hope we get back a canonical
- * version. This should match the backend's check_locale() function.
- */
-static char *
-get_canonical_locale_name(int category, const char *locale)
-{
- char *save;
- char *res;
-
- /* get the current setting, so we can restore it. */
- save = setlocale(category, NULL);
- if (!save)
- pg_fatal("failed to get the current locale");
-
- /* 'save' may be pointing at a modifiable scratch variable, so copy it. */
- save = pg_strdup(save);
-
- /* set the locale with setlocale, to see if it accepts it. */
- res = setlocale(category, locale);
-
- if (!res)
- pg_fatal("failed to get system locale name for \"%s\"", locale);
-
- res = pg_strdup(res);
-
- /* restore old value. */
- if (!setlocale(category, save))
- pg_fatal("failed to restore old locale \"%s\"", save);
-
- pg_free(save);
-
- return res;
-}
diff --git a/src/bin/pg_upgrade/info.c b/src/bin/pg_upgrade/info.c
index c1399c09b9..0490edfc45 100644
--- a/src/bin/pg_upgrade/info.c
+++ b/src/bin/pg_upgrade/info.c
@@ -20,6 +20,7 @@ static void create_rel_filename_map(const char *old_data, const char *new_data,
static void report_unmatched_relation(const RelInfo *rel, const DbInfo *db,
bool is_new_db);
static void free_db_and_rel_infos(DbInfoArr *db_arr);
+static void get_template0_info(ClusterInfo *cluster);
static void get_db_infos(ClusterInfo *cluster);
static void get_rel_infos(ClusterInfo *cluster, DbInfo *dbinfo);
static void free_rel_infos(RelInfoArr *rel_arr);
@@ -278,6 +279,7 @@ get_db_and_rel_infos(ClusterInfo *cluster)
if (cluster->dbarr.dbs != NULL)
free_db_and_rel_infos(&cluster->dbarr);
+ get_template0_info(cluster);
get_db_infos(cluster);
for (dbnum = 0; dbnum < cluster->dbarr.ndbs; dbnum++)
@@ -293,6 +295,54 @@ get_db_and_rel_infos(ClusterInfo *cluster)
}
+/*
+ * Get information about template0, which will be copied from the old cluster
+ * to the new cluster.
+ */
+static void
+get_template0_info(ClusterInfo *cluster)
+{
+ PGconn *conn = connectToServer(cluster, "template1");
+ DbInfo *dbinfo;
+ PGresult *dbres;
+ int i_datencoding;
+ int i_datlocprovider;
+ int i_datcollate;
+ int i_datctype;
+ int i_daticulocale;
+
+ dbres = executeQueryOrDie(conn,
+ "SELECT encoding, datlocprovider, "
+ " datcollate, datctype, daticulocale "
+ "FROM pg_catalog.pg_database "
+ "WHERE datname='template0'");
+
+ if (PQntuples(dbres) != 1)
+ pg_fatal("template0 not found");
+
+ cluster->template0 = pg_malloc(sizeof(DbInfo));
+ dbinfo = cluster->template0;
+
+ i_datencoding = PQfnumber(dbres, "encoding");
+ i_datlocprovider = PQfnumber(dbres, "datlocprovider");
+ i_datcollate = PQfnumber(dbres, "datcollate");
+ i_datctype = PQfnumber(dbres, "datctype");
+ i_daticulocale = PQfnumber(dbres, "daticulocale");
+
+ dbinfo->db_encoding = atoi(PQgetvalue(dbres, 0, i_datencoding));
+ dbinfo->db_collprovider = PQgetvalue(dbres, 0, i_datlocprovider)[0];
+ dbinfo->db_collate = pg_strdup(PQgetvalue(dbres, 0, i_datcollate));
+ dbinfo->db_ctype = pg_strdup(PQgetvalue(dbres, 0, i_datctype));
+ if (PQgetisnull(dbres, 0, i_daticulocale))
+ dbinfo->db_iculocale = NULL;
+ else
+ dbinfo->db_iculocale = pg_strdup(PQgetvalue(dbres, 0, i_daticulocale));
+
+ PQclear(dbres);
+ PQfinish(conn);
+}
+
+
/*
* get_db_infos()
*
diff --git a/src/bin/pg_upgrade/pg_upgrade.c b/src/bin/pg_upgrade/pg_upgrade.c
index e5597d3105..bd005056d1 100644
--- a/src/bin/pg_upgrade/pg_upgrade.c
+++ b/src/bin/pg_upgrade/pg_upgrade.c
@@ -51,6 +51,7 @@
#include "fe_utils/string_utils.h"
#include "pg_upgrade.h"
+static void set_locale_and_encoding(void);
static void prepare_new_cluster(void);
static void prepare_new_globals(void);
static void create_new_objects(void);
@@ -139,6 +140,8 @@ main(int argc, char **argv)
"Performing Upgrade\n"
"------------------");
+ set_locale_and_encoding();
+
prepare_new_cluster();
stop_postmaster(false);
@@ -366,6 +369,65 @@ setup(char *argv0, bool *live_check)
}
+/*
+ * Copy locale and encoding information into the new cluster's template0.
+ *
+ * We need to copy the encoding, datlocprovider, datcollate, datctype, and
+ * daticulocale. We don't need datcollversion because that's never set for
+ * template0.
+ */
+static void
+set_locale_and_encoding(void)
+{
+ PGconn *conn_new_template1;
+ char *datcollate_literal;
+ char *datctype_literal;
+ char *daticulocale_literal = NULL;
+ DbInfo *dbinfo = old_cluster.template0;
+
+ prep_status("Setting locale and encoding for new cluster");
+
+ /* escape literals with respect to new cluster */
+ conn_new_template1 = connectToServer(&new_cluster, "template1");
+
+ datcollate_literal = PQescapeLiteral(conn_new_template1,
+ dbinfo->db_collate,
+ strlen(dbinfo->db_collate));
+ datctype_literal = PQescapeLiteral(conn_new_template1,
+ dbinfo->db_ctype,
+ strlen(dbinfo->db_ctype));
+ if (dbinfo->db_iculocale)
+ daticulocale_literal = PQescapeLiteral(conn_new_template1,
+ dbinfo->db_iculocale,
+ strlen(dbinfo->db_iculocale));
+ else
+ daticulocale_literal = pg_strdup("NULL");
+
+ /* update template0 in new cluster */
+ PQclear(executeQueryOrDie(conn_new_template1,
+ "UPDATE pg_catalog.pg_database "
+ " SET encoding = %u, "
+ " datlocprovider = '%c', "
+ " datcollate = %s, "
+ " datctype = %s, "
+ " daticulocale = %s "
+ " WHERE datname = 'template0' ",
+ dbinfo->db_encoding,
+ dbinfo->db_collprovider,
+ datcollate_literal,
+ datctype_literal,
+ daticulocale_literal));
+
+ PQfreemem(datcollate_literal);
+ PQfreemem(datctype_literal);
+ PQfreemem(daticulocale_literal);
+
+ PQfinish(conn_new_template1);
+
+ check_ok();
+}
+
+
static void
prepare_new_cluster(void)
{
diff --git a/src/bin/pg_upgrade/pg_upgrade.h b/src/bin/pg_upgrade/pg_upgrade.h
index 5f2a116f23..5561fc0473 100644
--- a/src/bin/pg_upgrade/pg_upgrade.h
+++ b/src/bin/pg_upgrade/pg_upgrade.h
@@ -252,6 +252,7 @@ typedef enum
typedef struct
{
ControlData controldata; /* pg_control information */
+ DbInfo *template0; /* template0 info */
DbInfoArr dbarr; /* dbinfos array */
char *pgdata; /* pathname for cluster's $PGDATA directory */
char *pgconfig; /* pathname for cluster's config file
--
2.34.1
From 82073ce781b7f12e210d4534d83840dab691196f Mon Sep 17 00:00:00 2001
From: Jeff Davis <j...@j-davis.com>
Date: Wed, 8 Feb 2023 12:06:26 -0800
Subject: [PATCH v4 3/3] Use ICU by default at initdb time.
If the ICU locale is not specified, initialize the default collator
and retrieve the locale name from that.
Discussion: https://postgr.es/m/510d284759f6e943ce15096167760b2edcb2e700.ca...@j-davis.com
---
contrib/citext/expected/citext_utf8.out | 4 +-
contrib/citext/expected/citext_utf8_1.out | 4 +-
contrib/citext/sql/citext_utf8.sql | 4 +-
contrib/unaccent/expected/unaccent.out | 9 +++
contrib/unaccent/expected/unaccent_1.out | 8 ++
contrib/unaccent/sql/unaccent.sql | 11 +++
doc/src/sgml/ref/initdb.sgml | 48 +++++++-----
src/bin/initdb/Makefile | 4 +-
src/bin/initdb/initdb.c | 76 ++++++++++++++++++-
src/bin/initdb/t/001_initdb.pl | 7 +-
src/bin/pg_dump/t/002_pg_dump.pl | 2 +-
src/bin/scripts/t/020_createdb.pl | 2 +-
src/interfaces/ecpg/test/Makefile | 3 -
src/interfaces/ecpg/test/connect/test5.pgc | 2 +-
.../ecpg/test/expected/connect-test5.c | 2 +-
.../ecpg/test/expected/connect-test5.stderr | 2 +-
src/interfaces/ecpg/test/meson.build | 1 -
src/test/icu/t/010_database.pl | 2 +-
18 files changed, 149 insertions(+), 42 deletions(-)
create mode 100644 contrib/unaccent/expected/unaccent_1.out
diff --git a/contrib/citext/expected/citext_utf8.out b/contrib/citext/expected/citext_utf8.out
index 666b07ccec..85ce9c3b64 100644
--- a/contrib/citext/expected/citext_utf8.out
+++ b/contrib/citext/expected/citext_utf8.out
@@ -3,7 +3,9 @@
* and a Unicode-aware locale.
*/
SELECT getdatabaseencoding() <> 'UTF8' OR
- current_setting('lc_ctype') = 'C'
+ current_setting('lc_ctype') = 'C' OR
+ (SELECT datlocprovider='i' FROM pg_database
+ WHERE datname=current_database())
AS skip_test \gset
\if :skip_test
\quit
diff --git a/contrib/citext/expected/citext_utf8_1.out b/contrib/citext/expected/citext_utf8_1.out
index 433e985349..60ddebd841 100644
--- a/contrib/citext/expected/citext_utf8_1.out
+++ b/contrib/citext/expected/citext_utf8_1.out
@@ -3,7 +3,9 @@
* and a Unicode-aware locale.
*/
SELECT getdatabaseencoding() <> 'UTF8' OR
- current_setting('lc_ctype') = 'C'
+ current_setting('lc_ctype') = 'C' OR
+ (SELECT datlocprovider='i' FROM pg_database
+ WHERE datname=current_database())
AS skip_test \gset
\if :skip_test
\quit
diff --git a/contrib/citext/sql/citext_utf8.sql b/contrib/citext/sql/citext_utf8.sql
index d068000b42..e9c504e764 100644
--- a/contrib/citext/sql/citext_utf8.sql
+++ b/contrib/citext/sql/citext_utf8.sql
@@ -4,7 +4,9 @@
*/
SELECT getdatabaseencoding() <> 'UTF8' OR
- current_setting('lc_ctype') = 'C'
+ current_setting('lc_ctype') = 'C' OR
+ (SELECT datlocprovider='i' FROM pg_database
+ WHERE datname=current_database())
AS skip_test \gset
\if :skip_test
\quit
diff --git a/contrib/unaccent/expected/unaccent.out b/contrib/unaccent/expected/unaccent.out
index ee0ac71a1c..cef98ee60c 100644
--- a/contrib/unaccent/expected/unaccent.out
+++ b/contrib/unaccent/expected/unaccent.out
@@ -1,3 +1,12 @@
+-- unaccent is broken if the default collation is provided by ICU and
+-- LC_CTYPE=C
+SELECT current_setting('lc_ctype') = 'C' AND
+ (SELECT datlocprovider='i' FROM pg_database
+ WHERE datname=current_database())
+ AS skip_test \gset
+\if :skip_test
+\quit
+\endif
CREATE EXTENSION unaccent;
-- must have a UTF8 database
SELECT getdatabaseencoding();
diff --git a/contrib/unaccent/expected/unaccent_1.out b/contrib/unaccent/expected/unaccent_1.out
new file mode 100644
index 0000000000..0a4a3838ab
--- /dev/null
+++ b/contrib/unaccent/expected/unaccent_1.out
@@ -0,0 +1,8 @@
+-- unaccent is broken if the default collation is provided by ICU and
+-- LC_CTYPE=C
+SELECT current_setting('lc_ctype') = 'C' AND
+ (SELECT datlocprovider='i' FROM pg_database
+ WHERE datname=current_database())
+ AS skip_test \gset
+\if :skip_test
+\quit
diff --git a/contrib/unaccent/sql/unaccent.sql b/contrib/unaccent/sql/unaccent.sql
index 3fc0c706be..027dfb964a 100644
--- a/contrib/unaccent/sql/unaccent.sql
+++ b/contrib/unaccent/sql/unaccent.sql
@@ -1,3 +1,14 @@
+
+-- unaccent is broken if the default collation is provided by ICU and
+-- LC_CTYPE=C
+SELECT current_setting('lc_ctype') = 'C' AND
+ (SELECT datlocprovider='i' FROM pg_database
+ WHERE datname=current_database())
+ AS skip_test \gset
+\if :skip_test
+\quit
+\endif
+
CREATE EXTENSION unaccent;
-- must have a UTF8 database
diff --git a/doc/src/sgml/ref/initdb.sgml b/doc/src/sgml/ref/initdb.sgml
index 5b2bdac101..4f37386ea3 100644
--- a/doc/src/sgml/ref/initdb.sgml
+++ b/doc/src/sgml/ref/initdb.sgml
@@ -89,10 +89,28 @@ PostgreSQL documentation
and character set encoding. These can also be set separately for each
database when it is created. <command>initdb</command> determines those
settings for the template databases, which will serve as the default for
- all other databases. By default, <command>initdb</command> uses the
- locale provider <literal>libc</literal>, takes the locale settings from
- the environment, and determines the encoding from the locale settings.
- This is almost always sufficient, unless there are special requirements.
+ all other databases.
+ </para>
+
+ <para>
+ By default, <command>initdb</command> uses the ICU library to provide
+ locale services if the server was built with ICU support; otherwise it uses
+ the <literal>libc</literal> locale provider (see <xref
+ linkend="locale-providers"/>). To choose the specific ICU locale ID to
+ apply, use the option <option>--icu-locale</option>. Note that for
+ implementation reasons and to support legacy code,
+ <command>initdb</command> will still select and initialize libc locale
+ settings when the ICU locale provider is used.
+ </para>
+
+ <para>
+ Alternatively, <command>initdb</command> can use the locale provider
+ <literal>libc</literal>. To select this option, specify
+ <literal>--locale-provider=libc</literal>, or build the server without ICU
+ support. The <literal>libc</literal> locale provider takes the locale
+ settings from the environment, and determines the encoding from the locale
+ settings. This is almost always sufficient, unless there are special
+ requirements.
</para>
<para>
@@ -103,17 +121,6 @@ PostgreSQL documentation
categories can give nonsensical results, so this should be used with care.
</para>
- <para>
- Alternatively, the ICU library can be used to provide locale services.
- (Again, this only sets the default for subsequently created databases.) To
- select this option, specify <literal>--locale-provider=icu</literal>.
- To choose the specific ICU locale ID to apply, use the option
- <option>--icu-locale</option>. Note that
- for implementation reasons and to support legacy code,
- <command>initdb</command> will still select and initialize libc locale
- settings when the ICU locale provider is used.
- </para>
-
<para>
When <command>initdb</command> runs, it will print out the locale settings
it has chosen. If you have complex requirements or specified multiple
@@ -234,7 +241,8 @@ PostgreSQL documentation
<term><option>--icu-locale=<replaceable>locale</replaceable></option></term>
<listitem>
<para>
- Specifies the ICU locale ID, if the ICU locale provider is used.
+ Specifies the ICU locale ID, if the ICU locale provider is used. By
+ default, ICU obtains the ICU locale from the ICU default collator.
</para>
</listitem>
</varlistentry>
@@ -297,10 +305,12 @@ PostgreSQL documentation
<term><option>--locale-provider={<literal>libc</literal>|<literal>icu</literal>}</option></term>
<listitem>
<para>
- This option sets the locale provider for databases created in the
- new cluster. It can be overridden in the <command>CREATE
+ This option sets the locale provider for databases created in the new
+ cluster. It can be overridden in the <command>CREATE
DATABASE</command> command when new databases are subsequently
- created. The default is <literal>libc</literal>.
+ created. The default is <literal>icu</literal> if the server was
+ built with ICU support; otherwise the default is
+ <literal>libc</literal> (see <xref linkend="locale-providers"/>).
</para>
</listitem>
</varlistentry>
diff --git a/src/bin/initdb/Makefile b/src/bin/initdb/Makefile
index eab89c5501..d69bd89572 100644
--- a/src/bin/initdb/Makefile
+++ b/src/bin/initdb/Makefile
@@ -16,7 +16,7 @@ subdir = src/bin/initdb
top_builddir = ../../..
include $(top_builddir)/src/Makefile.global
-override CPPFLAGS := -I$(libpq_srcdir) -I$(top_srcdir)/src/timezone $(CPPFLAGS)
+override CPPFLAGS := -I$(libpq_srcdir) -I$(top_srcdir)/src/timezone $(ICU_CFLAGS) $(CPPFLAGS)
# Note: it's important that we link to encnames.o from libpgcommon, not
# from libpq, else we have risks of version skew if we run with a libpq
@@ -24,7 +24,7 @@ override CPPFLAGS := -I$(libpq_srcdir) -I$(top_srcdir)/src/timezone $(CPPFLAGS)
# should ensure that that happens.
#
# We need libpq only because fe_utils does.
-LDFLAGS_INTERNAL += -L$(top_builddir)/src/fe_utils -lpgfeutils $(libpq_pgport)
+LDFLAGS_INTERNAL += -L$(top_builddir)/src/fe_utils -lpgfeutils $(libpq_pgport) $(ICU_LIBS)
# use system timezone data?
ifneq (,$(with_system_tzdata))
diff --git a/src/bin/initdb/initdb.c b/src/bin/initdb/initdb.c
index 7a58c33ace..0776294499 100644
--- a/src/bin/initdb/initdb.c
+++ b/src/bin/initdb/initdb.c
@@ -53,6 +53,9 @@
#include <netdb.h>
#include <sys/socket.h>
#include <sys/stat.h>
+#ifdef USE_ICU
+#include <unicode/ucol.h>
+#endif
#include <unistd.h>
#include <signal.h>
#include <time.h>
@@ -133,7 +136,11 @@ static char *lc_monetary = NULL;
static char *lc_numeric = NULL;
static char *lc_time = NULL;
static char *lc_messages = NULL;
+#ifdef USE_ICU
+static char locale_provider = COLLPROVIDER_ICU;
+#else
static char locale_provider = COLLPROVIDER_LIBC;
+#endif
static char *icu_locale = NULL;
static const char *default_text_search_config = NULL;
static char *username = NULL;
@@ -2024,6 +2031,72 @@ check_icu_locale_encoding(int user_enc)
return true;
}
+/*
+ * Check that ICU accepts the locale name; or if not specified, retrieve the
+ * default ICU locale.
+ */
+static void
+check_icu_locale()
+{
+#ifdef USE_ICU
+ UCollator *collator;
+ UErrorCode status;
+
+ status = U_ZERO_ERROR;
+ collator = ucol_open(icu_locale, &status);
+ if (U_FAILURE(status))
+ {
+ if (icu_locale)
+ pg_fatal("could not open collator for locale \"%s\": %s",
+ icu_locale, u_errorName(status));
+ else
+ pg_fatal("could not open collator for default locale: %s",
+ u_errorName(status));
+ }
+
+ /* if not specified, get locale from default collator */
+ if (icu_locale == NULL)
+ {
+ const char *default_locale;
+
+ status = U_ZERO_ERROR;
+ default_locale = ucol_getLocaleByType(collator, ULOC_VALID_LOCALE,
+ &status);
+ if (U_FAILURE(status))
+ {
+ ucol_close(collator);
+ pg_fatal("could not determine default ICU locale");
+ }
+
+ if (U_ICU_VERSION_MAJOR_NUM >= 54)
+ {
+ const bool strict = true;
+ char *langtag;
+ int len;
+
+ len = uloc_toLanguageTag(default_locale, NULL, 0, strict, &status);
+ langtag = pg_malloc(len + 1);
+ status = U_ZERO_ERROR;
+ uloc_toLanguageTag(default_locale, langtag, len + 1, strict,
+ &status);
+
+ if (U_FAILURE(status))
+ {
+ ucol_close(collator);
+ pg_fatal("could not determine language tag for default locale \"%s\": %s",
+ default_locale, u_errorName(status));
+ }
+
+ icu_locale = langtag;
+ }
+ else
+ icu_locale = pg_strdup(default_locale);
+ }
+
+ ucol_close(collator);
+#endif
+}
+
/*
* set up the locale variables
*
@@ -2077,8 +2150,7 @@ setlocales(void)
if (locale_provider == COLLPROVIDER_ICU)
{
- if (!icu_locale)
- pg_fatal("ICU locale must be specified");
+ check_icu_locale();
/*
* In supported builds, the ICU locale ID will be checked by the
diff --git a/src/bin/initdb/t/001_initdb.pl b/src/bin/initdb/t/001_initdb.pl
index 772769acab..e5d214e09c 100644
--- a/src/bin/initdb/t/001_initdb.pl
+++ b/src/bin/initdb/t/001_initdb.pl
@@ -97,11 +97,6 @@ SKIP:
if ($ENV{with_icu} eq 'yes')
{
- command_fails_like(
- [ 'initdb', '--no-sync', '--locale-provider=icu', "$tempdir/data2" ],
- qr/initdb: error: ICU locale must be specified/,
- 'locale provider ICU requires --icu-locale');
-
command_ok(
[
'initdb', '--no-sync',
@@ -116,7 +111,7 @@ if ($ENV{with_icu} eq 'yes')
'--locale-provider=icu', '--icu-locale=@colNumeric=lower',
"$tempdir/dataX"
],
- qr/FATAL: could not open collator for locale/,
+ qr/error: could not open collator for locale/,
'fails for invalid ICU locale');
command_fails_like(
diff --git a/src/bin/pg_dump/t/002_pg_dump.pl b/src/bin/pg_dump/t/002_pg_dump.pl
index 72b19ee6cd..0808585866 100644
--- a/src/bin/pg_dump/t/002_pg_dump.pl
+++ b/src/bin/pg_dump/t/002_pg_dump.pl
@@ -1758,7 +1758,7 @@ my %tests = (
create_sql =>
"CREATE DATABASE dump_test2 LOCALE = 'C' TEMPLATE = template0;",
regexp => qr/^
- \QCREATE DATABASE dump_test2 \E.*\QLOCALE = 'C';\E
+ \QCREATE DATABASE dump_test2 \E.*\QLOCALE = 'C'\E
/xm,
like => { pg_dumpall_dbprivs => 1, },
},
diff --git a/src/bin/scripts/t/020_createdb.pl b/src/bin/scripts/t/020_createdb.pl
index 3ad4fbb00c..8ec58cdd64 100644
--- a/src/bin/scripts/t/020_createdb.pl
+++ b/src/bin/scripts/t/020_createdb.pl
@@ -13,7 +13,7 @@ program_version_ok('createdb');
program_options_handling_ok('createdb');
my $node = PostgreSQL::Test::Cluster->new('main');
-$node->init;
+$node->init(extra => ['--locale-provider=libc']);
$node->start;
$node->issues_sql_like(
diff --git a/src/interfaces/ecpg/test/Makefile b/src/interfaces/ecpg/test/Makefile
index d7a7d1d1ca..cf841a3a5b 100644
--- a/src/interfaces/ecpg/test/Makefile
+++ b/src/interfaces/ecpg/test/Makefile
@@ -14,9 +14,6 @@ override CPPFLAGS := \
'-DSHELLPROG="$(SHELL)"' \
$(CPPFLAGS)
-# default encoding for regression tests
-ENCODING = SQL_ASCII
-
ifneq ($(build_os),mingw32)
abs_builddir := $(shell pwd)
else
diff --git a/src/interfaces/ecpg/test/connect/test5.pgc b/src/interfaces/ecpg/test/connect/test5.pgc
index de29160089..d512553677 100644
--- a/src/interfaces/ecpg/test/connect/test5.pgc
+++ b/src/interfaces/ecpg/test/connect/test5.pgc
@@ -55,7 +55,7 @@ exec sql end declare section;
exec sql connect to 'unix:postgresql://localhost/ecpg2_regression' as main user :user USING "connectpw";
exec sql disconnect main;
- exec sql connect to unix:postgresql://localhost/ecpg2_regression?connect_timeout=180&client_encoding=latin1 as main user regress_ecpg_user1/connectpw;
+ exec sql connect to unix:postgresql://localhost/ecpg2_regression?connect_timeout=180&client_encoding=sql_ascii as main user regress_ecpg_user1/connectpw;
exec sql disconnect main;
exec sql connect to "unix:postgresql://200.46.204.71/ecpg2_regression" as main user regress_ecpg_user1/connectpw;
diff --git a/src/interfaces/ecpg/test/expected/connect-test5.c b/src/interfaces/ecpg/test/expected/connect-test5.c
index c1124c627f..ec1514ed9a 100644
--- a/src/interfaces/ecpg/test/expected/connect-test5.c
+++ b/src/interfaces/ecpg/test/expected/connect-test5.c
@@ -117,7 +117,7 @@ main(void)
#line 56 "test5.pgc"
- { ECPGconnect(__LINE__, 0, "unix:postgresql://localhost/ecpg2_regression?connect_timeout=180 & client_encoding=latin1" , "regress_ecpg_user1" , "connectpw" , "main", 0); }
+ { ECPGconnect(__LINE__, 0, "unix:postgresql://localhost/ecpg2_regression?connect_timeout=180 & client_encoding=sql_ascii" , "regress_ecpg_user1" , "connectpw" , "main", 0); }
#line 58 "test5.pgc"
{ ECPGdisconnect(__LINE__, "main");}
diff --git a/src/interfaces/ecpg/test/expected/connect-test5.stderr b/src/interfaces/ecpg/test/expected/connect-test5.stderr
index 01a6a0a13b..51cc18916a 100644
--- a/src/interfaces/ecpg/test/expected/connect-test5.stderr
+++ b/src/interfaces/ecpg/test/expected/connect-test5.stderr
@@ -50,7 +50,7 @@
[NO_PID]: sqlca: code: 0, state: 00000
[NO_PID]: ecpg_finish: connection main closed
[NO_PID]: sqlca: code: 0, state: 00000
-[NO_PID]: ECPGconnect: opening database ecpg2_regression on <DEFAULT> port <DEFAULT> with options connect_timeout=180 & client_encoding=latin1 for user regress_ecpg_user1
+[NO_PID]: ECPGconnect: opening database ecpg2_regression on <DEFAULT> port <DEFAULT> with options connect_timeout=180 & client_encoding=sql_ascii for user regress_ecpg_user1
[NO_PID]: sqlca: code: 0, state: 00000
[NO_PID]: ecpg_finish: connection main closed
[NO_PID]: sqlca: code: 0, state: 00000
diff --git a/src/interfaces/ecpg/test/meson.build b/src/interfaces/ecpg/test/meson.build
index d0be73ccf9..04c6819a79 100644
--- a/src/interfaces/ecpg/test/meson.build
+++ b/src/interfaces/ecpg/test/meson.build
@@ -69,7 +69,6 @@ ecpg_test_files = files(
ecpg_regress_args = [
'--dbname=ecpg1_regression,ecpg2_regression',
'--create-role=regress_ecpg_user1,regress_ecpg_user2',
- '--encoding=SQL_ASCII',
]
tests += {
diff --git a/src/test/icu/t/010_database.pl b/src/test/icu/t/010_database.pl
index 80ab1c7789..45d77c319a 100644
--- a/src/test/icu/t/010_database.pl
+++ b/src/test/icu/t/010_database.pl
@@ -12,7 +12,7 @@ if ($ENV{with_icu} ne 'yes')
}
my $node1 = PostgreSQL::Test::Cluster->new('node1');
-$node1->init;
+$node1->init(extra => ['--locale-provider=libc']);
$node1->start;
$node1->safe_psql('postgres',
--
2.34.1
