initdb: derive encoding from locale for ICU; similar to libc.

Previously, the default encoding was derived from the locale when
using libc; while the default was always UTF-8 when using ICU. That
would throw an error when the locale was not compatible with UTF-8.

This commit causes initdb to derive the default encoding from the
locale for both providers. If --no-locale is specified (or if the
locale is C or POSIX), the default encoding will be UTF-8 for ICU
(because ICU does not support SQL_ASCII) and SQL_ASCII for libc.

Per buildfarm failure on system "hoverfly" related to commit
27b6237.

Discussion: https://postgr.es/m/d191d5841347301a8f1238f609471ddd957fc47e.camel%40j-davis.com

Author: jeff-davis

contrib/unaccent/meson.build

@@ -37,6 +37,6 @@ tests += {
     'sql': [
       'unaccent',
     ],
-    'regress_args': ['--encoding=UTF8'],
+    'regress_args': ['--encoding=UTF8', '--no-locale'],
   },
 }

doc/src/sgml/ref/initdb.sgml

@@ -213,13 +213,19 @@ PostgreSQL documentation
       <term><option>--encoding=<replaceable class="parameter">encoding</replaceable></option></term>
       <listitem>
        <para>
-        Selects the encoding of the template databases. This will also
-        be the default encoding of any database you create later,
-        unless you override it then.  The default is derived from the locale,
-        if the libc locale provider is used, or <literal>UTF8</literal> if the
-        ICU locale provider is used.  The character sets supported by
-        the <productname>PostgreSQL</productname> server are described
-        in <xref linkend="multibyte-charset-supported"/>.
+        Selects the encoding of the template databases. This will also be the
+        default encoding of any database you create later, unless you override
+        it then. The character sets supported by the
+        <productname>PostgreSQL</productname> server are described in <xref
+        linkend="multibyte-charset-supported"/>.
+       </para>
+       <para>
+        By default, the template database encoding is derived from the
+        locale. If <xref linkend="app-initdb-option-no-locale"/> is specified
+        (or equivalently, if the locale is <literal>C</literal> or
+        <literal>POSIX</literal>), then the default is <literal>UTF8</literal>
+        for the ICU provider and <literal>SQL_ASCII</literal> for the
+        <literal>libc</literal> provider.
        </para>
       </listitem>
      </varlistentry>

src/bin/initdb/initdb.c

@@ -2350,18 +2350,19 @@ setup_locale_encoding(void)
 			   lc_time);
 	}
 
-	if (!encoding && locale_provider == COLLPROVIDER_ICU)
-	{
-		encodingid = PG_UTF8;
-		printf(_("The default database encoding has been set to \"%s\".\n"),
-			   pg_encoding_to_char(encodingid));
-	}
-	else if (!encoding)
+	if (!encoding)
 	{
 		int			ctype_enc;
 
 		ctype_enc = pg_get_encoding_from_locale(lc_ctype, true);
 
+		/*
+		 * If ctype_enc=SQL_ASCII, it's compatible with any encoding. ICU does
+		 * not support SQL_ASCII, so select UTF-8 instead.
+		 */
+		if (locale_provider == COLLPROVIDER_ICU && ctype_enc == PG_SQL_ASCII)
+			ctype_enc = PG_UTF8;
+
 		if (ctype_enc == -1)
 		{
 			/* Couldn't recognize the locale's codeset */

src/bin/pg_upgrade/t/002_pg_upgrade.pl

@@ -108,7 +108,7 @@ sub filter_dump
 
 my $original_encoding = "6"; # UTF-8
 my $original_provider = "c";
-my $original_collate = "C";
+my $original_locale = "C";
 my $original_iculocale = "";
 my $provider_field = "'c' AS datlocprovider";
 my $iculocale_field = "NULL AS daticulocale";
@@ -123,7 +123,7 @@ sub filter_dump
 my @initdb_params = @custom_opts;
 
 push @initdb_params, ('--encoding', 'UTF-8');
-push @initdb_params, ('--lc-collate', $original_collate);
+push @initdb_params, ('--locale', $original_locale);
 if ($original_provider eq "i")
 {
 	push @initdb_params, ('--locale-provider', 'icu');
@@ -136,16 +136,12 @@ sub filter_dump
 
 my $result;
 $result = $oldnode->safe_psql(
-	'postgres', "SELECT encoding, $provider_field, datcollate, $iculocale_field
+	'postgres', "SELECT encoding, $provider_field, datcollate, datctype, $iculocale_field
                  FROM pg_database WHERE datname='template0'");
-is($result, "$original_encoding|$original_provider|$original_collate|$original_iculocale",
+is($result, "$original_encoding|$original_provider|$original_locale|$original_locale|$original_iculocale",
 		"check locales in original cluster"
 	);
 
-# check ctype, which was acquired from environment by initdb
-my $original_ctype = $oldnode->safe_psql(
-	'postgres', q{SELECT datctype FROM pg_database WHERE datname='template0'});
-
 # The default location of the source code is the root of this directory.
 my $srcdir = abs_path("../../..");
 
@@ -224,7 +220,6 @@ sub filter_dump
 # cluster.
 push @initdb_params, ('--encoding', 'SQL_ASCII');
 push @initdb_params, ('--locale-provider', 'libc');
-push @initdb_params, ('--lc-ctype', 'C');
 
 $node_params{extra} = \@initdb_params;
 $newnode->init(%node_params);
@@ -401,7 +396,7 @@ sub filter_dump
 $result = $newnode->safe_psql(
 	'postgres', "SELECT encoding, $provider_field, datcollate, datctype, $iculocale_field
                  FROM pg_database WHERE datname='template0'");
-is($result, "$original_encoding|$original_provider|$original_collate|$original_ctype|$original_iculocale",
+is($result, "$original_encoding|$original_provider|$original_locale|$original_locale|$original_iculocale",
 		"check that locales in new cluster match original cluster"
 	);
 

src/bin/scripts/t/020_createdb.pl

@@ -41,7 +41,7 @@
 		[
 			'createdb',        '-T',
 			'template0',       '-E', 'UTF8', '--locale-provider=icu',
-			'--icu-locale=en', 'foobar5'
+			'--locale=C',      '--icu-locale=en', 'foobar5'
 		],
 		qr/statement: CREATE DATABASE foobar5 .* LOCALE_PROVIDER icu ICU_LOCALE 'en'/,
 		'create database with ICU locale specified');

src/test/icu/t/010_database.pl

@@ -54,7 +54,7 @@
 # Test error cases in CREATE DATABASE involving locale-related options
 
 my ($ret, $stdout, $stderr) = $node1->psql('postgres',
-	q{CREATE DATABASE dbicu LOCALE_PROVIDER icu TEMPLATE template0 ENCODING UTF8});
+	q{CREATE DATABASE dbicu LOCALE_PROVIDER icu LOCALE 'C' TEMPLATE template0 ENCODING UTF8});
 isnt($ret, 0,
 	"ICU locale must be specified for ICU provider: exit code not 0");
 like(