*: enable the new collation by default (#32016)

ref #32015
pingcap · Feb 8, 2022 · 3738dec · 3738dec
1 parent 077eb80
commit 3738dec
Show file tree

Hide file tree

Showing 47 changed files with 256 additions and 568 deletions.
diff --git a/cmd/explaintest/disable_new_collation.toml b/cmd/explaintest/disable_new_collation.toml
@@ -15,6 +15,7 @@
 lease = "0"
 mem-quota-query = 34359738368
 host = "127.0.0.1"
+new_collations_enabled_on_first_bootstrap = false
 
 [status]
 status-host = "127.0.0.1"

diff --git a/config/config.go b/config/config.go
@@ -744,10 +744,11 @@ var defaultConf = Config{
 		AutoTLS:                     false,
 		RSAKeySize:                  4096,
 	},
-	DeprecateIntegerDisplayWidth: false,
-	EnableEnumLengthLimit:        true,
-	StoresRefreshInterval:        defTiKVCfg.StoresRefreshInterval,
-	EnableForwarding:             defTiKVCfg.EnableForwarding,
+	DeprecateIntegerDisplayWidth:         false,
+	EnableEnumLengthLimit:                true,
+	StoresRefreshInterval:                defTiKVCfg.StoresRefreshInterval,
+	EnableForwarding:                     defTiKVCfg.EnableForwarding,
+	NewCollationsEnabledOnFirstBootstrap: true,
 }
 
 var (

diff --git a/config/config.toml.example b/config/config.toml.example
@@ -109,7 +109,7 @@ repair-table-list = []
 max-server-connections = 0
 
 # Whether new collations are enabled, as indicated by its name, this configuration entry take effect ONLY when a TiDB cluster bootstraps for the first time.
-new_collations_enabled_on_first_bootstrap = false
+new_collations_enabled_on_first_bootstrap = true
 
 # Don't register information of this TiDB to etcd, so this instance of TiDB won't appear in the services like dashboard.
 # This option is useful when you want to embed TiDB into your service(i.e. use TiDB as a library).

diff --git a/ddl/column_test.go b/ddl/column_test.go
@@ -37,7 +37,6 @@ import (
 	"github.com/pingcap/tidb/table/tables"
 	"github.com/pingcap/tidb/tablecodec"
 	"github.com/pingcap/tidb/types"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/stretchr/testify/require"
 	"github.com/stretchr/testify/suite"
 )
@@ -1171,9 +1170,6 @@ func (s *testColumnSuiteToVerify) TestDropColumns() {
 }
 
 func TestModifyColumn(t *testing.T) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
-
 	store, err := mockstore.NewMockStore()
 	require.NoError(t, err)
 	d, err := testNewDDLAndStart(

diff --git a/ddl/column_type_change_test.go b/ddl/column_type_change_test.go
@@ -41,7 +41,6 @@ import (
 	"github.com/pingcap/tidb/table"
 	"github.com/pingcap/tidb/table/tables"
 	"github.com/pingcap/tidb/tablecodec"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/dbterror"
 	"github.com/pingcap/tidb/util/testkit"
 )
@@ -440,8 +439,6 @@ func (s *testColumnTypeChangeSuite) TestColumnTypeChangeFromIntegerToOthers(c *C
 func (s *testColumnTypeChangeSuite) TestColumnTypeChangeBetweenVarcharAndNonVarchar(c *C) {
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test")
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk.MustExec("drop database if exists col_type_change_char;")
 	tk.MustExec("create database col_type_change_char;")
 	tk.MustExec("use col_type_change_char;")

diff --git a/ddl/db_integration_test.go b/ddl/db_integration_test.go
@@ -956,11 +956,11 @@ func (s *testIntegrationSuite4) TestChangingTableCharset(c *C) {
 
 	tk.MustExec("drop table t")
 	tk.MustExec("create table t(a varchar(5) charset utf8 collate utf8_unicode_ci) charset utf8 collate utf8_unicode_ci")
-	tk.MustExec("alter table t collate utf8_danish_ci")
+	tk.MustExec("alter table t collate utf8_general_ci")
 	tbl = testGetTableByName(c, s.ctx, "test", "t")
 	c.Assert(tbl, NotNil)
 	c.Assert(tbl.Meta().Charset, Equals, "utf8")
-	c.Assert(tbl.Meta().Collate, Equals, "utf8_danish_ci")
+	c.Assert(tbl.Meta().Collate, Equals, "utf8_general_ci")
 	for _, col := range tbl.Meta().Columns {
 		c.Assert(col.Charset, Equals, "utf8")
 		// Column collate should remain unchanged.
@@ -2836,7 +2836,7 @@ func (s *testIntegrationSuite3) TestStrictDoubleTypeCheck(c *C) {
 }
 
 func (s *testSerialDBSuite) TestDuplicateErrorMessage(c *C) {
-	defer collate.SetNewCollationEnabledForTest(false)
+	defer collate.SetNewCollationEnabledForTest(true)
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test")
 

diff --git a/ddl/db_partition_test.go b/ddl/db_partition_test.go
@@ -48,7 +48,6 @@ import (
 	ntestkit "github.com/pingcap/tidb/testkit"
 	"github.com/pingcap/tidb/types"
 	"github.com/pingcap/tidb/util/admin"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/israce"
 	"github.com/pingcap/tidb/util/logutil"
 	"github.com/pingcap/tidb/util/mock"
@@ -367,8 +366,6 @@ func (s *testIntegrationSuite2) TestCreateTableWithHashPartition(c *C) {
 }
 
 func (s *testSerialDBSuite1) TestCreateTableWithRangeColumnPartition(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test;")
 	tk.MustExec("drop table if exists log_message_1;")
@@ -3388,8 +3385,6 @@ func (s *testSerialDBSuite1) TestPartitionListWithTimeType(c *C) {
 }
 
 func (s *testSerialDBSuite1) TestPartitionListWithNewCollation(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk := testkit.NewTestKitWithInit(c, s.store)
 	tk.MustExec("use test;")
 	tk.MustExec("drop table if exists t;")

diff --git a/ddl/db_test.go b/ddl/db_test.go
@@ -58,7 +58,6 @@ import (
 	"github.com/pingcap/tidb/util"
 	"github.com/pingcap/tidb/util/admin"
 	"github.com/pingcap/tidb/util/codec"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/domainutil"
 	"github.com/pingcap/tidb/util/israce"
 	"github.com/pingcap/tidb/util/mock"
@@ -2867,8 +2866,6 @@ func (s *testSerialDBSuite) TestCreateTable(c *C) {
 	tk.MustExec("use test")
 	failSQL := "create table t_enum (a enum('e','e'));"
 	tk.MustGetErrCode(failSQL, errno.ErrDuplicatedValueInType)
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk = testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test")
 	failSQL = "create table t_enum (a enum('e','E')) charset=utf8 collate=utf8_general_ci;"
@@ -3173,8 +3170,8 @@ func (s *testDBSuite2) TestCreateTableWithSetCol(c *C) {
 	tk.MustGetErrCode(failedSQL, errno.ErrInvalidDefault)
 	failedSQL = "create table t_set (a set('1', '4', '10') default '1,4,11');"
 	tk.MustGetErrCode(failedSQL, errno.ErrInvalidDefault)
-	failedSQL = "create table t_set (a set('1', '4', '10') default '1 ,4');"
-	tk.MustGetErrCode(failedSQL, errno.ErrInvalidDefault)
+	// Success when the new collation is enabled.
+	tk.MustExec("create table t_set (a set('1', '4', '10') default '1 ,4');")
 	// The type of default value is int.
 	failedSQL = "create table t_set (a set('1', '4', '10') default 0);"
 	tk.MustGetErrCode(failedSQL, errno.ErrInvalidDefault)
@@ -3183,6 +3180,7 @@ func (s *testDBSuite2) TestCreateTableWithSetCol(c *C) {
 
 	// The type of default value is int.
 	// It's for successful cases
+	tk.MustExec("drop table if exists t_set")
 	tk.MustExec("create table t_set (a set('1', '4', '10', '21') default 1);")
 	tk.MustQuery("show create table t_set").Check(testkit.Rows("t_set CREATE TABLE `t_set` (\n" +
 		"  `a` set('1','4','10','21') DEFAULT '1'\n" +

diff --git a/ddl/serial_test.go b/ddl/serial_test.go
@@ -45,7 +45,6 @@ import (
 	"github.com/pingcap/tidb/store/mockstore"
 	"github.com/pingcap/tidb/tablecodec"
 	"github.com/pingcap/tidb/util/admin"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/gcutil"
 	"github.com/pingcap/tidb/util/mock"
 	"github.com/pingcap/tidb/util/testkit"
@@ -1521,9 +1520,6 @@ func (s *testSerialSuite) TestAutoRandomWithPreSplitRegion(c *C) {
 }
 
 func (s *testSerialSuite) TestModifyingColumn4NewCollations(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
-
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("create database dct")
 	tk.MustExec("use dct")
@@ -1557,8 +1553,6 @@ func (s *testSerialSuite) TestModifyingColumn4NewCollations(c *C) {
 }
 
 func (s *testSerialSuite) TestForbidUnsupportedCollations(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk := testkit.NewTestKit(c, s.store)
 
 	mustGetUnsupportedCollation := func(sql string, coll string) {

diff --git a/executor/analyze_test.go b/executor/analyze_test.go
@@ -1108,8 +1108,6 @@ func testAnalyzeIncremental(tk *testkit.TestKit, t *testing.T, dom *domain.Domai
 }
 
 func TestIssue20874(t *testing.T) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	store, clean := testkit.CreateMockStore(t)
 	defer clean()
 

diff --git a/executor/collation_test.go b/executor/collation_test.go
@@ -21,15 +21,11 @@ import (
 	"github.com/pingcap/tidb/parser/mysql"
 	"github.com/pingcap/tidb/types"
 	"github.com/pingcap/tidb/util/chunk"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/mock"
 	"github.com/stretchr/testify/require"
 )
 
 func TestVecGroupChecker(t *testing.T) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
-
 	tp := &types.FieldType{Tp: mysql.TypeVarchar}
 	col0 := &expression.Column{
 		RetType: tp,

diff --git a/executor/executor_test.go b/executor/executor_test.go
@@ -73,7 +73,6 @@ import (
 	"github.com/pingcap/tidb/types"
 	"github.com/pingcap/tidb/util"
 	"github.com/pingcap/tidb/util/admin"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/deadlockhistory"
 	"github.com/pingcap/tidb/util/gcutil"
 	"github.com/pingcap/tidb/util/israce"
@@ -322,14 +321,6 @@ func (s *testSuiteP1) TestShow(c *C) {
 	c.Assert(len(tk.MustQuery("show index in t").Rows()), Equals, 1)
 	c.Assert(len(tk.MustQuery("show index from t").Rows()), Equals, 1)
 
-	tk.MustQuery("show charset").Check(testkit.Rows(
-		"ascii US ASCII ascii_bin 1",
-		"binary binary binary 1",
-		"gbk Chinese Internal Code Specification gbk_bin 2",
-		"latin1 Latin1 latin1_bin 1",
-		"utf8 UTF-8 Unicode utf8_bin 3",
-		"utf8mb4 UTF-8 Unicode utf8mb4_bin 4",
-	))
 	c.Assert(len(tk.MustQuery("show master status").Rows()), Equals, 1)
 	tk.MustQuery("show create database test_show").Check(testkit.Rows("test_show CREATE DATABASE `test_show` /*!40100 DEFAULT CHARACTER SET utf8mb4 */"))
 	tk.MustQuery("show privileges").Check(testkit.Rows("Alter Tables To alter the table",
@@ -5730,54 +5721,9 @@ func (s *testSerialSuite2) TestUnsignedFeedback(c *C) {
 	c.Assert(result.Rows()[2][6], Equals, "keep order:false")
 }
 
-func (s *testSuiteWithCliBaseCharset) TestCharsetFeatureWithoutNewCollation(c *C) {
-	tk := testkit.NewTestKit(c, s.store)
-	tk.MustExec("use test")
-	tk.MustQuery("show charset").Check(testkit.Rows(
-		"ascii US ASCII ascii_bin 1",
-		"binary binary binary 1",
-		"gbk Chinese Internal Code Specification gbk_bin 2",
-		"latin1 Latin1 latin1_bin 1",
-		"utf8 UTF-8 Unicode utf8_bin 3",
-		"utf8mb4 UTF-8 Unicode utf8mb4_bin 4",
-	))
-	tk.MustQuery("show collation").Check(testkit.Rows(
-		"utf8mb4_bin utf8mb4 46 Yes Yes 1",
-		"latin1_bin latin1 47 Yes Yes 1",
-		"binary binary 63 Yes Yes 1",
-		"ascii_bin ascii 65 Yes Yes 1",
-		"utf8_bin utf8 83 Yes Yes 1",
-		"gbk_bin gbk 87 Yes Yes 1",
-	))
-	tk.MustExec("create table t5(a char(20), b char(20) charset utf8, c binary) charset gbk collate gbk_bin;")
-}
-
 func (s *testSuiteWithCliBaseCharset) TestCharsetFeature(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test")
-	tk.MustQuery("show charset").Check(testkit.Rows(
-		"ascii US ASCII ascii_bin 1",
-		"binary binary binary 1",
-		"gbk Chinese Internal Code Specification gbk_chinese_ci 2",
-		"latin1 Latin1 latin1_bin 1",
-		"utf8 UTF-8 Unicode utf8_bin 3",
-		"utf8mb4 UTF-8 Unicode utf8mb4_bin 4",
-	))
-	tk.MustQuery("show collation").Check(testkit.Rows(
-		"ascii_bin ascii 65 Yes Yes 1",
-		"binary binary 63 Yes Yes 1",
-		"gbk_bin gbk 87  Yes 1",
-		"gbk_chinese_ci gbk 28 Yes Yes 1",
-		"latin1_bin latin1 47 Yes Yes 1",
-		"utf8_bin utf8 83 Yes Yes 1",
-		"utf8_general_ci utf8 33  Yes 1",
-		"utf8_unicode_ci utf8 192  Yes 1",
-		"utf8mb4_bin utf8mb4 46 Yes Yes 1",
-		"utf8mb4_general_ci utf8mb4 45  Yes 1",
-		"utf8mb4_unicode_ci utf8mb4 224  Yes 1",
-	))
 
 	tk.MustExec("set names gbk;")
 	tk.MustQuery("select @@character_set_connection;").Check(testkit.Rows("gbk"))
@@ -5820,8 +5766,6 @@ func (s *testSuiteWithCliBaseCharset) TestCharsetFeature(c *C) {
 }
 
 func (s *testSuiteWithCliBaseCharset) TestCharsetFeatureCollation(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test")
 	tk.MustExec("drop table if exists t;")
@@ -5847,8 +5791,6 @@ func (s *testSuiteWithCliBaseCharset) TestCharsetFeatureCollation(c *C) {
 }
 
 func (s *testSuiteWithCliBaseCharset) TestCharsetWithPrefixIndex(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test")
 	tk.MustExec("drop table if exists t")

diff --git a/executor/insert_test.go b/executor/insert_test.go
@@ -30,7 +30,6 @@ import (
 	"github.com/pingcap/tidb/table"
 	"github.com/pingcap/tidb/types"
 	"github.com/pingcap/tidb/util"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/execdetails"
 	"github.com/pingcap/tidb/util/israce"
 	"github.com/pingcap/tidb/util/testkit"
@@ -1496,9 +1495,6 @@ func (s *testSuite10) TestInsertRuntimeStat(c *C) {
 }
 
 func (s *testSerialSuite) TestDuplicateEntryMessage(c *C) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
-
 	tk := testkit.NewTestKit(c, s.store)
 	tk.MustExec("use test;")
 	for _, enable := range []variable.ClusteredIndexDefMode{variable.ClusteredIndexDefModeOn, variable.ClusteredIndexDefModeOff, variable.ClusteredIndexDefModeIntOnly} {

diff --git a/executor/parallel_apply_test.go b/executor/parallel_apply_test.go
@@ -392,7 +392,6 @@ func (s *testSerialSuite) TestApplyWithOtherFeatures(c *C) {
 	tk.MustExec("set tidb_enable_parallel_apply=true")
 
 	// collation 1
-	collate.SetNewCollationEnabledForTest(true)
 	tk.MustExec("drop table if exists t, t1")
 	tk.MustExec("create table t(a varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci, b int)")
 	tk.MustExec("create table t1(a varchar(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci, b int)")
@@ -405,6 +404,7 @@ func (s *testSerialSuite) TestApplyWithOtherFeatures(c *C) {
 	sql = "select (select min(t1.b) from t1 where t1.a >= t.a and t1.b >= t.b), (select sum(t1.b) from t1 where t1.a >= t.a and t1.b >= t.b) from t"
 	tk.MustQuery(sql).Sort().Check(testkit.Rows("1 10", "2 9", "3 7", "4 4"))
 	collate.SetNewCollationEnabledForTest(false)
+	defer collate.SetNewCollationEnabledForTest(true)
 
 	// plan cache
 	orgEnable := core.PreparedPlanCacheEnabled()

diff --git a/executor/seqtest/seq_executor_test.go b/executor/seqtest/seq_executor_test.go
@@ -50,7 +50,6 @@ import (
 	"github.com/pingcap/tidb/store/mockstore"
 	"github.com/pingcap/tidb/tablecodec"
 	"github.com/pingcap/tidb/testkit"
-	"github.com/pingcap/tidb/util/collate"
 	"github.com/pingcap/tidb/util/gcutil"
 	"github.com/pingcap/tidb/util/testutil"
 	"github.com/stretchr/testify/require"
@@ -1194,8 +1193,6 @@ func TestCoprocessorPriority(t *testing.T) {
 }
 
 func TestShowForNewCollations(t *testing.T) {
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
 	store, clean := testkit.CreateMockStore(t)
 	defer clean()
 
@@ -1221,9 +1218,6 @@ func TestForbidUnsupportedCollations(t *testing.T) {
 	store, clean := testkit.CreateMockStore(t)
 	defer clean()
 
-	collate.SetNewCollationEnabledForTest(true)
-	defer collate.SetNewCollationEnabledForTest(false)
-
 	tk := testkit.NewTestKit(t, store)
 	mustGetUnsupportedCollation := func(sql string, coll string) {
 		tk.MustGetErrMsg(sql, fmt.Sprintf("[ddl:1273]Unsupported collation when new collation is enabled: '%s'", coll))