ddl: resolve table charset option from collation option (pingcap#13506)

ddl/db_integration_test.go

@@ -614,7 +614,7 @@ func (s *testIntegrationSuite3) TestChangingCharsetToUtf8(c *C) {
 	tk.MustExec("use test")
 	tk.MustExec("create table t1(a varchar(20) charset utf8)")
 	tk.MustExec("insert into t1 values (?)", "t1_value")
-
+	tk.MustExec("alter table t1 collate uTf8mB4_uNiCoDe_Ci charset Utf8mB4 charset uTF8Mb4 collate UTF8MB4_BiN")
 	tk.MustExec("alter table t1 modify column a varchar(20) charset utf8mb4")
 	tk.MustQuery("select * from t1;").Check(testkit.Rows("t1_value"))
 
@@ -624,34 +624,14 @@ func (s *testIntegrationSuite3) TestChangingCharsetToUtf8(c *C) {
 	tk.MustExec("alter table t modify column a varchar(20) charset latin1")
 	tk.MustQuery("select * from t;").Check(testkit.Rows("t_value"))
 
-	rs, err := tk.Exec("alter table t modify column a varchar(20) charset utf8")
-	if rs != nil {
-		rs.Close()
-	}
-	c.Assert(err, NotNil)
-	c.Assert(err.Error(), Equals, "[ddl:8200]Unsupported modify charset from latin1 to utf8")
-	rs, err = tk.Exec("alter table t modify column a varchar(20) charset utf8mb4")
-	if rs != nil {
-		rs.Close()
-	}
-	c.Assert(err, NotNil)
-	c.Assert(err.Error(), Equals, "[ddl:8200]Unsupported modify charset from latin1 to utf8mb4")
+	tk.MustGetErrCode("alter table t modify column a varchar(20) charset utf8", mysql.ErrUnsupportedDDLOperation)
+	tk.MustGetErrCode("alter table t modify column a varchar(20) charset utf8mb4", mysql.ErrUnsupportedDDLOperation)
+	tk.MustGetErrCode("alter table t modify column a varchar(20) charset utf8 collate utf8_bin", mysql.ErrUnsupportedDDLOperation)
+	tk.MustGetErrCode("alter table t modify column a varchar(20) charset utf8mb4 collate utf8mb4_general_ci", mysql.ErrUnsupportedDDLOperation)
 
-	rs, err = tk.Exec("alter table t modify column a varchar(20) charset utf8mb4 collate utf8bin")
-	if rs != nil {
-		rs.Close()
-	}
-	c.Assert(err, NotNil)
-	rs, err = tk.Exec("alter table t modify column a varchar(20) charset utf8 collate utf8_bin")
-	if rs != nil {
-		rs.Close()
-	}
-	c.Assert(err, NotNil)
-	rs, err = tk.Exec("alter table t modify column a varchar(20) charset utf8mb4 collate utf8mb4_general_ci")
-	if rs != nil {
-		rs.Close()
-	}
-	c.Assert(err, NotNil)
+	tk.MustGetErrCode("alter table t modify column a varchar(20) charset utf8mb4 collate utf8bin", mysql.ErrUnknownCollation)
+	tk.MustGetErrCode("alter table t collate LATIN1_GENERAL_CI charset utf8 collate utf8_bin", mysql.ErrConflictingDeclarations)
+	tk.MustGetErrCode("alter table t collate LATIN1_GENERAL_CI collate UTF8MB4_UNICODE_ci collate utf8_bin", mysql.ErrCollationCharsetMismatch)
 }
 
 func (s *testIntegrationSuite4) TestChangingTableCharset(c *C) {

ddl/db_test.go

@@ -1953,21 +1953,24 @@ func (s *testDBSuite1) TestCreateTable(c *C) {
 
 	s.tk.MustExec("drop table t")
 
-	_, err = s.tk.Exec("CREATE TABLE `t` (`a` int) DEFAULT CHARSET=abcdefg")
-	c.Assert(err, NotNil)
+	s.tk.MustGetErrCode("CREATE TABLE `t` (`a` int) DEFAULT CHARSET=abcdefg", mysql.ErrUnknownCharacterSet)
 
-	_, err = s.tk.Exec("CREATE TABLE `collateTest` (`a` int, `b` varchar(10)) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_slovak_ci")
-	c.Assert(err, IsNil)
-	result := s.tk.MustQuery("show create table collateTest")
-	got := result.Rows()[0][1]
-	c.Assert(got, Equals, "CREATE TABLE `collateTest` (\n  `a` int(11) DEFAULT NULL,\n  `b` varchar(10) COLLATE utf8_slovak_ci DEFAULT NULL\n) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_slovak_ci")
+	s.tk.MustExec("CREATE TABLE `collateTest` (`a` int, `b` varchar(10)) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_slovak_ci")
+	expects := "CREATE TABLE `collateTest` (\n  `a` int(11) DEFAULT NULL,\n  `b` varchar(10) COLLATE utf8_slovak_ci DEFAULT NULL\n) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_slovak_ci"
+	s.tk.MustQuery("show create table collateTest").Check(testkit.Rows(expects))
+
+	s.tk.MustGetErrCode("CREATE TABLE `collateTest2` (`a` int) CHARSET utf8 COLLATE utf8mb4_unicode_ci", mysql.ErrCollationCharsetMismatch)
+	s.tk.MustGetErrCode("CREATE TABLE `collateTest3` (`a` int) COLLATE utf8mb4_unicode_ci CHARSET utf8", mysql.ErrConflictingDeclarations)
+
+	s.tk.MustExec("CREATE TABLE `collateTest4` (`a` int) COLLATE utf8_uniCOde_ci")
+	expects = "CREATE TABLE `collateTest4` (\n  `a` int(11) DEFAULT NULL\n) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci"
+	s.tk.MustQuery("show create table collateTest4").Check(testkit.Rows(expects))
 
 	s.tk.MustExec("create database test2 default charset utf8 collate utf8_general_ci")
 	s.tk.MustExec("use test2")
 	s.tk.MustExec("create table dbCollateTest (a varchar(10))")
-	result = s.tk.MustQuery("show create table dbCollateTest")
-	got = result.Rows()[0][1]
-	c.Assert(got, Equals, "CREATE TABLE `dbCollateTest` (\n  `a` varchar(10) COLLATE utf8_general_ci DEFAULT NULL\n) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_general_ci")
+	expects = "CREATE TABLE `dbCollateTest` (\n  `a` varchar(10) COLLATE utf8_general_ci DEFAULT NULL\n) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_general_ci"
+	s.tk.MustQuery("show create table dbCollateTest").Check(testkit.Rows(expects))
 
 	// test for enum column
 	s.tk.MustExec("use test")

ddl/ddl_api.go

@@ -1309,7 +1309,11 @@ func buildTableInfoWithCheck(ctx sessionctx.Context, d *ddl, s *ast.CreateTableS
 		return nil, errors.Trace(err)
 	}
 
-	tableCharset, tableCollate := findTableOptionCharsetAndCollate(s.Options)
+	tableCharset, tableCollate, err := getCharsetAndCollateInTableOption(0, s.Options)
+	if err != nil {
+		return nil, err
+	}
+
 	// The column charset haven't been resolved here.
 	cols, newConstraints, err := buildColumnsAndConstraints(ctx, colDefs, s.Constraints, tableCharset, tableCollate, dbCharset, dbCollate)
 	if err != nil {
@@ -1326,6 +1330,8 @@ func buildTableInfoWithCheck(ctx sessionctx.Context, d *ddl, s *ast.CreateTableS
 	if err != nil {
 		return nil, errors.Trace(err)
 	}
+	tbInfo.Collate = tableCollate
+	tbInfo.Charset = tableCharset
 
 	pi, err := buildTablePartitionInfo(ctx, d, s)
 	if err != nil {
@@ -1348,7 +1354,6 @@ func buildTableInfoWithCheck(ctx sessionctx.Context, d *ddl, s *ast.CreateTableS
 		tbInfo.Partition = pi
 	}
 
-	// The specified charset will be handled in handleTableOptions
 	if err = handleTableOptions(s.Options, tbInfo); err != nil {
 		return nil, errors.Trace(err)
 	}
@@ -1759,32 +1764,6 @@ func resolveDefaultTableCharsetAndCollation(tbInfo *model.TableInfo, dbCharset,
 	return
 }
 
-func findTableOptionCharsetAndCollate(options []*ast.TableOption) (tableCharset, tableCollate string) {
-	var findCnt int
-	for i := len(options) - 1; i >= 0; i-- {
-		op := options[i]
-		if len(tableCharset) == 0 && op.Tp == ast.TableOptionCharset {
-			// find the last one.
-			tableCharset = op.StrValue
-			findCnt++
-			if findCnt == 2 {
-				break
-			}
-			continue
-		}
-		if len(tableCollate) == 0 && op.Tp == ast.TableOptionCollate {
-			// find the last one.
-			tableCollate = op.StrValue
-			findCnt++
-			if findCnt == 2 {
-				break
-			}
-			continue
-		}
-	}
-	return tableCharset, tableCollate
-}
-
 // handleTableOptions updates tableInfo according to table options.
 func handleTableOptions(options []*ast.TableOption, tbInfo *model.TableInfo) error {
 	for _, op := range options {
@@ -1793,10 +1772,6 @@ func handleTableOptions(options []*ast.TableOption, tbInfo *model.TableInfo) err
 			tbInfo.AutoIncID = int64(op.UintValue)
 		case ast.TableOptionComment:
 			tbInfo.Comment = op.StrValue
-		case ast.TableOptionCharset:
-			tbInfo.Charset = op.StrValue
-		case ast.TableOptionCollate:
-			tbInfo.Collate = op.StrValue
 		case ast.TableOptionCompression:
 			tbInfo.Compression = op.StrValue
 		case ast.TableOptionShardRowID:
@@ -1815,7 +1790,6 @@ func handleTableOptions(options []*ast.TableOption, tbInfo *model.TableInfo) err
 	if tbInfo.PreSplitRegions > tbInfo.ShardRowIDBits {
 		tbInfo.PreSplitRegions = tbInfo.ShardRowIDBits
 	}
-
 	return nil
 }
 
@@ -1829,40 +1803,37 @@ func isIgnorableSpec(tp ast.AlterTableType) bool {
 // getCharsetAndCollateInTableOption will iterate the charset and collate in the options,
 // and returns the last charset and collate in options. If there is no charset in the options,
 // the returns charset will be "", the same as collate.
-func getCharsetAndCollateInTableOption(startIdx int, options []*ast.TableOption) (ca, co string, err error) {
-	charsets := make([]string, 0, len(options))
-	collates := make([]string, 0, len(options))
+func getCharsetAndCollateInTableOption(startIdx int, options []*ast.TableOption) (chs, coll string, err error) {
 	for i := startIdx; i < len(options); i++ {
 		opt := options[i]
 		// we set the charset to the last option. example: alter table t charset latin1 charset utf8 collate utf8_bin;
 		// the charset will be utf8, collate will be utf8_bin
 		switch opt.Tp {
 		case ast.TableOptionCharset:
-			charsets = append(charsets, opt.StrValue)
+			info, err := charset.GetCharsetDesc(opt.StrValue)
+			if err != nil {
+				return "", "", err
+			}
+			if len(chs) == 0 {
+				chs = info.Name
+			} else if chs != info.Name {
+				return "", "", ErrConflictingDeclarations.GenWithStackByArgs(chs, info.Name)
+			}
+			if len(coll) == 0 {
+				coll = info.DefaultCollation
+			}
 		case ast.TableOptionCollate:
-			collates = append(collates, opt.StrValue)
-		}
-	}
-
-	if len(charsets) > 1 {
-		return "", "", ErrConflictingDeclarations.GenWithStackByArgs(charsets[0], charsets[1])
-	}
-	if len(charsets) == 1 {
-		if charsets[0] == "" {
-			return "", "", ErrUnknownCharacterSet.GenWithStackByArgs("")
-		}
-		ca = charsets[0]
-	}
-	if len(collates) != 0 {
-		for i := range collates {
-			if collates[i] == "" {
-				return "", "", ErrUnknownCollation.GenWithStackByArgs("")
+			info, err := charset.GetCollationByName(opt.StrValue)
+			if err != nil {
+				return "", "", err
 			}
-			if len(ca) != 0 && !charset.ValidCharsetAndCollation(ca, collates[i]) {
-				return "", "", ErrCollationCharsetMismatch.GenWithStackByArgs(collates[i], ca)
+			if len(chs) == 0 {
+				chs = info.CharsetName
+			} else if chs != info.CharsetName {
+				return "", "", ErrCollationCharsetMismatch.GenWithStackByArgs(info.Name, chs)
 			}
+			coll = info.Name
 		}
-		co = collates[len(collates)-1]
 	}
 	return
 }

go.sum

@@ -197,7 +197,6 @@ github.com/pingcap/tipb v0.0.0-20191120020146-6161b015e21e/go.mod h1:RtkHW8WbcNx
 github.com/pkg/errors v0.8.0/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkg/errors v0.8.1 h1:iURUrRGxPUNPdy5/HRSm+Yj6okJ6UtLINN0Q9M4+h3I=
 github.com/pkg/errors v0.8.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
-github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/prometheus/client_golang v0.8.0/go.mod h1:7SWBe2y4D6OKWSNQJUaRYU/AaXPKyh/dDVn+NZz0KFw=
 github.com/prometheus/client_golang v0.9.1/go.mod h1:7SWBe2y4D6OKWSNQJUaRYU/AaXPKyh/dDVn+NZz0KFw=