parser: revert latin1 as an alias for utf8mb4

expression/integration_test.go

@@ -653,7 +653,7 @@ func TestStringBuiltin(t *testing.T) {
 	result = tk.MustQuery("select ord('123'), ord(123), ord(''), ord('你好'), ord(NULL), ord('👍')")
 	result.Check(testkit.Rows("49 49 0 14990752 <nil> 4036989325"))
 	result = tk.MustQuery("select ord(X''), ord(X'6161'), ord(X'e4bd'), ord(X'e4bda0'), ord(_ascii'你'), ord(_latin1'你')")
-	result.Check(testkit.Rows("0 97 228 228 228 14990752"))
+	result.Check(testkit.Rows("0 97 228 228 228 228"))
 
 	// for space
 	result = tk.MustQuery(`select space(0), space(2), space(-1), space(1.1), space(1.9)`)

parser/charset/encoding_latin1.go

@@ -14,11 +14,12 @@
 package charset
 
 import (
+	"bytes"
 	"golang.org/x/text/encoding"
 )
 
 // EncodingLatin1Impl is the instance of encodingLatin1.
-// In TiDB, latin1 is an alias for utf8, so uses utf8 implementation for latin1.
+// TiDB uses utf8 implementation for latin1 charset because of the backward compatibility.
 var EncodingLatin1Impl = &encodingLatin1{encodingUTF8{encodingBase{enc: encoding.Nop}}}
 
 func init() {
@@ -34,3 +35,25 @@ type encodingLatin1 struct {
 func (e *encodingLatin1) Name() string {
 	return CharsetLatin1
 }
+
+// Peek implements Encoding interface.
+func (e *encodingLatin1) Peek(src []byte) []byte {
+	if len(src) == 0 {
+		return src
+	}
+	return src[:1]
+}
+
+// IsValid implements Encoding interface.
+func (e *encodingLatin1) IsValid(src []byte) bool {
+	return true
+}
+
+// Tp implements Encoding interface.
+func (e *encodingLatin1) Tp() EncodingTp {
+	return EncodingTpLatin1
+}
+
+func (e *encodingLatin1) Transform(dest *bytes.Buffer, src []byte, op Op) ([]byte, error) {
+	return src, nil
+}

parser/mysql/charset.go

@@ -593,9 +593,9 @@ const (
 	MaxBytesOfCharacter = 4
 )
 
-// IsUTF8Charset checks if charset is utf8, utf8mb4 or latin1.
+// IsUTF8Charset checks if charset is utf8, utf8mb4.
 func IsUTF8Charset(charset string) bool {
-	return charset == UTF8Charset || charset == UTF8MB4Charset || charset == Latin1Charset
+	return charset == UTF8Charset || charset == UTF8MB4Charset
 }
 
 // RangeGraph defines valid unicode characters to use in column names. It strictly follows MySQL's definition.