let latin1 use UTF8MB3 instead of UTF8MB4

Defined2014 · May 17, 2022 · f7ae84b · f7ae84b
1 parent e255739
commit f7ae84b
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 3 deletions.
diff --git a/parser/charset/encoding_latin1.go b/parser/charset/encoding_latin1.go
@@ -18,16 +18,22 @@ import (
 )
 
 // EncodingLatin1Impl is the instance of encodingLatin1.
-// In TiDB, latin1 is an alias for utf8, so uses utf8 implementation for latin1.
-var EncodingLatin1Impl = &encodingLatin1{encodingUTF8{encodingBase{enc: encoding.Nop}}}
+// In TiDB, latin1 is an alias for utf8, so uses utf8mb3 implementation for latin1.
+var EncodingLatin1Impl = &encodingUTF8MB3Strict{
+	encodingUTF8{
+		encodingBase{
+			enc: encoding.Nop,
+		},
+	},
+}
 
 func init() {
 	EncodingLatin1Impl.self = EncodingLatin1Impl
 }
 
 // encodingLatin1 compatibles with latin1 in old version TiDB.
 type encodingLatin1 struct {
-	encodingUTF8
+	encodingUTF8MB3Strict
 }
 
 // Name implements Encoding interface.

diff --git a/parser/charset/encoding_test.go b/parser/charset/encoding_test.go
@@ -130,6 +130,15 @@ func TestEncodingValidate(t *testing.T) {
 		{charset.CharsetUTF8, oxfffefd, "???", 0, false},
 		{charset.CharsetUTF8, "中文" + oxfffefd, "中文???", 6, false},
 		{charset.CharsetUTF8, string(utf8.RuneError), "�", 3, true},
+		{charset.CharsetLatin1, "", "", 0, true},
+		{charset.CharsetLatin1, "qwerty", "qwerty", 6, true},
+		{charset.CharsetLatin1, "qwÊrty", "qwÊrty", 7, true},
+		{charset.CharsetLatin1, "qwÊ合法字符串", "qwÊ合法字符串", 19, true},
+		{charset.CharsetLatin1, "😂", "?", 0, false},
+		{charset.CharsetLatin1, "valid_str😂", "valid_str?", 9, false},
+		{charset.CharsetLatin1, oxfffefd, "???", 0, false},
+		{charset.CharsetLatin1, "中文" + oxfffefd, "中文???", 6, false},
+		{charset.CharsetLatin1, string(utf8.RuneError), "�", 3, true},
 		{charset.CharsetGBK, "", "", 0, true},
 		{charset.CharsetGBK, "asdf", "asdf", 4, true},
 		{charset.CharsetGBK, "中文", "中文", 6, true},