Fix show character set

This fixes the implementation for `show character set` by returning all supported collations and fixing like as well. We also move a bunch of things from `utf8` to the explicit `utf8mb3` names so that we're better compatible if MySQL ever decides to change the alias meaning. Signed-off-by: Dirkjan Bussink <[email protected]>
vitessio · Jul 21, 2023 · 2503965 · 2503965
1 parent 62c8334
commit 2503965
Show file tree

Hide file tree

Showing 14 changed files with 173 additions and 150 deletions.
diff --git a/go/mysql/flavor_mysql.go b/go/mysql/flavor_mysql.go
@@ -361,7 +361,7 @@ const TablesWithSize80 = `SELECT t.table_name,
 		i.allocated_size
 	FROM information_schema.tables t
 		LEFT JOIN information_schema.innodb_tablespaces i
-	ON i.name = CONCAT(t.table_schema, '/', t.table_name) COLLATE utf8_general_ci
+	ON i.name = CONCAT(t.table_schema, '/', t.table_name) COLLATE utf8mb3_general_ci
 	WHERE
 		t.table_schema = database() AND not t.create_options <=> 'partitioned'
 UNION ALL
@@ -374,7 +374,7 @@ UNION ALL
 		SUM(i.allocated_size)
 	FROM information_schema.tables t
 		LEFT JOIN information_schema.innodb_tablespaces i
-	ON i.name LIKE (CONCAT(t.table_schema, '/', t.table_name, '#p#%') COLLATE utf8_general_ci )
+	ON i.name LIKE (CONCAT(t.table_schema, '/', t.table_name, '#p#%') COLLATE utf8mb3_general_ci )
 	WHERE
 		t.table_schema = database() AND t.create_options <=> 'partitioned'
 	GROUP BY

diff --git a/go/vt/dbconfigs/dbconfigs.go b/go/vt/dbconfigs/dbconfigs.go
@@ -26,6 +26,8 @@ import (
 
 	"github.com/spf13/pflag"
 
+	"vitess.io/vitess/go/mysql/collations"
+
 	"vitess.io/vitess/go/vt/servenv"
 	"vitess.io/vitess/go/vt/vttls"
 
@@ -416,6 +418,6 @@ func NewTestDBConfigs(genParams, appDebugParams mysql.ConnParams, dbname string)
 		replParams:         genParams,
 		externalReplParams: genParams,
 		DBName:             dbname,
-		Charset:            "utf8mb4_general_ci",
+		Charset:            collations.Default().Get().Name(),
 	}
 }
diff --git a/go/vt/sqlparser/constants.go b/go/vt/sqlparser/constants.go
@@ -211,7 +211,7 @@ const (
 	Utf16Str    = "_utf16"
 	Utf16leStr  = "_utf16le"
 	Utf32Str    = "_utf32"
-	Utf8Str     = "_utf8"
+	Utf8mb3Str  = "_utf8mb3"
 	Utf8mb4Str  = "_utf8mb4"
 	NStringStr  = "N"
 

diff --git a/go/vt/sqlparser/parse_test.go b/go/vt/sqlparser/parse_test.go
@@ -2599,7 +2599,8 @@ var (
 	}, {
 		input: "select 1 from t where foo = _binary 'bar'",
 	}, {
-		input: "select 1 from t where foo = _utf8 'bar' and bar = _latin1 'sjösjuk'",
+		input:  "select 1 from t where foo = _utf8 'bar' and bar = _latin1 'sjösjuk'",
+		output: "select 1 from t where foo = _utf8mb3 'bar' and bar = _latin1 'sjösjuk'",
 	}, {
 		input:  "select 1 from t where foo = _binary'bar'",
 		output: "select 1 from t where foo = _binary 'bar'",
@@ -2610,10 +2611,10 @@ var (
 		output: "select 1 from t where foo = _utf8mb4 'bar'",
 	}, {
 		input:  "select 1 from t where foo = _utf8mb3 'bar'",
-		output: "select 1 from t where foo = _utf8 'bar'",
+		output: "select 1 from t where foo = _utf8mb3 'bar'",
 	}, {
-		input:  "select 1 from t where foo = _utf8mb3'bar'",
-		output: "select 1 from t where foo = _utf8 'bar'",
+		input:  "select 1 from t where foo = _utf8'bar'",
+		output: "select 1 from t where foo = _utf8mb3 'bar'",
 	}, {
 		input: "select match(a) against ('foo') from t",
 	}, {
@@ -4036,13 +4037,13 @@ func TestIntroducers(t *testing.T) {
 		output: "select _utf32 'x' from dual",
 	}, {
 		input:  "select _utf8 'x'",
-		output: "select _utf8 'x' from dual",
+		output: "select _utf8mb3 'x' from dual",
 	}, {
 		input:  "select _utf8mb4 'x'",
 		output: "select _utf8mb4 'x' from dual",
 	}, {
 		input:  "select _utf8mb3 'x'",
-		output: "select _utf8 'x' from dual",
+		output: "select _utf8mb3 'x' from dual",
 	}}
 	for _, tcase := range validSQL {
 		t.Run(tcase.input, func(t *testing.T) {

diff --git a/go/vt/sqlparser/sql.go b/go/vt/sqlparser/sql.go
diff --git a/go/vt/sqlparser/sql.y b/go/vt/sqlparser/sql.y
@@ -1895,15 +1895,15 @@ underscore_charsets:
   }
 | UNDERSCORE_UTF8
   {
-    $$ = Utf8Str
+    $$ = Utf8mb3Str
   }
 | UNDERSCORE_UTF8MB4
   {
     $$ = Utf8mb4Str
   }
 | UNDERSCORE_UTF8MB3
   {
-    $$ = Utf8Str
+    $$ = Utf8mb3Str
   }
 
 literal_or_null: