planner: do not use like to build range when new collation is enabled

expression/integration_serial_test.go

@@ -96,6 +96,26 @@ func TestIssue17891(t *testing.T) {
 	tk.MustExec("create table test(id int, value set ('a','b','c') charset utf8mb4 collate utf8mb4_general_ci default 'a,B ,C');")
 }
 
+func TestIssue31174(t *testing.T) {
+	collate.SetNewCollationEnabledForTest(true)
+	defer collate.SetNewCollationEnabledForTest(false)
+
+	store, clean := testkit.CreateMockStore(t)
+	defer clean()
+
+	tk := testkit.NewTestKit(t, store)
+	tk.MustExec("use test")
+	tk.MustExec("drop table if exists t")
+	tk.MustExec("create table t(a char(4) collate utf8_general_ci primary key /*T![clustered_index] clustered */);")
+	tk.MustExec("insert into t values('`?');")
+	// The 'like' condition can not be used to construct the range.
+	tk.MustQuery("explain format = 'brief' select * from t where a like '`%';").Check(testkit.Rows(""+
+		"TableReader 8000.00 root  data:Selection",
+		"└─Selection 8000.00 cop[tikv]  like(test.t.a, \"`%\", 92)",
+		"  └─TableFullScan 10000.00 cop[tikv] table:t keep order:false, stats:pseudo"))
+	tk.MustQuery("select * from t where a like '`%';").Check(testkit.Rows("`?"))
+}
+
 func TestIssue20268(t *testing.T) {
 	collate.SetNewCollationEnabledForTest(true)
 	defer collate.SetNewCollationEnabledForTest(false)

util/ranger/checker.go

@@ -112,6 +112,16 @@ func (c *conditionChecker) checkScalarFunction(scalar *expression.ScalarFunction
 
 func (c *conditionChecker) checkLikeFunc(scalar *expression.ScalarFunction) bool {
 	_, collation := scalar.CharsetAndCollation()
+	if collate.NewCollationEnabled() && !collate.IsBinCollation(collation) {
+		// The algorithm constructs the range in byte-level: for example, ab% is mapped to [ab, ac] by adding 1 to the last byte.
+		// However, this is incorrect for non-binary collation strings because the sort key order is not the same as byte order.
+		// For example, "`%" is mapped to the range [`, a](where ` is 0x60 and a is 0x61).
+		// Because the collation utf8_general_ci is case-insensitive, a and A have the same sort key.
+		// Finally, the range comes to be [`, A], which is actually an empty range.
+		// See https://github.com/pingcap/tidb/issues/31174 for more details.
+		// In short, when the column type is non-binary collation string, we cannot use `like` expressions to generate the range.
+		return false
+	}
 	if !collate.CompatibleCollate(scalar.GetArgs()[0].GetType().Collate, collation) {
 		return false
 	}