stats: fix combined index low-bound check (#7814)

pingcap · Oct 8, 2018 · d5d8ba0 · d5d8ba0
1 parent 05666f5
commit d5d8ba0
Show file tree

Hide file tree

Showing 2 changed files with 32 additions and 2 deletions.
diff --git a/statistics/histogram.go b/statistics/histogram.go
@@ -682,9 +682,8 @@ func (hg *Histogram) outOfRange(val types.Datum) bool {
 	if hg.Bounds == nil {
 		return true
 	}
-	len := hg.Bounds.NumRows()
 	return chunk.Compare(hg.Bounds.GetRow(0), 0, &val) > 0 ||
-		chunk.Compare(hg.Bounds.GetRow(len-1), 0, &val) < 0
+		chunk.Compare(hg.Bounds.GetRow(hg.Bounds.NumRows()-1), 0, &val) < 0
 }
 
 // ErrorRate is the error rate of estimate row count by bucket and cm sketch.
@@ -857,3 +856,22 @@ func (idx *Index) getRowCount(sc *stmtctx.StatementContext, indexRanges []*range
 	}
 	return totalCount, nil
 }
+
+func (idx *Index) outOfRange(val types.Datum) bool {
+	if idx.Bounds == nil {
+		return true
+	}
+	withInLowBoundOrPrefixMatch := chunk.Compare(idx.Bounds.GetRow(0), 0, &val) <= 0 ||
+		matchPrefix(idx.Bounds.GetRow(0), 0, &val)
+	withInHighBound := chunk.Compare(idx.Bounds.GetRow(idx.Bounds.NumRows()-1), 0, &val) >= 0
+	return !withInLowBoundOrPrefixMatch || !withInHighBound
+}
+
+// matchPrefix checks whether ad is the prefix of value
+func matchPrefix(row chunk.Row, colIdx int, ad *types.Datum) bool {
+	switch ad.Kind() {
+	case types.KindString, types.KindBytes, types.KindBinaryLiteral, types.KindMysqlBit:
+		return strings.HasPrefix(row.GetString(colIdx), ad.GetString())
+	}
+	return false
+}
diff --git a/statistics/selectivity_test.go b/statistics/selectivity_test.go
@@ -229,6 +229,18 @@ func (s *testSelectivitySuite) TestDiscreteDistribution(c *C) {
 		"└─IndexScan_8 0.00 cop table:t, index:a, b, range:[\"tw\" -inf,\"tw\" 0), keep order:false"))
 }
 
+func (s *testSelectivitySuite) TestSelectCombinedLowBound(c *C) {
+	testKit := testkit.NewTestKit(c, s.store)
+	testKit.MustExec("use test")
+	testKit.MustExec("drop table if exists t")
+	testKit.MustExec("create table t(id int auto_increment, kid int, pid int, primary key(id), key(kid, pid))")
+	testKit.MustExec("insert into t (kid, pid) values (1,2), (1,3), (1,4),(1, 11), (1, 12), (1, 13), (1, 14), (2, 2), (2, 3), (2, 4)")
+	testKit.MustExec("analyze table t")
+	testKit.MustQuery("explain select * from t where kid = 1").Check(testkit.Rows(
+		"IndexReader_9 7.00 root index:IndexScan_8",
+		"└─IndexScan_8 7.00 cop table:t, index:kid, pid, range:[1,1], keep order:false"))
+}
+
 func getRange(start, end int64) []*ranger.Range {
 	ran := &ranger.Range{
 		LowVal:  []types.Datum{types.NewIntDatum(start)},