executor.go

// Copyright 2015 PingCAP, Inc.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// See the License for the specific language governing permissions and
// limitations under the License.

package executor

import (
	"container/heap"
	"sort"
	"sync"
	"sync/atomic"

	"github.com/juju/errors"
	"github.com/pingcap/tidb/ast"
	"github.com/pingcap/tidb/context"
	"github.com/pingcap/tidb/expression"
	"github.com/pingcap/tidb/infoschema"
	"github.com/pingcap/tidb/inspectkv"
	"github.com/pingcap/tidb/kv"
	"github.com/pingcap/tidb/model"
	"github.com/pingcap/tidb/mysql"
	"github.com/pingcap/tidb/plan"
	"github.com/pingcap/tidb/sessionctx"
	"github.com/pingcap/tidb/sessionctx/forupdate"
	"github.com/pingcap/tidb/sessionctx/variable"
	"github.com/pingcap/tidb/table"
	"github.com/pingcap/tidb/tablecodec"
	"github.com/pingcap/tidb/terror"
	"github.com/pingcap/tidb/util/codec"
	"github.com/pingcap/tidb/util/distinct"
	"github.com/pingcap/tidb/util/types"
)

var (
	_ Executor = &ApplyExec{}
	_ Executor = &CheckTableExec{}
	_ Executor = &DistinctExec{}
	_ Executor = &DummyScanExec{}
	_ Executor = &ExistsExec{}
	_ Executor = &HashAggExec{}
	_ Executor = &HashJoinExec{}
	_ Executor = &HashSemiJoinExec{}
	_ Executor = &LimitExec{}
	_ Executor = &MaxOneRowExec{}
	_ Executor = &ProjectionExec{}
	_ Executor = &ReverseExec{}
	_ Executor = &SelectionExec{}
	_ Executor = &SelectLockExec{}
	_ Executor = &ShowDDLExec{}
	_ Executor = &SortExec{}
	_ Executor = &StreamAggExec{}
	_ Executor = &TableDualExec{}
	_ Executor = &TableScanExec{}
	_ Executor = &TopnExec{}
	_ Executor = &TrimExec{}
	_ Executor = &UnionExec{}
)

// Error instances.
var (
	ErrUnknownPlan     = terror.ClassExecutor.New(codeUnknownPlan, "Unknown plan")
	ErrPrepareMulti    = terror.ClassExecutor.New(codePrepareMulti, "Can not prepare multiple statements")
	ErrStmtNotFound    = terror.ClassExecutor.New(codeStmtNotFound, "Prepared statement not found")
	ErrSchemaChanged   = terror.ClassExecutor.New(codeSchemaChanged, "Schema has changed")
	ErrWrongParamCount = terror.ClassExecutor.New(codeWrongParamCount, "Wrong parameter count")
	ErrRowKeyCount     = terror.ClassExecutor.New(codeRowKeyCount, "Wrong row key entry count")
	ErrPrepareDDL      = terror.ClassExecutor.New(codePrepareDDL, "Can not prepare DDL statements")
	ErrPasswordNoMatch = terror.ClassExecutor.New(CodePasswordNoMatch, "Can't find any matching row in the user table")
)

// Error codes.
const (
	codeUnknownPlan     terror.ErrCode = 1
	codePrepareMulti    terror.ErrCode = 2
	codeStmtNotFound    terror.ErrCode = 3
	codeSchemaChanged   terror.ErrCode = 4
	codeWrongParamCount terror.ErrCode = 5
	codeRowKeyCount     terror.ErrCode = 6
	codePrepareDDL      terror.ErrCode = 7
	// MySQL error code
	CodePasswordNoMatch terror.ErrCode = 1133
	CodeCannotUser      terror.ErrCode = 1396
)

// Row represents a result set row, it may be returned from a table, a join, or a projection.
type Row struct {
	// Data is the output record data for current Plan.
	Data []types.Datum
	// RowKeys contains all table row keys in the row.
	RowKeys []*RowKeyEntry
}

// RowKeyEntry represents a row key read from a table.
type RowKeyEntry struct {
	// The table which this row come from.
	Tbl table.Table
	// Row key.
	Handle int64
	// Table alias name.
	TableAsName *model.CIStr
}

// Executor executes a query.
type Executor interface {
	Next() (*Row, error)
	Close() error
	Schema() expression.Schema
}

// ShowDDLExec represents a show DDL executor.
type ShowDDLExec struct {
	schema expression.Schema
	ctx    context.Context
	done   bool
}

// Schema implements the Executor Schema interface.
func (e *ShowDDLExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *ShowDDLExec) Next() (*Row, error) {
	if e.done {
		return nil, nil
	}

	txn, err := e.ctx.GetTxn(false)
	if err != nil {
		return nil, errors.Trace(err)
	}

	ddlInfo, err := inspectkv.GetDDLInfo(txn)
	if err != nil {
		return nil, errors.Trace(err)
	}
	bgInfo, err := inspectkv.GetBgDDLInfo(txn)
	if err != nil {
		return nil, errors.Trace(err)
	}

	var ddlOwner, ddlJob string
	if ddlInfo.Owner != nil {
		ddlOwner = ddlInfo.Owner.String()
	}
	if ddlInfo.Job != nil {
		ddlJob = ddlInfo.Job.String()
	}

	var bgOwner, bgJob string
	if bgInfo.Owner != nil {
		bgOwner = bgInfo.Owner.String()
	}
	if bgInfo.Job != nil {
		bgJob = bgInfo.Job.String()
	}

	row := &Row{}
	row.Data = types.MakeDatums(
		ddlInfo.SchemaVer,
		ddlOwner,
		ddlJob,
		bgInfo.SchemaVer,
		bgOwner,
		bgJob,
	)
	e.done = true

	return row, nil
}

// Close implements the Executor Close interface.
func (e *ShowDDLExec) Close() error {
	return nil
}

// CheckTableExec represents a check table executor.
// It is built from the "admin check table" statement, and it checks if the
// index matches the records in the table.
type CheckTableExec struct {
	tables []*ast.TableName
	ctx    context.Context
	done   bool
}

// Schema implements the Executor Schema interface.
func (e *CheckTableExec) Schema() expression.Schema {
	return nil
}

// Next implements the Executor Next interface.
func (e *CheckTableExec) Next() (*Row, error) {
	if e.done {
		return nil, nil
	}

	dbName := model.NewCIStr(e.ctx.GetSessionVars().CurrentDB)
	is := sessionctx.GetDomain(e.ctx).InfoSchema()

	for _, t := range e.tables {
		tb, err := is.TableByName(dbName, t.Name)
		if err != nil {
			return nil, errors.Trace(err)
		}
		for _, idx := range tb.Indices() {
			txn, err := e.ctx.GetTxn(false)
			if err != nil {
				return nil, errors.Trace(err)
			}
			err = inspectkv.CompareIndexData(txn, tb, idx)
			if err != nil {
				return nil, errors.Errorf("%v err:%v", t.Name, err)
			}
		}
	}
	e.done = true

	return nil, nil
}

// Close implements plan.Plan Close interface.
func (e *CheckTableExec) Close() error {
	return nil
}

// SelectLockExec represents a select lock executor.
// It is built from the "SELECT .. FOR UPDATE" or the "SELECT .. LOCK IN SHARE MODE" statement.
// For "SELECT .. FOR UPDATE" statement, it locks every row key from source Executor.
// After the execution, the keys are buffered in transaction, and will be sent to KV
// when doing commit. If there is any key already locked by another transaction,
// the transaction will rollback and retry.
type SelectLockExec struct {
	Src    Executor
	Lock   ast.SelectLockType
	ctx    context.Context
	schema expression.Schema
}

// Schema implements the Executor Schema interface.
func (e *SelectLockExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *SelectLockExec) Next() (*Row, error) {
	row, err := e.Src.Next()
	if err != nil {
		return nil, errors.Trace(err)
	}
	if row == nil {
		return nil, nil
	}
	if len(row.RowKeys) != 0 && e.Lock == ast.SelectLockForUpdate {
		forupdate.SetForUpdate(e.ctx)
		txn, err := e.ctx.GetTxn(false)
		if err != nil {
			return nil, errors.Trace(err)
		}
		for _, k := range row.RowKeys {
			lockKey := tablecodec.EncodeRowKeyWithHandle(k.Tbl.Meta().ID, k.Handle)
			err = txn.LockKeys(lockKey)
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
	}
	return row, nil
}

// Close implements the Executor Close interface.
func (e *SelectLockExec) Close() error {
	return e.Src.Close()
}

// LimitExec represents limit executor
// It ignores 'Offset' rows from src, then returns 'Count' rows at maximum.
type LimitExec struct {
	Src    Executor
	Offset uint64
	Count  uint64
	Idx    uint64
	schema expression.Schema
}

// Schema implements the Executor Schema interface.
func (e *LimitExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *LimitExec) Next() (*Row, error) {
	for e.Idx < e.Offset {
		srcRow, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if srcRow == nil {
			return nil, nil
		}
		e.Idx++
	}
	if e.Idx >= e.Count+e.Offset {
		return nil, nil
	}
	srcRow, err := e.Src.Next()
	if err != nil {
		return nil, errors.Trace(err)
	}
	if srcRow == nil {
		return nil, nil
	}
	e.Idx++
	return srcRow, nil
}

// Close implements the Executor Close interface.
func (e *LimitExec) Close() error {
	e.Idx = 0
	return e.Src.Close()
}

// orderByRow binds a row to its order values, so it can be sorted.
type orderByRow struct {
	key []types.Datum
	row *Row
}

// DistinctExec represents Distinct executor.
// It ignores duplicate rows from source Executor by using a *distinct.Checker which maintains
// a map to check duplication.
// Because every distinct row will be added to the map, the memory usage might be very high.
type DistinctExec struct {
	Src     Executor
	checker *distinct.Checker
	schema  expression.Schema
}

// Schema implements the Executor Schema interface.
func (e *DistinctExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *DistinctExec) Next() (*Row, error) {
	if e.checker == nil {
		e.checker = distinct.CreateDistinctChecker()
	}
	for {
		row, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if row == nil {
			return nil, nil
		}
		ok, err := e.checker.Check(types.DatumsToInterfaces(row.Data))
		if err != nil {
			return nil, errors.Trace(err)
		}
		if !ok {
			continue
		}
		return row, nil
	}
}

// Close implements the Executor Close interface.
func (e *DistinctExec) Close() error {
	return e.Src.Close()
}

// ReverseExec produces reverse ordered result, it is used to wrap executors that do not support reverse scan.
type ReverseExec struct {
	Src    Executor
	rows   []*Row
	cursor int
	done   bool
}

// Schema implements the Executor Schema interface.
func (e *ReverseExec) Schema() expression.Schema {
	return e.Src.Schema()
}

// Next implements the Executor Next interface.
func (e *ReverseExec) Next() (*Row, error) {
	if !e.done {
		for {
			row, err := e.Src.Next()
			if err != nil {
				return nil, errors.Trace(err)
			}
			if row == nil {
				break
			}
			e.rows = append(e.rows, row)
		}
		e.cursor = len(e.rows) - 1
		e.done = true
	}
	if e.cursor < 0 {
		return nil, nil
	}
	row := e.rows[e.cursor]
	e.cursor--
	return row, nil
}

// Close implements the Executor Close interface.
func (e *ReverseExec) Close() error {
	return e.Src.Close()
}

func init() {
	// While doing optimization in the plan package, we need to execute uncorrelated subquery,
	// but the plan package cannot import the executor package because of the dependency cycle.
	// So we assign a function implemented in the executor package to the plan package to avoid the dependency cycle.
	plan.EvalSubquery = func(p plan.PhysicalPlan, is infoschema.InfoSchema, ctx context.Context) (d []types.Datum, err error) {
		e := &executorBuilder{is: is, ctx: ctx}
		exec := e.build(p)
		row, err := exec.Next()
		if err != nil {
			return d, errors.Trace(err)
		}
		if row == nil {
			return
		}
		return row.Data, nil
	}
	tableMySQLErrCodes := map[terror.ErrCode]uint16{
		CodeCannotUser:      mysql.ErrCannotUser,
		CodePasswordNoMatch: mysql.ErrPasswordNoMatch,
	}
	terror.ErrClassToMySQLCodes[terror.ClassExecutor] = tableMySQLErrCodes
}

// HashJoinExec implements the hash join algorithm.
type HashJoinExec struct {
	hashTable     map[string][]*Row
	smallHashKey  []*expression.Column
	bigHashKey    []*expression.Column
	smallExec     Executor
	bigExec       Executor
	prepared      bool
	ctx           context.Context
	smallFilter   expression.Expression
	bigFilter     expression.Expression
	otherFilter   expression.Expression
	schema        expression.Schema
	outer         bool
	leftSmall     bool
	cursor        int
	defaultValues []types.Datum
	// targetTypes means the target the type that both smallHashKey and bigHashKey should convert to.
	targetTypes []*types.FieldType

	finished atomic.Value
	// For sync multiple join workers.
	wg sync.WaitGroup
	// closeCh add a lock for closing executor.
	closeCh chan struct{}

	// Concurrent channels.
	concurrency      int
	bigTableRows     []chan []*Row
	bigTableErr      chan error
	hashJoinContexts []*hashJoinCtx

	// Channels for output.
	resultErr  chan error
	resultRows chan *Row
}

// hashJoinCtx holds the variables needed to do a hash join in one of many concurrent goroutines.
type hashJoinCtx struct {
	bigFilter   expression.Expression
	otherFilter expression.Expression
	// Buffer used for encode hash keys.
	datumBuffer   []types.Datum
	hashKeyBuffer []byte
}

// Close implements the Executor Close interface.
func (e *HashJoinExec) Close() error {
	e.finished.Store(true)
	<-e.closeCh
	e.prepared = false
	e.cursor = 0
	return e.smallExec.Close()
}

// makeJoinRow simply creates a new row that appends row b to row a.
func makeJoinRow(a *Row, b *Row) *Row {
	ret := &Row{
		RowKeys: make([]*RowKeyEntry, 0, len(a.RowKeys)+len(b.RowKeys)),
		Data:    make([]types.Datum, 0, len(a.Data)+len(b.Data)),
	}
	ret.RowKeys = append(ret.RowKeys, a.RowKeys...)
	ret.RowKeys = append(ret.RowKeys, b.RowKeys...)
	ret.Data = append(ret.Data, a.Data...)
	ret.Data = append(ret.Data, b.Data...)
	return ret
}

// getHashKey gets the hash key when given a row and hash columns.
// It will return a boolean value representing if the hash key has null, a byte slice representing the result hash code.
func getHashKey(sc *variable.StatementContext, cols []*expression.Column, row *Row, targetTypes []*types.FieldType,
	vals []types.Datum, bytes []byte) (bool, []byte, error) {
	var err error
	for i, col := range cols {
		vals[i], err = col.Eval(row.Data, nil)
		if err != nil {
			return false, nil, errors.Trace(err)
		}
		if vals[i].IsNull() {
			return true, nil, nil
		}
		if targetTypes[i].Tp != col.RetType.Tp {
			vals[i], err = vals[i].ConvertTo(sc, targetTypes[i])
			if err != nil {
				return false, nil, errors.Trace(err)
			}
		}
	}
	if len(vals) == 0 {
		return false, nil, nil
	}
	bytes, err = codec.EncodeValue(bytes, vals...)
	return false, bytes, errors.Trace(err)
}

// Schema implements the Executor Schema interface.
func (e *HashJoinExec) Schema() expression.Schema {
	return e.schema
}

var batchSize = 128

// fetchBigExec fetches rows from the big table in a background goroutine
// and sends the rows to multiple channels which will be read by multiple join workers.
func (e *HashJoinExec) fetchBigExec() {
	cnt := 0
	defer func() {
		for _, cn := range e.bigTableRows {
			close(cn)
		}
		e.bigExec.Close()
		e.wg.Done()
	}()
	curBatchSize := 1
	for {
		rows := make([]*Row, 0, batchSize)
		done := false
		for i := 0; i < curBatchSize; i++ {
			if e.finished.Load().(bool) {
				return
			}
			row, err := e.bigExec.Next()
			if err != nil {
				e.bigTableErr <- errors.Trace(err)
				done = true
				break
			}
			if row == nil {
				done = true
				break
			}
			rows = append(rows, row)
		}
		idx := cnt % e.concurrency
		e.bigTableRows[idx] <- rows
		cnt++
		if done {
			break
		}
		if curBatchSize < batchSize {
			curBatchSize *= 2
		}
	}
}

// prepare runs the first time when 'Next' is called, it starts one worker goroutine to fetch rows from the big table,
// and reads all data from the small table to build a hash table, then starts multiple join worker goroutines.
func (e *HashJoinExec) prepare() error {
	e.closeCh = make(chan struct{}, 1)
	e.finished.Store(false)
	e.bigTableRows = make([]chan []*Row, e.concurrency)
	e.wg = sync.WaitGroup{}
	for i := 0; i < e.concurrency; i++ {
		e.bigTableRows[i] = make(chan []*Row, e.concurrency*batchSize)
	}
	e.bigTableErr = make(chan error, 1)

	// Start a worker to fetch big table rows.
	e.wg.Add(1)
	go e.fetchBigExec()

	e.hashTable = make(map[string][]*Row)
	e.cursor = 0
	sc := e.ctx.GetSessionVars().StmtCtx
	for {
		row, err := e.smallExec.Next()
		if err != nil {
			return errors.Trace(err)
		}
		if row == nil {
			e.smallExec.Close()
			break
		}

		matched := true
		if e.smallFilter != nil {
			matched, err = expression.EvalBool(e.smallFilter, row.Data, e.ctx)
			if err != nil {
				return errors.Trace(err)
			}
			if !matched {
				continue
			}
		}
		hasNull, hashcode, err := getHashKey(sc, e.smallHashKey, row, e.targetTypes, e.hashJoinContexts[0].datumBuffer, nil)
		if err != nil {
			return errors.Trace(err)
		}
		if hasNull {
			continue
		}
		if rows, ok := e.hashTable[string(hashcode)]; !ok {
			e.hashTable[string(hashcode)] = []*Row{row}
		} else {
			e.hashTable[string(hashcode)] = append(rows, row)
		}
	}

	e.resultRows = make(chan *Row, e.concurrency*1000)
	e.resultErr = make(chan error, 1)

	for i := 0; i < e.concurrency; i++ {
		e.wg.Add(1)
		go e.runJoinWorker(i)
	}
	go e.waitJoinWorkersAndCloseResultChan()

	e.prepared = true
	return nil
}

func (e *HashJoinExec) waitJoinWorkersAndCloseResultChan() {
	e.wg.Wait()
	close(e.resultRows)
	e.hashTable = nil
	close(e.closeCh)
}

// doJoin does join job in one goroutine.
func (e *HashJoinExec) runJoinWorker(idx int) {
	for {
		var (
			bigRows []*Row
			ok      bool
			err     error
		)
		select {
		case bigRows, ok = <-e.bigTableRows[idx]:
		case err = <-e.bigTableErr:
		}
		if err != nil {
			e.resultErr <- errors.Trace(err)
			break
		}
		if !ok || e.finished.Load().(bool) {
			break
		}
		for _, bigRow := range bigRows {
			succ := e.joinOneBigRow(e.hashJoinContexts[idx], bigRow)
			if !succ {
				break
			}
		}
	}
	e.wg.Done()
}

// joinOneBigRow creates result rows from a row in a big table and sends them to resultRows channel.
// Every matching row generates a result row.
// If there are no matching rows and it is outer join, a null filled result row is created.
func (e *HashJoinExec) joinOneBigRow(ctx *hashJoinCtx, bigRow *Row) bool {
	var (
		matchedRows []*Row
		err         error
	)
	bigMatched := true
	if e.bigFilter != nil {
		bigMatched, err = expression.EvalBool(ctx.bigFilter, bigRow.Data, e.ctx)
		if err != nil {
			e.resultErr <- errors.Trace(err)
			return false
		}
	}
	if bigMatched {
		matchedRows, err = e.constructMatchedRows(ctx, bigRow)
		if err != nil {
			e.resultErr <- errors.Trace(err)
			return false
		}
	}
	for _, r := range matchedRows {
		e.resultRows <- r
	}
	if len(matchedRows) == 0 && e.outer {
		r := e.fillRowWithDefaultValues(bigRow)
		e.resultRows <- r
	}
	return true
}

// constructMatchedRows creates matching result rows from a row in the big table.
func (e *HashJoinExec) constructMatchedRows(ctx *hashJoinCtx, bigRow *Row) (matchedRows []*Row, err error) {
	sc := e.ctx.GetSessionVars().StmtCtx
	hasNull, hashcode, err := getHashKey(sc, e.bigHashKey, bigRow, e.targetTypes, ctx.datumBuffer, ctx.hashKeyBuffer[0:0:cap(ctx.hashKeyBuffer)])
	if err != nil {
		return nil, errors.Trace(err)
	}

	if hasNull {
		return
	}
	rows, ok := e.hashTable[string(hashcode)]
	if !ok {
		return
	}
	// match eq condition
	for _, smallRow := range rows {
		otherMatched := true
		var matchedRow *Row
		if e.leftSmall {
			matchedRow = makeJoinRow(smallRow, bigRow)
		} else {
			matchedRow = makeJoinRow(bigRow, smallRow)
		}
		if e.otherFilter != nil {
			otherMatched, err = expression.EvalBool(ctx.otherFilter, matchedRow.Data, e.ctx)
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
		if otherMatched {
			matchedRows = append(matchedRows, matchedRow)
		}
	}

	return matchedRows, nil
}

// fillRowWithDefaultValues creates a result row filled with default values from a row in the big table.
// It is used for outer join, when a row from outer table doesn't have any matching rows.
func (e *HashJoinExec) fillRowWithDefaultValues(bigRow *Row) (returnRow *Row) {
	smallRow := &Row{
		Data: make([]types.Datum, len(e.smallExec.Schema())),
	}
	copy(smallRow.Data, e.defaultValues)
	if e.leftSmall {
		returnRow = makeJoinRow(smallRow, bigRow)
	} else {
		returnRow = makeJoinRow(bigRow, smallRow)
	}
	return returnRow
}

// Next implements the Executor Next interface.
func (e *HashJoinExec) Next() (*Row, error) {
	if !e.prepared {
		if err := e.prepare(); err != nil {
			return nil, errors.Trace(err)
		}
	}
	var (
		row *Row
		err error
		ok  bool
	)
	select {
	case row, ok = <-e.resultRows:
	case err, ok = <-e.resultErr:
	}
	if err != nil {
		e.finished.Store(true)
		return nil, errors.Trace(err)
	}
	if !ok {
		return nil, nil
	}
	return row, nil
}

// HashSemiJoinExec implements the hash join algorithm for semi join.
type HashSemiJoinExec struct {
	hashTable    map[string][]*Row
	smallHashKey []*expression.Column
	bigHashKey   []*expression.Column
	smallExec    Executor
	bigExec      Executor
	prepared     bool
	ctx          context.Context
	smallFilter  expression.Expression
	bigFilter    expression.Expression
	otherFilter  expression.Expression
	schema       expression.Schema
	// In auxMode, the result row always returns with an extra column which stores a boolean
	// or NULL value to indicate if this row is matched.
	auxMode           bool
	targetTypes       []*types.FieldType
	smallTableHasNull bool
	// If anti is true, semi join only output the unmatched row.
	anti bool
}

// Close implements the Executor Close interface.
func (e *HashSemiJoinExec) Close() error {
	e.prepared = false
	e.hashTable = make(map[string][]*Row)
	e.smallTableHasNull = false
	err := e.smallExec.Close()
	if err != nil {
		return errors.Trace(err)
	}
	return e.bigExec.Close()
}

// Schema implements the Executor Schema interface.
func (e *HashSemiJoinExec) Schema() expression.Schema {
	return e.schema
}

// Prepare runs the first time when 'Next' is called and it reads all data from the small table and stores
// them in a hash table.
func (e *HashSemiJoinExec) prepare() error {
	e.hashTable = make(map[string][]*Row)
	sc := e.ctx.GetSessionVars().StmtCtx
	for {
		row, err := e.smallExec.Next()
		if err != nil {
			return errors.Trace(err)
		}
		if row == nil {
			e.smallExec.Close()
			break
		}

		matched := true
		if e.smallFilter != nil {
			matched, err = expression.EvalBool(e.smallFilter, row.Data, e.ctx)
			if err != nil {
				return errors.Trace(err)
			}
			if !matched {
				continue
			}
		}
		hasNull, hashcode, err := getHashKey(sc, e.smallHashKey, row, e.targetTypes, make([]types.Datum, len(e.smallHashKey)), nil)
		if err != nil {
			return errors.Trace(err)
		}
		if hasNull {
			e.smallTableHasNull = true
			continue
		}
		if rows, ok := e.hashTable[string(hashcode)]; !ok {
			e.hashTable[string(hashcode)] = []*Row{row}
		} else {
			e.hashTable[string(hashcode)] = append(rows, row)
		}
	}

	e.prepared = true
	return nil
}

func (e *HashSemiJoinExec) rowIsMatched(bigRow *Row) (matched bool, hasNull bool, err error) {
	sc := e.ctx.GetSessionVars().StmtCtx
	hasNull, hashcode, err := getHashKey(sc, e.bigHashKey, bigRow, e.targetTypes, make([]types.Datum, len(e.smallHashKey)), nil)
	if err != nil {
		return false, false, errors.Trace(err)
	}
	if hasNull {
		return false, true, nil
	}
	rows, ok := e.hashTable[string(hashcode)]
	if !ok {
		return
	}
	// match eq condition
	for _, smallRow := range rows {
		matched = true
		if e.otherFilter != nil {
			var matchedRow *Row
			matchedRow = makeJoinRow(bigRow, smallRow)
			matched, err = expression.EvalBool(e.otherFilter, matchedRow.Data, e.ctx)
			if err != nil {
				return false, false, errors.Trace(err)
			}
		}
		if matched {
			return
		}
	}
	return
}

// Next implements the Executor Next interface.
func (e *HashSemiJoinExec) Next() (*Row, error) {
	if !e.prepared {
		if err := e.prepare(); err != nil {
			return nil, errors.Trace(err)
		}
	}

	for {
		bigRow, err := e.bigExec.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if bigRow == nil {
			e.bigExec.Close()
			return nil, nil
		}

		matched := true
		if e.bigFilter != nil {
			matched, err = expression.EvalBool(e.bigFilter, bigRow.Data, e.ctx)
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
		isNull := false
		if matched {
			matched, isNull, err = e.rowIsMatched(bigRow)
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
		if !matched && e.smallTableHasNull {
			isNull = true
		}
		if e.anti && !isNull {
			matched = !matched
		}
		// For the auxMode subquery, we return the row with a Datum indicating if it's a match,
		// For the non-auxMode subquery, we return the matching row only.
		if e.auxMode {
			if isNull {
				bigRow.Data = append(bigRow.Data, types.NewDatum(nil))
			} else {
				bigRow.Data = append(bigRow.Data, types.NewDatum(matched))
			}
			return bigRow, nil
		}
		if matched {
			return bigRow, nil
		}
	}
}

// HashAggExec deals with all the aggregate functions.
// It is built from the Aggregate Plan. When Next() is called, it reads all the data from Src
// and updates all the items in AggFuncs.
type HashAggExec struct {
	Src               Executor
	schema            expression.Schema
	executed          bool
	hasGby            bool
	aggType           plan.AggregationType
	ctx               context.Context
	AggFuncs          []expression.AggregationFunction
	groupMap          map[string]bool
	groups            [][]byte
	currentGroupIndex int
	GroupByItems      []expression.Expression
}

// Close implements the Executor Close interface.
func (e *HashAggExec) Close() error {
	e.executed = false
	e.groups = nil
	e.currentGroupIndex = 0
	for _, agg := range e.AggFuncs {
		agg.Clear()
	}
	return e.Src.Close()
}

// Schema implements the Executor Schema interface.
func (e *HashAggExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *HashAggExec) Next() (*Row, error) {
	// In this stage we consider all data from src as a single group.
	if !e.executed {
		e.groupMap = make(map[string]bool)
		for {
			hasMore, err := e.innerNext()
			if err != nil {
				return nil, errors.Trace(err)
			}
			if !hasMore {
				break
			}
		}
		e.executed = true
		if (len(e.groups) == 0) && !e.hasGby {
			// If no groupby and no data, we should add an empty group.
			// For example:
			// "select count(c) from t;" should return one row [0]
			// "select count(c) from t group by c1;" should return empty result set.
			e.groups = append(e.groups, []byte{})
		}
	}
	if e.currentGroupIndex >= len(e.groups) {
		return nil, nil
	}
	retRow := &Row{Data: make([]types.Datum, 0, len(e.AggFuncs))}
	groupKey := e.groups[e.currentGroupIndex]
	for _, af := range e.AggFuncs {
		retRow.Data = append(retRow.Data, af.GetGroupResult(groupKey))
	}
	e.currentGroupIndex++
	return retRow, nil
}

func (e *HashAggExec) getGroupKey(row *Row) ([]byte, error) {
	if e.aggType == plan.FinalAgg {
		val, err := e.GroupByItems[0].Eval(row.Data, e.ctx)
		if err != nil {
			return nil, errors.Trace(err)
		}
		return val.GetBytes(), nil
	}
	if !e.hasGby {
		return []byte{}, nil
	}
	vals := make([]types.Datum, 0, len(e.GroupByItems))
	for _, item := range e.GroupByItems {
		v, err := item.Eval(row.Data, e.ctx)
		if err != nil {
			return nil, errors.Trace(err)
		}
		vals = append(vals, v)
	}
	bs, err := codec.EncodeValue([]byte{}, vals...)
	if err != nil {
		return nil, errors.Trace(err)
	}
	return bs, nil
}

// Fetch a single row from src and update each aggregate function.
// If the first return value is false, it means there is no more data from src.
func (e *HashAggExec) innerNext() (ret bool, err error) {
	var srcRow *Row
	if e.Src != nil {
		srcRow, err = e.Src.Next()
		if err != nil {
			return false, errors.Trace(err)
		}
		if srcRow == nil {
			return false, nil
		}
	} else {
		// If Src is nil, only one row should be returned.
		if e.executed {
			return false, nil
		}
	}
	e.executed = true
	groupKey, err := e.getGroupKey(srcRow)
	if err != nil {
		return false, errors.Trace(err)
	}
	if _, ok := e.groupMap[string(groupKey)]; !ok {
		e.groupMap[string(groupKey)] = true
		e.groups = append(e.groups, groupKey)
	}
	for _, af := range e.AggFuncs {
		af.Update(srcRow.Data, groupKey, e.ctx)
	}
	return true, nil
}

// StreamAggExec deals with all the aggregate functions.
// It assumes all the input datas is sorted by group by key.
// When Next() is called, it will return a result for the same group.
type StreamAggExec struct {
	Src                Executor
	schema             expression.Schema
	executed           bool
	hasData            bool
	Ctx                context.Context
	AggFuncs           []expression.AggregationFunction
	GroupByItems       []expression.Expression
	curGroupEncodedKey []byte
	curGroupKey        []types.Datum
	tmpGroupKey        []types.Datum
}

// Close implements the Executor Close interface.
func (e *StreamAggExec) Close() error {
	e.executed = false
	e.hasData = false
	for _, agg := range e.AggFuncs {
		agg.Clear()
	}
	return e.Src.Close()
}

// Schema implements the Executor Schema interface.
func (e *StreamAggExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *StreamAggExec) Next() (*Row, error) {
	if e.executed {
		return nil, nil
	}
	retRow := &Row{Data: make([]types.Datum, 0, len(e.AggFuncs))}
	for {
		row, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		var newGroup bool
		if row == nil {
			newGroup = true
			e.executed = true
		} else {
			e.hasData = true
			newGroup, err = e.meetNewGroup(row)
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
		if newGroup {
			for _, af := range e.AggFuncs {
				retRow.Data = append(retRow.Data, af.GetStreamResult())
			}
		}
		if e.executed {
			break
		}
		for _, af := range e.AggFuncs {
			err = af.StreamUpdate(row.Data, e.Ctx)
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
		if newGroup {
			break
		}
	}
	if !e.hasData && len(e.GroupByItems) > 0 {
		return nil, nil
	}
	return retRow, nil
}

// meetNewGroup returns a value that represents if the new group is different from last group.
func (e *StreamAggExec) meetNewGroup(row *Row) (bool, error) {
	if len(e.GroupByItems) == 0 {
		return false, nil
	}
	e.tmpGroupKey = e.tmpGroupKey[:0]
	matched, firstGroup := true, false
	if len(e.curGroupKey) == 0 {
		matched, firstGroup = false, true
	}
	sc := e.Ctx.GetSessionVars().StmtCtx
	for i, item := range e.GroupByItems {
		v, err := item.Eval(row.Data, e.Ctx)
		if err != nil {
			return false, errors.Trace(err)
		}
		if matched {
			c, err := v.CompareDatum(sc, e.curGroupKey[i])
			if err != nil {
				return false, errors.Trace(err)
			}
			matched = c == 0
		}
		e.tmpGroupKey = append(e.tmpGroupKey, v)
	}
	if matched {
		return false, nil
	}
	e.curGroupKey = e.tmpGroupKey
	var err error
	e.curGroupEncodedKey, err = codec.EncodeValue(e.curGroupEncodedKey[0:0:cap(e.curGroupEncodedKey)], e.curGroupKey...)
	if err != nil {
		return false, errors.Trace(err)
	}
	return !firstGroup, nil
}

// ProjectionExec represents a select fields executor.
type ProjectionExec struct {
	Src      Executor
	schema   expression.Schema
	executed bool
	ctx      context.Context
	exprs    []expression.Expression
}

// Schema implements the Executor Schema interface.
func (e *ProjectionExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *ProjectionExec) Next() (retRow *Row, err error) {
	var rowKeys []*RowKeyEntry
	var srcRow *Row
	if e.Src != nil {
		srcRow, err = e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if srcRow == nil {
			return nil, nil
		}
		rowKeys = srcRow.RowKeys
	} else {
		// If Src is nil, only one row should be returned.
		if e.executed {
			return nil, nil
		}
	}
	e.executed = true
	row := &Row{
		RowKeys: rowKeys,
		Data:    make([]types.Datum, 0, len(e.exprs)),
	}
	for _, expr := range e.exprs {
		val, err := expr.Eval(srcRow.Data, e.ctx)
		if err != nil {
			return nil, errors.Trace(err)
		}
		row.Data = append(row.Data, val)
	}
	return row, nil
}

// Close implements the Executor Close interface.
func (e *ProjectionExec) Close() error {
	if e.Src != nil {
		return e.Src.Close()
	}
	return nil
}

// TableDualExec represents a dual table executor.
type TableDualExec struct {
	schema   expression.Schema
	executed bool
}

// Init implements the Executor Init interface.
func (e *TableDualExec) Init() {
	e.executed = false
}

// Schema implements the Executor Schema interface.
func (e *TableDualExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *TableDualExec) Next() (*Row, error) {
	if e.executed {
		return nil, nil
	}
	e.executed = true
	return &Row{}, nil
}

// Close implements the Executor interface.
func (e *TableDualExec) Close() error {
	return nil
}

// SelectionExec represents a filter executor.
type SelectionExec struct {
	Src       Executor
	Condition expression.Expression
	ctx       context.Context
	schema    expression.Schema
}

// Schema implements the Executor Schema interface.
func (e *SelectionExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *SelectionExec) Next() (*Row, error) {
	for {
		srcRow, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if srcRow == nil {
			return nil, nil
		}
		match, err := expression.EvalBool(e.Condition, srcRow.Data, e.ctx)
		if err != nil {
			return nil, errors.Trace(err)
		}
		if match {
			return srcRow, nil
		}
	}
}

// Close implements the Executor Close interface.
func (e *SelectionExec) Close() error {
	return e.Src.Close()
}

// TableScanExec is a table scan executor without result fields.
type TableScanExec struct {
	t          table.Table
	asName     *model.CIStr
	ctx        context.Context
	ranges     []plan.TableRange
	seekHandle int64
	iter       kv.Iterator
	cursor     int
	schema     expression.Schema
	columns    []*model.ColumnInfo

	isInfoSchema     bool
	infoSchemaRows   [][]types.Datum
	infoSchemaCursor int
}

// Schema implements the Executor Schema interface.
func (e *TableScanExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor interface.
func (e *TableScanExec) Next() (*Row, error) {
	if e.isInfoSchema {
		return e.nextForInfoSchema()
	}
	for {
		if e.cursor >= len(e.ranges) {
			return nil, nil
		}
		ran := e.ranges[e.cursor]
		if e.seekHandle < ran.LowVal {
			e.seekHandle = ran.LowVal
		}
		if e.seekHandle > ran.HighVal {
			e.cursor++
			continue
		}
		handle, found, err := e.t.Seek(e.ctx, e.seekHandle)
		if err != nil {
			return nil, errors.Trace(err)
		}
		if !found {
			return nil, nil
		}
		if handle > ran.HighVal {
			// The handle is out of the current range, but may be in following ranges.
			// We seek to the range that may contains the handle, so we
			// don't need to seek key again.
			inRange := e.seekRange(handle)
			if !inRange {
				// The handle may be less than the current range low value, can not
				// return directly.
				continue
			}
		}
		row, err := e.getRow(handle)
		if err != nil {
			return nil, errors.Trace(err)
		}
		e.seekHandle = handle + 1
		return row, nil
	}
}

func (e *TableScanExec) nextForInfoSchema() (*Row, error) {
	if e.infoSchemaRows == nil {
		columns := make([]*table.Column, len(e.schema))
		for i, v := range e.columns {
			columns[i] = table.ToColumn(v)
		}
		err := e.t.IterRecords(e.ctx, nil, columns, func(h int64, rec []types.Datum, cols []*table.Column) (bool, error) {
			e.infoSchemaRows = append(e.infoSchemaRows, rec)
			return true, nil
		})
		if err != nil {
			return nil, errors.Trace(err)
		}
	}
	if e.infoSchemaCursor >= len(e.infoSchemaRows) {
		return nil, nil
	}
	row := &Row{Data: e.infoSchemaRows[e.infoSchemaCursor]}
	e.infoSchemaCursor++
	return row, nil
}

// seekRange increments the range cursor to the range
// with high value greater or equal to handle.
func (e *TableScanExec) seekRange(handle int64) (inRange bool) {
	for {
		e.cursor++
		if e.cursor >= len(e.ranges) {
			return false
		}
		ran := e.ranges[e.cursor]
		if handle < ran.LowVal {
			return false
		}
		if handle > ran.HighVal {
			continue
		}
		return true
	}
}

func (e *TableScanExec) getRow(handle int64) (*Row, error) {
	row := &Row{}
	var err error

	columns := make([]*table.Column, len(e.schema))
	for i, v := range e.columns {
		columns[i] = table.ToColumn(v)
	}
	row.Data, err = e.t.RowWithCols(e.ctx, handle, columns)
	if err != nil {
		return nil, errors.Trace(err)
	}

	// Put rowKey to the tail of record row.
	rke := &RowKeyEntry{
		Tbl:         e.t,
		Handle:      handle,
		TableAsName: e.asName,
	}
	row.RowKeys = append(row.RowKeys, rke)
	return row, nil
}

// Close implements the Executor Close interface.
func (e *TableScanExec) Close() error {
	e.iter = nil
	e.cursor = 0
	return nil
}

// SortExec represents sorting executor.
type SortExec struct {
	Src     Executor
	ByItems []*plan.ByItems
	Rows    []*orderByRow
	ctx     context.Context
	Idx     int
	fetched bool
	err     error
	schema  expression.Schema
}

// Close implements the Executor Close interface.
func (e *SortExec) Close() error {
	e.fetched = false
	e.Rows = nil
	return e.Src.Close()
}

// Schema implements the Executor Schema interface.
func (e *SortExec) Schema() expression.Schema {
	return e.schema
}

// Len returns the number of rows.
func (e *SortExec) Len() int {
	return len(e.Rows)
}

// Swap implements sort.Interface Swap interface.
func (e *SortExec) Swap(i, j int) {
	e.Rows[i], e.Rows[j] = e.Rows[j], e.Rows[i]
}

// Less implements sort.Interface Less interface.
func (e *SortExec) Less(i, j int) bool {
	sc := e.ctx.GetSessionVars().StmtCtx
	for index, by := range e.ByItems {
		v1 := e.Rows[i].key[index]
		v2 := e.Rows[j].key[index]

		ret, err := v1.CompareDatum(sc, v2)
		if err != nil {
			e.err = errors.Trace(err)
			return true
		}

		if by.Desc {
			ret = -ret
		}

		if ret < 0 {
			return true
		} else if ret > 0 {
			return false
		}
	}

	return false
}

// Next implements the Executor Next interface.
func (e *SortExec) Next() (*Row, error) {
	if !e.fetched {
		for {
			srcRow, err := e.Src.Next()
			if err != nil {
				return nil, errors.Trace(err)
			}
			if srcRow == nil {
				break
			}
			orderRow := &orderByRow{
				row: srcRow,
				key: make([]types.Datum, len(e.ByItems)),
			}
			for i, byItem := range e.ByItems {
				orderRow.key[i], err = byItem.Expr.Eval(srcRow.Data, e.ctx)
				if err != nil {
					return nil, errors.Trace(err)
				}
			}
			e.Rows = append(e.Rows, orderRow)
		}
		sort.Sort(e)
		e.fetched = true
	}
	if e.err != nil {
		return nil, errors.Trace(e.err)
	}
	if e.Idx >= len(e.Rows) {
		return nil, nil
	}
	row := e.Rows[e.Idx].row
	e.Idx++
	return row, nil
}

// TopnExec implements a Top-N algorithm and it is built from a SELECT statement with ORDER BY and LIMIT.
// Instead of sorting all the rows fetched from the table, it keeps the Top-N elements only in a heap to reduce memory usage.
type TopnExec struct {
	SortExec
	limit      *plan.Limit
	totalCount int
	heapSize   int
}

// Less implements heap.Interface Less interface.
func (e *TopnExec) Less(i, j int) bool {
	sc := e.ctx.GetSessionVars().StmtCtx
	for index, by := range e.ByItems {
		v1 := e.Rows[i].key[index]
		v2 := e.Rows[j].key[index]

		ret, err := v1.CompareDatum(sc, v2)
		if err != nil {
			e.err = errors.Trace(err)
			return true
		}

		if by.Desc {
			ret = -ret
		}

		if ret > 0 {
			return true
		} else if ret < 0 {
			return false
		}
	}

	return false
}

// Len implements heap.Interface Len interface.
func (e *TopnExec) Len() int {
	return e.heapSize
}

// Push implements heap.Interface Push interface.
func (e *TopnExec) Push(x interface{}) {
	e.Rows = append(e.Rows, x.(*orderByRow))
	e.heapSize++
}

// Pop implements heap.Interface Pop interface.
func (e *TopnExec) Pop() interface{} {
	e.heapSize--
	return nil
}

// Next implements the Executor Next interface.
func (e *TopnExec) Next() (*Row, error) {
	if !e.fetched {
		e.Idx = int(e.limit.Offset)
		e.totalCount = int(e.limit.Offset + e.limit.Count)
		e.Rows = make([]*orderByRow, 0, e.totalCount+1)
		e.heapSize = 0
		for {
			srcRow, err := e.Src.Next()
			if err != nil {
				return nil, errors.Trace(err)
			}
			if srcRow == nil {
				break
			}
			// build orderRow from srcRow.
			orderRow := &orderByRow{
				row: srcRow,
				key: make([]types.Datum, len(e.ByItems)),
			}
			for i, byItem := range e.ByItems {
				orderRow.key[i], err = byItem.Expr.Eval(srcRow.Data, e.ctx)
				if err != nil {
					return nil, errors.Trace(err)
				}
			}
			if e.totalCount == e.heapSize {
				// An equivalent of Push and Pop. We don't use the standard Push and Pop
				// to reduce the number of comparisons.
				e.Rows = append(e.Rows, orderRow)
				if e.Less(0, e.heapSize) {
					e.Swap(0, e.heapSize)
					heap.Fix(e, 0)
				}
				e.Rows = e.Rows[:e.heapSize]
			} else {
				heap.Push(e, orderRow)
			}
		}
		if e.limit.Offset == 0 {
			sort.Sort(&e.SortExec)
		} else {
			for i := 0; i < int(e.limit.Count) && e.Len() > 0; i++ {
				heap.Pop(e)
			}
		}
		e.fetched = true
	}
	if e.Idx >= len(e.Rows) {
		return nil, nil
	}
	row := e.Rows[e.Idx].row
	e.Idx++
	return row, nil
}

// ApplyExec represents apply executor.
// Apply gets one row from outer executor and gets one row from inner executor according to outer row.
type ApplyExec struct {
	schema      expression.Schema
	Src         Executor
	outerSchema []*expression.CorrelatedColumn
	innerExec   Executor
	// checker checks if an Src row with an inner row matches the condition,
	// and if it needs to check more inner rows.
	checker *conditionChecker
}

// conditionChecker checks if all or any of the row match this condition.
type conditionChecker struct {
	cond        expression.Expression
	trimLen     int
	ctx         context.Context
	all         bool
	dataHasNull bool
}

// Check returns finished for checking if the input row can determine the final result,
// and returns data for the evaluation result.
func (c *conditionChecker) check(rowData []types.Datum) (finished bool, data types.Datum, err error) {
	data, err = c.cond.Eval(rowData, c.ctx)
	if err != nil {
		return false, data, errors.Trace(err)
	}
	var matched int64
	if data.IsNull() {
		c.dataHasNull = true
		matched = 0
	} else {
		matched, err = data.ToBool(c.ctx.GetSessionVars().StmtCtx)
		if err != nil {
			return false, data, errors.Trace(err)
		}
	}
	return (matched != 0) != c.all, data, nil
}

// Reset resets dataHasNull to false, so it can be reused for the next row from ApplyExec.Src.
func (c *conditionChecker) reset() {
	c.dataHasNull = false
}

// Schema implements the Executor Schema interface.
func (e *ApplyExec) Schema() expression.Schema {
	return e.schema
}

// Close implements the Executor Close interface.
func (e *ApplyExec) Close() error {
	if e.checker != nil {
		e.checker.dataHasNull = false
	}
	return e.Src.Close()
}

// Next implements the Executor Next interface.
func (e *ApplyExec) Next() (*Row, error) {
	srcRow, err := e.Src.Next()
	if err != nil {
		return nil, errors.Trace(err)
	}
	if srcRow == nil {
		return nil, nil
	}
	for {
		for _, col := range e.outerSchema {
			idx := col.Index
			*col.Data = srcRow.Data[idx]
		}
		innerRow, err := e.innerExec.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		trimLen := len(srcRow.Data)
		if innerRow != nil {
			srcRow.Data = append(srcRow.Data, innerRow.Data...)
		}
		if e.checker == nil {
			e.innerExec.Close()
			return srcRow, nil
		}
		if innerRow == nil {
			// When inner exec finishes, we need to append a result column to true, false or NULL.
			var result types.Datum
			if e.checker.all {
				result = types.NewDatum(true)
			} else {
				// If 'any' meets a null value, the result will be null.
				if e.checker.dataHasNull {
					result = types.NewDatum(nil)
				} else {
					result = types.NewDatum(false)
				}
			}
			srcRow.Data = append(srcRow.Data, result)
			e.checker.reset()
			e.innerExec.Close()
			return srcRow, nil
		}
		finished, data, err := e.checker.check(srcRow.Data)
		if err != nil {
			return nil, errors.Trace(err)
		}
		srcRow.Data = srcRow.Data[:trimLen]
		if finished {
			e.checker.reset()
			e.innerExec.Close()
			srcRow.Data = append(srcRow.Data, data)
			return srcRow, nil
		}
	}
}

// ExistsExec represents exists executor.
type ExistsExec struct {
	schema    expression.Schema
	Src       Executor
	evaluated bool
}

// Schema implements the Executor Schema interface.
func (e *ExistsExec) Schema() expression.Schema {
	return e.schema
}

// Close implements the Executor Close interface.
func (e *ExistsExec) Close() error {
	e.evaluated = false
	return e.Src.Close()
}

// Next implements the Executor Next interface.
// We always return one row with one column which has true or false value.
func (e *ExistsExec) Next() (*Row, error) {
	if !e.evaluated {
		e.evaluated = true
		srcRow, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		return &Row{Data: []types.Datum{types.NewDatum(srcRow != nil)}}, nil
	}
	return nil, nil
}

// MaxOneRowExec checks if the number of rows that a query returns is at maximum one.
// It's built from subquery expression.
type MaxOneRowExec struct {
	schema    expression.Schema
	Src       Executor
	evaluated bool
}

// Schema implements the Executor Schema interface.
func (e *MaxOneRowExec) Schema() expression.Schema {
	return e.schema
}

// Close implements the Executor Close interface.
func (e *MaxOneRowExec) Close() error {
	e.evaluated = false
	return e.Src.Close()
}

// Next implements the Executor Next interface.
func (e *MaxOneRowExec) Next() (*Row, error) {
	if !e.evaluated {
		e.evaluated = true
		srcRow, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if srcRow == nil {
			return &Row{Data: make([]types.Datum, len(e.schema))}, nil
		}
		srcRow1, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if srcRow1 != nil {
			return nil, errors.New("subquery returns more than 1 row")
		}
		return srcRow, nil
	}
	return nil, nil
}

// TrimExec truncates extra columns in the Src rows.
// Some columns in src rows are not needed in the result.
// For example, in the 'SELECT a from t order by b' statement,
// 'b' is needed for ordering, but not needed in the result.
type TrimExec struct {
	schema expression.Schema
	Src    Executor
	len    int
}

// Schema implements the Executor Schema interface.
func (e *TrimExec) Schema() expression.Schema {
	return e.schema
}

// Close implements the Executor Close interface.
func (e *TrimExec) Close() error {
	return e.Src.Close()
}

// Next implements the Executor Next interface.
func (e *TrimExec) Next() (*Row, error) {
	row, err := e.Src.Next()
	if err != nil {
		return nil, errors.Trace(err)
	}
	if row == nil {
		return nil, nil
	}
	row.Data = row.Data[:e.len]
	return row, nil
}

// UnionExec represents union executor.
// UnionExec has multiple source Executors, it executes them sequentially, and do conversion to the same type
// as source Executors may has different field type, we need to do conversion.
type UnionExec struct {
	schema expression.Schema
	Srcs   []Executor
	ctx    context.Context
	cursor int
}

// Schema implements the Executor Schema interface.
func (e *UnionExec) Schema() expression.Schema {
	return e.schema
}

// Next implements the Executor Next interface.
func (e *UnionExec) Next() (*Row, error) {
	for {
		if e.cursor >= len(e.Srcs) {
			return nil, nil
		}
		sel := e.Srcs[e.cursor]
		row, err := sel.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if row == nil {
			e.cursor++
			continue
		}
		if e.cursor != 0 {
			for i := range row.Data {
				// The column value should be casted as the same type of the first select statement in corresponding position.
				col := e.schema[i]
				var val types.Datum
				val, err = row.Data[i].ConvertTo(e.ctx.GetSessionVars().StmtCtx, col.RetType)
				if err != nil {
					return nil, errors.Trace(err)
				}
				row.Data[i] = val
			}
		}
		return row, nil
	}
}

// Close implements the Executor Close interface.
func (e *UnionExec) Close() error {
	e.cursor = 0
	for _, sel := range e.Srcs {
		er := sel.Close()
		if er != nil {
			return errors.Trace(er)
		}
	}
	return nil
}

// DummyScanExec returns zero results, when some where condition never match, there won't be any
// rows to return, so DummyScan is used to avoid real scan on KV.
type DummyScanExec struct {
	schema expression.Schema
}

// Schema implements the Executor Schema interface.
func (e *DummyScanExec) Schema() expression.Schema {
	return e.schema
}

// Close implements the Executor Close interface.
func (e *DummyScanExec) Close() error {
	return nil
}

// Next implements the Executor Next interface.
func (e *DummyScanExec) Next() (*Row, error) {
	return nil, nil
}

// CacheExec represents Cache executor.
// it stores the return values of the executor of its child node.
type CacheExec struct {
	schema      expression.Schema
	Src         Executor
	storedRows  []*Row
	cursor      int
	srcFinished bool
}

// Schema implements the Executor Schema interface.
func (e *CacheExec) Schema() expression.Schema {
	return e.schema
}

// Close implements the Executor Close interface.
func (e *CacheExec) Close() error {
	e.cursor = 0
	return nil
}

// Next implements the Executor Next interface.
func (e *CacheExec) Next() (*Row, error) {
	if e.srcFinished && e.cursor >= len(e.storedRows) {
		return nil, nil
	}
	if !e.srcFinished {
		row, err := e.Src.Next()
		if err != nil {
			return nil, errors.Trace(err)
		}
		if row == nil {
			e.srcFinished = true
			err := e.Src.Close()
			if err != nil {
				return nil, errors.Trace(err)
			}
		}
		e.storedRows = append(e.storedRows, row)
	}
	row := e.storedRows[e.cursor]
	e.cursor++
	return row, nil
}