Files
tidb/optimizer/plan/range.go
2015-12-09 17:03:16 +08:00

428 lines
12 KiB
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 plan
import (
"fmt"
"sort"
"github.com/pingcap/tidb/ast"
"github.com/pingcap/tidb/parser/opcode"
"github.com/pingcap/tidb/util/types"
)
type rangePoint struct {
value interface{}
excl bool // exclude
start bool
}
func (rp rangePoint) String() string {
if rp.start {
symbol := "["
if rp.excl {
symbol = "("
}
return fmt.Sprintf("%s%v", symbol, rp.value)
}
symbol := "]"
if rp.excl {
symbol = ")"
}
return fmt.Sprintf("%v%s", rp.value, symbol)
}
type rangePointSorter struct {
points []rangePoint
err error
}
func (r *rangePointSorter) Len() int {
return len(r.points)
}
func (r *rangePointSorter) Less(i, j int) bool {
a := r.points[i]
b := r.points[j]
if a.value == nil && b.value == nil {
return r.equalValueLess(a, b)
} else if b.value == nil {
return false
} else if a.value == nil {
return true
}
// a and b both not nil
if a.value == MinNotNullVal && b.value == MinNotNullVal {
return r.equalValueLess(a, b)
} else if b.value == MinNotNullVal {
return false
} else if a.value == MinNotNullVal {
return true
}
// a and b both not min value
if a.value == MaxVal && b.value == MaxVal {
return r.equalValueLess(a, b)
} else if a.value == MaxVal {
return false
} else if b.value == MaxVal {
return true
}
n, err := types.Compare(a.value, b.value)
if err != nil {
r.err = err
return true
}
if n == 0 {
return r.equalValueLess(a, b)
}
return n < 0
}
func (r *rangePointSorter) equalValueLess(a, b rangePoint) bool {
if a.start && b.start {
return !a.excl && b.excl
} else if a.start {
return !b.excl
} else if b.start {
return a.excl || b.excl
}
return a.excl && !b.excl
}
func (r *rangePointSorter) Swap(i, j int) {
r.points[i], r.points[j] = r.points[j], r.points[i]
}
type rangeBuilder struct {
err error
}
func (r *rangeBuilder) build(expr ast.ExprNode) []rangePoint {
switch x := expr.(type) {
case *ast.BinaryOperationExpr:
return r.buildFromBinop(x)
case *ast.PatternInExpr:
return r.buildFromIn(x)
case *ast.ParenthesesExpr:
return r.build(x.Expr)
case *ast.BetweenExpr:
return r.buildFromBetween(x)
case *ast.IsNullExpr:
return r.buildFromIsNull(x)
case *ast.IsTruthExpr:
return r.buildFromIsTruth(x)
case *ast.PatternLikeExpr:
return r.buildFromPatternLike(x)
case *ast.ColumnNameExpr:
return r.buildFromColumnName(x)
}
return fullRange
}
func (r *rangeBuilder) buildFromBinop(x *ast.BinaryOperationExpr) []rangePoint {
if x.Op == opcode.OrOr {
return r.union(r.build(x.L), r.build(x.R))
} else if x.Op == opcode.AndAnd {
return r.intersection(r.build(x.L), r.build(x.R))
}
// This has been checked that the binary operation is comparison operation, and one of
// the operand is column name expression.
var value interface{}
var op opcode.Op
if _, ok := x.L.(*ast.ValueExpr); ok {
value = x.L.GetValue()
switch x.Op {
case opcode.GE:
op = opcode.LE
case opcode.GT:
op = opcode.LT
case opcode.LT:
op = opcode.GT
case opcode.LE:
op = opcode.GE
default:
op = x.Op
}
} else {
value = x.R.GetValue()
op = x.Op
}
switch op {
case opcode.EQ:
startPoint := rangePoint{value: value, start: true}
endPoint := rangePoint{value: value}
return []rangePoint{startPoint, endPoint}
case opcode.NE:
startPoint1 := rangePoint{value: MinNotNullVal, start: true}
endPoint1 := rangePoint{value: value, excl: true}
startPoint2 := rangePoint{value: value, start: true, excl: true}
endPoint2 := rangePoint{value: MaxVal}
return []rangePoint{startPoint1, endPoint1, startPoint2, endPoint2}
case opcode.LT:
startPoint := rangePoint{value: MinNotNullVal, start: true}
endPoint := rangePoint{value: value, excl: true}
return []rangePoint{startPoint, endPoint}
case opcode.LE:
startPoint := rangePoint{value: MinNotNullVal, start: true}
endPoint := rangePoint{value: value}
return []rangePoint{startPoint, endPoint}
case opcode.GT:
startPoint := rangePoint{value: value, start: true, excl: true}
endPoint := rangePoint{value: MaxVal}
return []rangePoint{startPoint, endPoint}
case opcode.GE:
startPoint := rangePoint{value: value, start: true}
endPoint := rangePoint{value: MaxVal}
return []rangePoint{startPoint, endPoint}
}
return nil
}
func (r *rangeBuilder) buildFromIn(x *ast.PatternInExpr) []rangePoint {
if x.Not {
r.err = ErrUnsupportedType.Gen("NOT IN is not supported")
return fullRange
}
var rangePoints []rangePoint
for _, v := range x.List {
startPoint := rangePoint{value: v.GetValue(), start: true}
endPoint := rangePoint{value: v.GetValue()}
rangePoints = append(rangePoints, startPoint, endPoint)
}
sorter := rangePointSorter{points: rangePoints}
sort.Sort(&sorter)
if sorter.err != nil {
r.err = sorter.err
}
return rangePoints
}
func (r *rangeBuilder) buildFromBetween(x *ast.BetweenExpr) []rangePoint {
if x.Not {
binop1 := &ast.BinaryOperationExpr{Op: opcode.LT, L: x.Expr, R: x.Left}
binop2 := &ast.BinaryOperationExpr{Op: opcode.GT, L: x.Expr, R: x.Right}
range1 := r.buildFromBinop(binop1)
range2 := r.buildFromBinop(binop2)
return r.union(range1, range2)
}
binop1 := &ast.BinaryOperationExpr{Op: opcode.GE, L: x.Expr, R: x.Left}
binop2 := &ast.BinaryOperationExpr{Op: opcode.LE, L: x.Expr, R: x.Right}
range1 := r.buildFromBinop(binop1)
range2 := r.buildFromBinop(binop2)
return r.intersection(range1, range2)
}
func (r *rangeBuilder) buildFromIsNull(x *ast.IsNullExpr) []rangePoint {
if x.Not {
startPoint := rangePoint{value: MinNotNullVal, start: true}
endPoint := rangePoint{value: MaxVal}
return []rangePoint{startPoint, endPoint}
}
startPoint := rangePoint{start: true}
endPoint := rangePoint{}
return []rangePoint{startPoint, endPoint}
}
func (r *rangeBuilder) buildFromIsTruth(x *ast.IsTruthExpr) []rangePoint {
if x.True != 0 {
if x.Not {
startPoint1 := rangePoint{start: true}
endPoint1 := rangePoint{}
startPoint2 := rangePoint{value: int64(0), start: true}
endPoint2 := rangePoint{value: int64(0)}
return []rangePoint{startPoint1, endPoint1, startPoint2, endPoint2}
}
startPoint1 := rangePoint{value: MinNotNullVal, start: true}
endPoint1 := rangePoint{value: int64(0), excl: true}
startPoint2 := rangePoint{value: int64(0), excl: true, start: true}
endPoint2 := rangePoint{value: MaxVal}
return []rangePoint{startPoint1, endPoint1, startPoint2, endPoint2}
}
if x.Not {
startPoint1 := rangePoint{start: true}
endPoint1 := rangePoint{value: int64(0), excl: true}
startPoint2 := rangePoint{value: int64(0), start: true, excl: true}
endPoint2 := rangePoint{value: MaxVal}
return []rangePoint{startPoint1, endPoint1, startPoint2, endPoint2}
}
startPoint := rangePoint{value: int64(0), start: true}
endPoint := rangePoint{value: int64(0)}
return []rangePoint{startPoint, endPoint}
}
func (r *rangeBuilder) buildFromPatternLike(x *ast.PatternLikeExpr) []rangePoint {
if x.Not {
// Pattern not like is not supported.
r.err = ErrUnsupportedType.Gen("NOT LIKE is not supported.")
return fullRange
}
pattern := x.Pattern.GetValue().(string)
lowValue := make([]byte, 0, len(pattern))
// unscape the pattern
var exclude bool
for i := 0; i < len(pattern); i++ {
if pattern[i] == x.Escape {
i++
if i < len(pattern) {
lowValue = append(lowValue, pattern[i])
} else {
lowValue = append(lowValue, x.Escape)
}
continue
}
if pattern[i] == '%' {
break
} else if pattern[i] == '_' {
exclude = true
break
}
lowValue = append(lowValue, pattern[i])
}
if len(lowValue) == 0 {
return []rangePoint{{value: MinNotNullVal, start: true}, {value: MaxVal}}
}
startPoint := rangePoint{value: string(lowValue), start: true, excl: exclude}
highValue := make([]byte, len(lowValue))
copy(highValue, lowValue)
endPoint := rangePoint{excl: true}
for i := len(highValue) - 1; i >= 0; i-- {
highValue[i]++
if highValue[i] != 0 {
endPoint.value = string(highValue)
break
}
if i == 0 {
endPoint.value = MaxVal
break
}
}
return []rangePoint{startPoint, endPoint}
}
func (r *rangeBuilder) buildFromColumnName(x *ast.ColumnNameExpr) []rangePoint {
// column name expression is equivalent to column name is true.
startPoint1 := rangePoint{value: MinNotNullVal, start: true}
endPoint1 := rangePoint{value: int64(0), excl: true}
startPoint2 := rangePoint{value: int64(0), excl: true, start: true}
endPoint2 := rangePoint{value: MaxVal}
return []rangePoint{startPoint1, endPoint1, startPoint2, endPoint2}
}
func (r *rangeBuilder) intersection(a, b []rangePoint) []rangePoint {
return r.merge(a, b, false)
}
func (r *rangeBuilder) union(a, b []rangePoint) []rangePoint {
return r.merge(a, b, true)
}
func (r *rangeBuilder) merge(a, b []rangePoint, union bool) []rangePoint {
sorter := rangePointSorter{points: append(a, b...)}
sort.Sort(&sorter)
if sorter.err != nil {
r.err = sorter.err
return nil
}
var (
merged []rangePoint
inRangeCount int
requiredInRangeCount int
)
if union {
requiredInRangeCount = 1
} else {
requiredInRangeCount = 2
}
for _, val := range sorter.points {
if val.start {
inRangeCount++
if inRangeCount == requiredInRangeCount {
// just reached the required in range count, a new range started.
merged = append(merged, val)
}
} else {
if inRangeCount == requiredInRangeCount {
// just about to leave the required in range count, the range is ended.
merged = append(merged, val)
}
inRangeCount--
}
}
return merged
}
// buildIndexRanges build index ranges from range points.
// Only the first column in the index is built, extra column ranges will be appended by
// appendIndexRanges.
func (r *rangeBuilder) buildIndexRanges(rangePoints []rangePoint) []*IndexRange {
indexRanges := make([]*IndexRange, 0, len(rangePoints)/2)
for i := 0; i < len(rangePoints); i += 2 {
startPoint := rangePoints[i]
endPoint := rangePoints[i+1]
ir := &IndexRange{
LowVal: []interface{}{startPoint.value},
LowExclude: startPoint.excl,
HighVal: []interface{}{endPoint.value},
HighExclude: endPoint.excl,
}
indexRanges = append(indexRanges, ir)
}
return indexRanges
}
// appendIndexRanges appends additional column ranges for multi-column index.
// The additional column ranges can only be appended to point ranges.
// for example we have an index (a, b), if the condition is (a > 1 and b = 2)
// then we can not build a conjunctive ranges for this index.
func (r *rangeBuilder) appendIndexRanges(origin []*IndexRange, rangePoints []rangePoint) []*IndexRange {
var newIndexRanges []*IndexRange
for i := 0; i < len(origin); i++ {
oRange := origin[i]
if !oRange.IsPoint() {
newIndexRanges = append(newIndexRanges, oRange)
} else {
newIndexRanges = append(newIndexRanges, r.appendIndexRange(oRange, rangePoints)...)
}
}
return newIndexRanges
}
func (r *rangeBuilder) appendIndexRange(origin *IndexRange, rangePoints []rangePoint) []*IndexRange {
newRanges := make([]*IndexRange, 0, len(rangePoints)/2)
for i := 0; i < len(rangePoints); i += 2 {
startPoint := rangePoints[i]
lowVal := make([]interface{}, len(origin.LowVal)+1)
copy(lowVal, origin.LowVal)
lowVal[len(origin.LowVal)] = startPoint.value
endPoint := rangePoints[i+1]
highVal := make([]interface{}, len(origin.HighVal)+1)
copy(highVal, origin.HighVal)
highVal[len(origin.HighVal)] = endPoint.value
ir := &IndexRange{
LowVal: lowVal,
LowExclude: startPoint.excl,
HighVal: highVal,
HighExclude: endPoint.excl,
}
newRanges = append(newRanges, ir)
}
return newRanges
}