pgplan

package
v1.6.0 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Sep 11, 2024 License: MIT Imports: 4 Imported by: 0

Documentation

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

This section is empty.

Types

type Agg 

type Agg struct{ Plan }

func (Agg) Kind 

func (Agg) Kind() NodeKind

type Append 

type Append struct{ Plan }

Append is the concatenation of the results of sub-plans. Combine the results of the child operations. This can be the result of an explicit UNION ALL statement, or the need for a parent operation to consume the results of two or more children together. https://www.pgmustard.com/docs/explain/append

func (Append) Kind 

func (Append) Kind() NodeKind

type BadNode 

type BadNode struct{ Plan }

BadNode is returned whenever a plan is not parseable.

func (BadNode) Kind 

func (BadNode) Kind() NodeKind

type BitmapAnd 

type BitmapAnd struct{ Plan }

func (BitmapAnd) Kind 

func (BitmapAnd) Kind() NodeKind

type BitmapHeapScan 

type BitmapHeapScan struct{ Plan }

func (BitmapHeapScan) Kind 

func (BitmapHeapScan) Kind() NodeKind

type BitmapIndexScan 

type BitmapIndexScan struct{ Plan }

func (BitmapIndexScan) Kind 

func (BitmapIndexScan) Kind() NodeKind

type BitmapOr 

type BitmapOr struct{ Plan }

func (BitmapOr) Kind 

func (BitmapOr) Kind() NodeKind

type CteScan 

type CteScan struct{ Plan }

func (CteScan) Kind 

func (CteScan) Kind() NodeKind

type CustomScan 

type CustomScan struct{ Plan }

func (CustomScan) Kind 

func (CustomScan) Kind() NodeKind

type ForeignScan 

type ForeignScan struct{ Plan }

func (ForeignScan) Kind 

func (ForeignScan) Kind() NodeKind

type FunctionScan 

type FunctionScan struct{ Plan }

func (FunctionScan) Kind 

func (FunctionScan) Kind() NodeKind

type Gather 

type Gather struct{ Plan }

func (Gather) Kind 

func (Gather) Kind() NodeKind

type GatherMerge 

type GatherMerge struct{ Plan }

func (GatherMerge) Kind 

func (GatherMerge) Kind() NodeKind

type Group 

type Group struct{ Plan }

func (Group) Kind 

func (Group) Kind() NodeKind

type Hash 

type Hash struct{ Plan }

func (Hash) Kind 

func (Hash) Kind() NodeKind

type HashJoin 

type HashJoin struct{ Plan }

func (HashJoin) Kind 

func (HashJoin) Kind() NodeKind

type IncrementalSort 

type IncrementalSort struct{ Plan }

func (IncrementalSort) Kind 

func (IncrementalSort) Kind() NodeKind

type IndexOnlyScan 

type IndexOnlyScan struct{ Plan }

func (IndexOnlyScan) Kind 

func (IndexOnlyScan) Kind() NodeKind

type IndexScan 

type IndexScan struct{ Plan }

func (IndexScan) Kind 

func (IndexScan) Kind() NodeKind

type Join 

type Join struct{ Plan }

func (Join) Kind 

func (Join) Kind() NodeKind

type Limit 

type Limit struct{ Plan }

func (Limit) Kind 

func (Limit) Kind() NodeKind

type LockRows 

type LockRows struct{ Plan }

func (LockRows) Kind 

func (LockRows) Kind() NodeKind

type Material 

type Material struct{ Plan }

func (Material) Kind 

func (Material) Kind() NodeKind

type MergeAppend 

type MergeAppend struct {
	Plan
	SortKey []string
}

MergeAppend combines the sorted results of the child operations, in a way that preserves their sort order. Can be used for combining already-sorted rows from table partitions. https://www.pgmustard.com/docs/explain/merge-append

func (MergeAppend) Kind 

func (MergeAppend) Kind() NodeKind

type MergeJoin 

type MergeJoin struct{ Plan }

func (MergeJoin) Kind 

func (MergeJoin) Kind() NodeKind

type ModifyTable 

type ModifyTable struct {
	Plan
	Operation    Operation
	RelationName string
	Schema       string
	Alias        string
}

ModifyTable applies rows produced by subplan(s) to result table(s), by inserting, updating, or deleting.

func (ModifyTable) Kind 

func (ModifyTable) Kind() NodeKind

type NamedTuplestoreScan 

type NamedTuplestoreScan struct{ Plan }

func (NamedTuplestoreScan) Kind 

func (NamedTuplestoreScan) Kind() NodeKind

type NestLoop 

type NestLoop struct{ Plan }

func (NestLoop) Kind 

func (NestLoop) Kind() NodeKind

type Node 

type Node interface {
	Kind() NodeKind
	// Output returns the output columns of the node. The format of each output
	// column depends on the type of node.
	Output() []string
	// Children returns the direct children of the node, or nil if none exist.
	Children() []Node
}

Node is the super-type of all Postgres plan nodes. https://doxygen.postgresql.org/nodes_8h.html#a83ba1e84fa23f6619c3d29036b160919

func ExplainQuery 

func ExplainQuery(conn *pgx.Conn, sql string) (Node, error)

ExplainQuery executes an explain query and parses the plan.

func ParseNode 

func ParseNode(rawPlan map[string]interface{}) (Node, error)

type NodeKind 

type NodeKind string

NodeKind is the top-level node plan type that Postgres plans for executing query. https://www.postgresql.org/docs/13/executor.html 

const (
	KindBadNode             NodeKind = "BadNode"
	KindResult              NodeKind = "Result"
	KindProjectSet          NodeKind = "ProjectSet"
	KindModifyTable         NodeKind = "ModifyTable"
	KindAppend              NodeKind = "Append"
	KindMergeAppend         NodeKind = "MergeAppend"
	KindRecursiveUnion      NodeKind = "RecursiveUnion"
	KindBitmapAnd           NodeKind = "BitmapAnd"
	KindBitmapOr            NodeKind = "BitmapOr"
	KindScan                NodeKind = "Scan"
	KindSeqScan             NodeKind = "SeqScan"
	KindSampleScan          NodeKind = "SampleScan"
	KindIndexScan           NodeKind = "IndexScan"
	KindIndexOnlyScan       NodeKind = "IndexOnlyScan"
	KindBitmapIndexScan     NodeKind = "BitmapIndexScan"
	KindBitmapHeapScan      NodeKind = "BitmapHeapScan"
	KindTidScan             NodeKind = "TidScan"
	KindSubqueryScan        NodeKind = "SubqueryScan"
	KindFunctionScan        NodeKind = "FunctionScan"
	KindValuesScan          NodeKind = "ValuesScan"
	KindTableFuncScan       NodeKind = "TableFuncScan"
	KindCteScan             NodeKind = "CteScan"
	KindNamedTuplestoreScan NodeKind = "NamedTuplestoreScan"
	KindWorkTableScan       NodeKind = "WorkTableScan"
	KindForeignScan         NodeKind = "ForeignScan"
	KindCustomScan          NodeKind = "CustomScan"
	KindJoin                NodeKind = "Join"
	KindNestLoop            NodeKind = "NestLoop"
	KindMergeJoin           NodeKind = "MergeJoin"
	KindHashJoin            NodeKind = "HashJoin"
	KindMaterial            NodeKind = "Material"
	KindSort                NodeKind = "Sort"
	KindIncrementalSort     NodeKind = "IncrementalSort"
	KindGroup               NodeKind = "Group"
	KindAgg                 NodeKind = "Agg"
	KindWindowAgg           NodeKind = "WindowAgg"
	KindUnique              NodeKind = "Unique"
	KindGather              NodeKind = "Gather"
	KindGatherMerge         NodeKind = "GatherMerge"
	KindHash                NodeKind = "Hash"
	KindSetOp               NodeKind = "SetOp"
	KindLockRows            NodeKind = "LockRows"
	KindLimit               NodeKind = "Limit"
)

type Operation 

type Operation string

Operation for a ModifyTable node. 

const (
	OperationInsert Operation = "Insert"
	OperationUpdate Operation = "Update"
	OperationDelete Operation = "Delete"
)

type ParentRelationship 

type ParentRelationship string

ParentRelationship describes why this operation needs to be run in order to facilitate the parent operation. 

const (
	// ParentRelationshipNone means this node is a top-level node. All nodes with
	// a parent have set relationship that is not none.
	ParentRelationshipNone ParentRelationship = ""
	// ParentRelationshipOuter is the most common node. It means take in the rows
	// from this operation as input, process them and pass them on.
	ParentRelationshipOuter ParentRelationship = "Outer"
	// ParentRelationshipInner is only (but always) on second child of join
	// operations. Means a node is the inner part of a loop.
	ParentRelationshipInner ParentRelationship = "Inner"
	// ParentRelationshipMember is for all children of Append and ModifyTable
	// nodes.
	ParentRelationshipMember ParentRelationship = "Member"
	// ParentRelationshipInitPlan is calculations performed before query starts
	// executing.
	ParentRelationshipInitPlan ParentRelationship = "InitPlan"
	// ParentRelationshipSubquery means the node is a subquery of a parent node.
	// Since Postgres always uses subquery scans to feed subquery data to parent
	// queries, only ever appears on the children of subquery scans.
	ParentRelationshipSubquery ParentRelationship = "Subquery"
	// ParentRelationshipSubPlan is like a Subquery, represents a new query, but
	// used when a subquery scan is not necessary.
	ParentRelationshipSubPlan ParentRelationship = "SubPlan"
)

type Plan 

type Plan struct {
	// Estimated execution costs for plan (see costsize.c for more info).
	StartupCost float64 // cost expended before fetching any tuples
	TotalCost   float64 // total cost (assuming all tuples fetched)

	// Planner's estimate of result size of this plan step.
	PlanRows  float64 // number of rows plan is expected to emit
	PlanWidth int     // average row width in bytes

	// Information needed for parallel query.
	ParallelAware bool // engage parallel-aware logic?
	ParallelSafe  bool // OK to use as part of parallel plan?

	// Relationship from this node to its parent. Always set for descendant nodes.
	ParentRelationship ParentRelationship

	// How to execute a node. Used for Agg and SetOp nodes.
	Strategy Strategy

	// Custom plan, if any.
	CustomPlanProvider string

	// The column expressions (target list), if any.
	Outs []string

	// Child nodes, if any.
	Nodes []Node
}

Plan nodes "derive" from the Plan structure by having the Plan structure as the first field. This ensures that everything works when nodes are cast to Plan's. (node pointers are frequently cast to Plan* when passed around generically in the executor) https://sourcegraph.com/github.com/postgres/postgres@8facf1ea00b7a0c08c755a0392212b83e04ae28a/-/blob/src/include/nodes/plannodes.h#L110:16

func (Plan) Children 

func (p Plan) Children() []Node

func (Plan) Output 

func (p Plan) Output() []string

type ProjectSet 

type ProjectSet struct{ Plan }

ProjectSet appears when the SELECT or ORDER BY clause of the query. They basically just execute the set-returning function(s) for each tuple until none of the functions return any more records. https://www.postgresql.org/message-id/CAKJS1f9pWUwxaD%2B0kxOOUuwaBcpGQtCKi3DKE8ob_uHN-JTJhw%40mail.gmail.com

func (ProjectSet) Kind 

func (ProjectSet) Kind() NodeKind

type RecursiveUnion 

type RecursiveUnion struct{ Plan }

func (RecursiveUnion) Kind 

func (RecursiveUnion) Kind() NodeKind

type Result 

type Result struct{ Plan }

Result is if no outer plan, evaluate a variable-free targetlist. If outer plan, return tuples from outer plan (after a level of projection as shown by targetlist). https://sourcegraph.com/github.com/postgres/postgres@8facf1ea00b7a0c08c755a0392212b83e04ae28a/-/blob/src/include/nodes/plannodes.h#L180:1

func (Result) Kind 

func (Result) Kind() NodeKind

type SampleScan 

type SampleScan struct{ Plan }

func (SampleScan) Kind 

func (SampleScan) Kind() NodeKind

type Scan 

type Scan struct{ Plan }

func (Scan) Kind 

func (Scan) Kind() NodeKind

type SeqScan 

type SeqScan struct{ Plan }

func (SeqScan) Kind 

func (SeqScan) Kind() NodeKind

type SetOp 

type SetOp struct{ Plan }

func (SetOp) Kind 

func (SetOp) Kind() NodeKind

type Sort 

type Sort struct {
	Plan
	SortKey []string
}

func (Sort) Kind 

func (Sort) Kind() NodeKind

type Strategy 

type Strategy string

Strategy determines overall execution strategies for Agg plan nodes and SetOp nodes. https://source graph.com/github.com/postgres/postgres@8facf1ea00b7a0c08c755a0392212b83e04ae28a/-/blob/src/include/nodes/nodes.h?subtree=true#L759:14 

const (
	// StrategyPlain is a simple agg across all input rows.
	StrategyPlain Strategy = "Plain"
	// StrategySorted is a for grouped agg and SetOp, input must be sorted.
	StrategySorted Strategy = "Sorted"
	// StrategyHashed is a for grouped agg and SetOp, uses internal hashtable.
	StrategyHashed Strategy = "Hashed"
	// StrategyMixed is a grouped agg, hash and sort both used.
	StrategyMixed Strategy = "Mixed"
	// StrategyUnknown is a for unknown aggregates.
	StrategyUnknown Strategy = "???"
)

type SubqueryScan 

type SubqueryScan struct{ Plan }

func (SubqueryScan) Kind 

func (SubqueryScan) Kind() NodeKind

type TableFuncScan 

type TableFuncScan struct{ Plan }

func (TableFuncScan) Kind 

func (TableFuncScan) Kind() NodeKind

type TidScan 

type TidScan struct{ Plan }

func (TidScan) Kind 

func (TidScan) Kind() NodeKind

type Unique 

type Unique struct{ Plan }

Unique is a very simple node type that just filters out duplicate tuples from a stream of sorted tuples from its subplan. https://sourcegraph.com/github.com/postgres/postgres@8facf1ea00b7a0c08c755a0392212b83e04ae28a/-/blob/src/include/nodes/plannodes.h?subtree=true#L864:16

func (Unique) Kind 

func (Unique) Kind() NodeKind

type ValuesScan 

type ValuesScan struct{ Plan }

func (ValuesScan) Kind 

func (ValuesScan) Kind() NodeKind

type WindowAgg 

type WindowAgg struct{ Plan }

func (WindowAgg) Kind 

func (WindowAgg) Kind() NodeKind

type WorkTableScan 

type WorkTableScan struct{ Plan }

func (WorkTableScan) Kind 

func (WorkTableScan) Kind() NodeKind

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.