org.apache.spark.sql

Class SparkSessionExtensions

Object

org.apache.spark.sql.SparkSessionExtensions

public class SparkSessionExtensions
extends Object

:: Experimental :: Holder for injection points to the SparkSession. We make NO guarantee about the stability regarding binary compatibility and source compatibility of methods here.

This current provides the following extension points:

• Analyzer Rules.
• Check Analysis Rules.
• Cache Plan Normalization Rules.
• Optimizer Rules.
• Pre CBO Rules.
• Planning Strategies.
• Customized Parser.
• (External) Catalog listeners.
• Columnar Rules.
• Adaptive Query Post Planner Strategy Rules.
• Adaptive Query Stage Preparation Rules.
• Adaptive Query Execution Runtime Optimizer Rules.
• Adaptive Query Stage Optimizer Rules.

The extensions can be used by calling withExtensions on the SparkSession.Builder, for example:

   SparkSession.builder()
     .master("...")
     .config("...", true)
     .withExtensions { extensions =>
       extensions.injectResolutionRule { session =>
         ...
       }
       extensions.injectParser { (session, parser) =>
         ...
       }
     }
     .getOrCreate()
 

The extensions can also be used by setting the Spark SQL configuration property spark.sql.extensions. Multiple extensions can be set using a comma-separated list. For example:

   SparkSession.builder()
     .master("...")
     .config("spark.sql.extensions", "org.example.MyExtensions,org.example.YourExtensions")
     .getOrCreate()

   class MyExtensions extends Function1[SparkSessionExtensions, Unit] {
     override def apply(extensions: SparkSessionExtensions): Unit = {
       extensions.injectResolutionRule { session =>
         ...
       }
       extensions.injectParser { (session, parser) =>
         ...
       }
     }
   }

   class YourExtensions extends SparkSessionExtensionsProvider {
     override def apply(extensions: SparkSessionExtensions): Unit = {
       extensions.injectResolutionRule { session =>
         ...
       }
       extensions.injectFunction(...)
     }
   }
 

Note that none of the injected builders should assume that the SparkSession is fully initialized and should not touch the session's internals (e.g. the SessionState).

Constructor Summary

Constructors

Constructor and Description
SparkSessionExtensions()

Method Summary

Modifier and Type	Method and Description
scala.collection.Seq<org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>>	buildPlanNormalizationRules(SparkSession session)
void	injectCheckRule(scala.Function1<SparkSession,scala.Function1<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan,scala.runtime.BoxedUnit>> builder) Inject an check analysis Rule builder into the SparkSession.
void	injectColumnar(scala.Function1<SparkSession,org.apache.spark.sql.execution.ColumnarRule> builder) Inject a rule that can override the columnar execution of an executor.
void	injectFunction(scala.Tuple3<org.apache.spark.sql.catalyst.FunctionIdentifier,org.apache.spark.sql.catalyst.expressions.ExpressionInfo,scala.Function1<scala.collection.Seq<org.apache.spark.sql.catalyst.expressions.Expression>,org.apache.spark.sql.catalyst.expressions.Expression>> functionDescription) Injects a custom function into the FunctionRegistry at runtime for all sessions.
void	injectOptimizerRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder) Inject an optimizer Rule builder into the SparkSession.
void	injectParser(scala.Function2<SparkSession,org.apache.spark.sql.catalyst.parser.ParserInterface,org.apache.spark.sql.catalyst.parser.ParserInterface> builder) Inject a custom parser into the SparkSession.
void	injectPlannerStrategy(scala.Function1<SparkSession,org.apache.spark.sql.execution.SparkStrategy> builder) Inject a planner Strategy builder into the SparkSession.
void	injectPlanNormalizationRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder) Inject a plan normalization Rule builder into the SparkSession.
void	injectPostHocResolutionRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder) Inject an analyzer Rule builder into the SparkSession.
void	injectPreCBORule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder) Inject an optimizer Rule builder that rewrites logical plans into the SparkSession.
void	injectQueryPostPlannerStrategyRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.execution.SparkPlan>> builder) Inject a rule that applied between plannerStrategy and queryStagePrepRules, so it can get the whole plan before injecting exchanges.
void	injectQueryStageOptimizerRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.execution.SparkPlan>> builder) Inject a rule that can override the query stage optimizer phase of adaptive query execution.
void	injectQueryStagePrepRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.execution.SparkPlan>> builder) Inject a rule that can override the query stage preparation phase of adaptive query execution.
void	injectResolutionRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder) Inject an analyzer resolution Rule builder into the SparkSession.
void	injectRuntimeOptimizerRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder) Inject a runtime Rule builder into the SparkSession.
void	injectTableFunction(scala.Tuple3<org.apache.spark.sql.catalyst.FunctionIdentifier,org.apache.spark.sql.catalyst.expressions.ExpressionInfo,scala.Function1<scala.collection.Seq<org.apache.spark.sql.catalyst.expressions.Expression>,org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> functionDescription) Injects a custom function into the TableFunctionRegistry at runtime for all sessions.

Methods inherited from class Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

SparkSessionExtensions

public SparkSessionExtensions()

Method Detail

buildPlanNormalizationRules

public scala.collection.Seq<org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> buildPlanNormalizationRules(SparkSession session)

injectCheckRule

public void injectCheckRule(scala.Function1<SparkSession,scala.Function1<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan,scala.runtime.BoxedUnit>> builder)

Inject an check analysis Rule builder into the SparkSession. The injected rules will be executed after the analysis phase. A check analysis rule is used to detect problems with a LogicalPlan and should throw an exception when a problem is found.

Parameters:

builder - (undocumented)

injectColumnar

public void injectColumnar(scala.Function1<SparkSession,org.apache.spark.sql.execution.ColumnarRule> builder)

Inject a rule that can override the columnar execution of an executor.

Parameters:

builder - (undocumented)

injectFunction

public void injectFunction(scala.Tuple3<org.apache.spark.sql.catalyst.FunctionIdentifier,org.apache.spark.sql.catalyst.expressions.ExpressionInfo,scala.Function1<scala.collection.Seq<org.apache.spark.sql.catalyst.expressions.Expression>,org.apache.spark.sql.catalyst.expressions.Expression>> functionDescription)

Injects a custom function into the FunctionRegistry at runtime for all sessions.

Parameters:

functionDescription - (undocumented)

injectOptimizerRule

public void injectOptimizerRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder)

Inject an optimizer Rule builder into the SparkSession. The injected rules will be executed during the operator optimization batch. An optimizer rule is used to improve the quality of an analyzed logical plan; these rules should never modify the result of the LogicalPlan.

Parameters:

builder - (undocumented)

injectParser

public void injectParser(scala.Function2<SparkSession,org.apache.spark.sql.catalyst.parser.ParserInterface,org.apache.spark.sql.catalyst.parser.ParserInterface> builder)

Inject a custom parser into the SparkSession. Note that the builder is passed a session and an initial parser. The latter allows for a user to create a partial parser and to delegate to the underlying parser for completeness. If a user injects more parsers, then the parsers are stacked on top of each other.

Parameters:

builder - (undocumented)

injectPlanNormalizationRule

public void injectPlanNormalizationRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder)

Inject a plan normalization Rule builder into the SparkSession. The injected rules will be executed just before query caching decisions are made. Such rules can be used to improve the cache hit rate by normalizing different plans to the same form. These rules should never modify the result of the LogicalPlan.

Parameters:

builder - (undocumented)

injectPlannerStrategy

public void injectPlannerStrategy(scala.Function1<SparkSession,org.apache.spark.sql.execution.SparkStrategy> builder)

Inject a planner Strategy builder into the SparkSession. The injected strategy will be used to convert a LogicalPlan into a executable SparkPlan.

Parameters:

builder - (undocumented)

injectPostHocResolutionRule

public void injectPostHocResolutionRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder)

Inject an analyzer Rule builder into the SparkSession. These analyzer rules will be executed after resolution.

Parameters:

builder - (undocumented)

injectPreCBORule

public void injectPreCBORule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder)

Inject an optimizer Rule builder that rewrites logical plans into the SparkSession. The injected rules will be executed once after the operator optimization batch and before any cost-based optimization rules that depend on stats.

Parameters:

builder - (undocumented)

injectQueryPostPlannerStrategyRule

public void injectQueryPostPlannerStrategyRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.execution.SparkPlan>> builder)

Inject a rule that applied between plannerStrategy and queryStagePrepRules, so it can get the whole plan before injecting exchanges. Note, these rules can only be applied within AQE.

Parameters:

builder - (undocumented)

injectQueryStageOptimizerRule

public void injectQueryStageOptimizerRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.execution.SparkPlan>> builder)

Inject a rule that can override the query stage optimizer phase of adaptive query execution.

Parameters:

builder - (undocumented)

injectQueryStagePrepRule

public void injectQueryStagePrepRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.execution.SparkPlan>> builder)

Inject a rule that can override the query stage preparation phase of adaptive query execution.

Parameters:

builder - (undocumented)

injectResolutionRule

public void injectResolutionRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder)

Inject an analyzer resolution Rule builder into the SparkSession. These analyzer rules will be executed as part of the resolution phase of analysis.

Parameters:

builder - (undocumented)

injectRuntimeOptimizerRule

public void injectRuntimeOptimizerRule(scala.Function1<SparkSession,org.apache.spark.sql.catalyst.rules.Rule<org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> builder)

Inject a runtime Rule builder into the SparkSession. The injected rules will be executed after built-in org.apache.spark.sql.execution.adaptive.AQEOptimizer rules are applied. A runtime optimizer rule is used to improve the quality of a logical plan during execution which can leverage accurate statistics from shuffle.

Note that, it does not work if adaptive query execution is disabled.

Parameters:

builder - (undocumented)

injectTableFunction

public void injectTableFunction(scala.Tuple3<org.apache.spark.sql.catalyst.FunctionIdentifier,org.apache.spark.sql.catalyst.expressions.ExpressionInfo,scala.Function1<scala.collection.Seq<org.apache.spark.sql.catalyst.expressions.Expression>,org.apache.spark.sql.catalyst.plans.logical.LogicalPlan>> functionDescription)

Injects a custom function into the TableFunctionRegistry at runtime for all sessions.

Parameters:

functionDescription - (undocumented)