com.ibm.wala.ipa.slicer

Class Slicer.SliceProblem

java.lang.Object

com.ibm.wala.ipa.slicer.Slicer.SliceProblem

All Implemented Interfaces:

PartiallyBalancedTabulationProblem<Statement,PDG,Object>, TabulationProblem<Statement,PDG,Object>

Enclosing class:

Slicer

public static class Slicer.SliceProblem
extends Object
implements PartiallyBalancedTabulationProblem<Statement,PDG,Object>

Tabulation problem representing slicing

Constructor Summary

Constructors

Constructor and Description
Slicer.SliceProblem(Collection<Statement> roots, ISDG sdg, boolean backward)

Method Summary

Methods

Modifier and Type	Method and Description
TabulationDomain<Object,Statement>	getDomain()
Statement	getFakeEntry(Statement node) If n is reached by a partially balanced parenthesis, what is the entry node we should use as the root of the PathEdge to n? Note that the result must in fact be an entry node of the procedure containing n.
IPartiallyBalancedFlowFunctions<Statement>	getFunctionMap()
IMergeFunction	getMergeFunction() Special case: if supportsMerge(), then the problem is not really IFDS anymore.
ISupergraph<Statement,PDG>	getSupergraph()
Collection<PathEdge<Statement>>	initialSeeds() Define the set of path edges to start propagation with.

Methods inherited from class java.lang.Object

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

Constructor Detail

Slicer.SliceProblem

public Slicer.SliceProblem(Collection<Statement> roots,
                   ISDG sdg,
                   boolean backward)

Method Detail

getDomain

public TabulationDomain<Object,Statement> getDomain()

Specified by:

getDomain in interface TabulationProblem<Statement,PDG,Object>

getFunctionMap

public IPartiallyBalancedFlowFunctions<Statement> getFunctionMap()

Specified by:

getFunctionMap in interface PartiallyBalancedTabulationProblem<Statement,PDG,Object>

Specified by:

getFunctionMap in interface TabulationProblem<Statement,PDG,Object>

getMergeFunction

public IMergeFunction getMergeFunction()

Description copied from interface: TabulationProblem

Special case: if supportsMerge(), then the problem is not really IFDS anymore. (TODO: rename it?). Instead, we perform a merge operation before propagating at every program point. This way, we can implement standard interprocedural dataflow and ESP-style property simulation, and various other things.

Specified by:

getMergeFunction in interface TabulationProblem<Statement,PDG,Object>

Returns:

the merge function, or null if !supportsMerge()

getSupergraph

public ISupergraph<Statement,PDG> getSupergraph()

Specified by:

getSupergraph in interface TabulationProblem<Statement,PDG,Object>

initialSeeds

public Collection<PathEdge<Statement>> initialSeeds()

Description copied from interface: TabulationProblem

Define the set of path edges to start propagation with.

Specified by:

initialSeeds in interface TabulationProblem<Statement,PDG,Object>

getFakeEntry

public Statement getFakeEntry(Statement node)

Description copied from interface: PartiallyBalancedTabulationProblem

If n is reached by a partially balanced parenthesis, what is the entry node we should use as the root of the PathEdge to n? Note that the result must in fact be an entry node of the procedure containing n.

Specified by:

getFakeEntry in interface PartiallyBalancedTabulationProblem<Statement,PDG,Object>