Packages

•  package root

  Definition Classes

  root

•  package org

  Definition Classes

  root

•  package opalj

  OPAL is a Scala-based framework for the static analysis, manipulation and creation of Java bytecode.

  OPAL is a Scala-based framework for the static analysis, manipulation and creation of Java bytecode. OPAL is designed with performance, scalability and adaptability in mind.

  Its main components are:

  • a library (Common) which provides generally useful data-structures and algorithms for static analyses.
  • a framework for implementing lattice based static analyses (Static Analysis Infrastructure)
  • a framework for parsing Java bytecode (Bytecode Infrastructure) that can be used to create arbitrary representations.
  • a library to create a one-to-one in-memory representation of Java bytecode (Bytecode Disassembler).
  • a library to create a representation of Java bytecode that facilitates writing simple static analyses (Bytecode Representation - org.opalj.br).
  • a scalable, easily customizable framework for the abstract interpretation of Java bytecode (Abstract Interpretation Framework - org.opalj.ai).
  • a library to extract dependencies between code elements and to facilitate checking architecture definitions.
  • a library for the lightweight manipulation and creation of Java bytecode (Bytecode Assembler).

  General Design Decisions

  Thread Safety

  Unless explicitly noted, OPAL is thread safe. I.e., the classes defined by OPAL can be considered to be thread safe unless otherwise stated. (For example, it is possible to read and process class files concurrently without explicit synchronization on the client side.)

  No null Values

  Unless explicitly noted, OPAL does not null values I.e., fields that are accessible will never contain null values and methods will never return null. If a method accepts null as a value for a parameter or returns a null value it is always explicitly documented. In general, the behavior of methods that are passed null values is undefined unless explicitly documented.

  No Typecasts for Collections

  For efficiency reasons, OPAL sometimes uses mutable data-structures internally. After construction time, these data-structures are generally represented using their generic interfaces (e.g., scala.collection.{Set,Map}). However, a downcast (e.g., to add/remove elements) is always forbidden as it would effectively prevent thread-safety.

  Assertions

  OPAL makes heavy use of Scala's Assertion Facility to facilitate writing correct code. Hence, for production builds (after thorough testing(!)) it is highly recommend to build OPAL again using -Xdisable-assertions.

  Definition Classes

  org

•  package ai

  Implementation of an abstract interpretation (ai) framework – also referred to as OPAL.

  Implementation of an abstract interpretation (ai) framework – also referred to as OPAL.

  Please note that OPAL/the abstract interpreter just refers to the classes and traits defined in this package (ai). The classes and traits defined in the sub-packages (in particular in domain) are not considered to be part of the core of OPAL/the abstract interpreter.

  Definition Classes

  opalj

  Note

  This framework assumes that the analyzed bytecode is valid; i.e., the JVM's bytecode verifier would be able to verify the code. Furthermore, load-time errors (e.g., LinkageErrors) are – by default – completely ignored to facilitate the analysis of parts of a project. In general, if the presented bytecode is not valid, the result is undefined (i.e., OPAL may report meaningless results, crash or run indefinitely).

  See also

  org.opalj.ai.AI - Implements the abstract interpreter that processes a methods code and uses an analysis-specific domain to perform the abstract computations.

  org.opalj.ai.Domain - The core interface between the abstract interpretation framework and the abstract domain that is responsible for performing the abstract computations.

•  package domain

  This package contains definitions of common domains that can be used for the implementation of analyses.

  This package contains definitions of common domains that can be used for the implementation of analyses.

  Types of Domains

  In general, we distinguish two types of domains. First, domains that define a general interface (on top of the one defined by Domain), but do not directly provide an implementation. Hence, whenever you develop a new Domain you should consider implementing/using these domains to maximize reusability. Second, Domains that implement a specific interface (trait). In this case, we further distinguish between domains that provide a default implementation (per interface only one of these Domains can be used to create a final Domain) and those that can be stacked and basically refine the overall functionality.

  Examples

  • Domains That Define a General Interface
    • Origin defines two types which domains that provide information abou the origin of a value should consider to implement.
    • TheProject defines a standard mechanism how a domain can access the current project.
    • ...
  • Domains That Provide a Default Implementation
    • Origin defines the functionality to return a value's origin if the value supports that.
    • org.opalj.ai.domain.TheProject default implementation of the class hierarchy related methods using the project's class hierarchy.
    • org.opalj.ai.domain.DefaultHandlingOfMethodResults basically implements a Domain's methods related to return instructions an uncaught exceptions.
    • ...
  • Domains That Implement Stackable Functionality
    • org.opalj.ai.domain.RecordThrownExceptions records information about all uncaught exceptions by intercepting a Domain's respective methods. However, it does provide a default implementation. Hence, a typical pattern is:

  class MyDomain extends Domain with ...
      with DefaultHandlingOfMethodResults with RecordThrownExceptions

  Thread Safety

  Unless explicitly documented, a domain is never thread-safe. The general programming model is to use one Domain object per code block/method and therefore, thread-safety is not required for Domains that are used for the evaluation of methods. However domains that are used to adapt/transfer values should be thread safe (see org.opalj.ai.domain.ValuesCoordinatingDomain for further details).

  Definition Classes

  ai

•  package l1

  Commonly useful methods.

  Commonly useful methods.

  Definition Classes

  domain

•  trait IntegerRangeValues extends IntegerValuesDomain with IntegerRangeValuesFactory with ConcreteIntegerValues

  This domain represents integer values using ranges.

  This domain represents integer values using ranges.

  The cardinality of the range can be configured to satisfy different needs with regard to the desired precision (maxCardinalityOfIntegerRanges). Often, a very small cardinality (e.g., between 2 and 8) may be completely sufficient and a large cardinality does not add the overall precision significantly and just increases the analysis time.

  Constraint Propagation

  This domain facilitates and performs constraint propagation (e.g., intEstablishValue, intEstablishIsLessThan,...). Two integer (range) values (ir1,ir2) are reference equal (eq in Scala) iff both represent the same runtime value.

  In other words, the implementation ensures that two int values that are known to have the same value – even though the precise value may not be known – are represented using the same object. Furthermore, two int values that are not known to represent the same value at runtime are always represented using different objects. For example, consider the following sequence:

  • pcA+0/t1: iadd (Stack: 1 :: AnIntegerValue :: ...; Registers: <ignored>)
  • pcA+1/t2: dup (Stack: v(pcA/t1) :: ...; Registers: <ignored>)
  • pcA+2/t3: iflt true:+10 (Stack: v(pcA/t1) :: v(pcA/t1) :: ...; Registers: <ignored>)
  • pcA+3/t4: ... (Stack: v(pcA/t1) >= 0 :: ...; Registers: <ignored>)
  • pcA+XYZ...
  • pcA+12/t5: ... (Stack: v(pcA/t1) < 0 :: ...; Registers: <ignored>)

  Here, the test (iflt) of the topmost stack value against the constant 0 constraints the second topmost stack value. Both (abstract) values are guaranteed to represent the same value at runtime even though the concrete value may be unknown. In this case, the value was even created at the same point in time.

  In case of this domain the reference of the Domain(Integer)Value is used to identify those values that were created at the same point in time and hence, have the same properties.

  E.g., consider the following fictitious sequence:

  • iconst2 ...
    • Stack: EMPTY
    • Locals: EMPTY
  • dup ...
    • Stack: IntegerRangeValue(2,2)@123456;
    • Locals: EMPTY
  • istore_0 ...
    • Stack: IntegerRangeValue(2,2)@123456 <- IntegerRangeValue(2,2)@123456;
    • Locals: EMPTY
  • iconst2 ...
    • Stack: IntegerRangeValue(2,2)@123456;
    • Locals: 0=IntegerRangeValue(2,2)@123456, 1=EMPTY
  • istore_1 ...
    • Stack: IntegerRangeValue(2,2)@654321 <- IntegerRangeValue(2,2)@123456;
    • Locals: 0=IntegerRangeValue(2,2)@123456, 1=EMPTY
  • ...
    • Stack: IntegerRangeValue(2,2)@123456;
    • Locals: 0=IntegerRangeValue(2,2)@123456, 1=IntegerRangeValue(2,2)@654321

  Additionally, if the sequence would be part of a loop, the next iteration would create new IntegerRangeValues.

  Implementation Requirements

  Subclasses are required to create new instances of IntegerRangeValues and AnIntegerValue whenever a computation is performed that may affect the runtime value. If this property is not satisfied the implemented constraint propagation mechanism will produce unpredictable results as it may constrain unrelated values! This is true for concrete ranges as well as AnIntegerValues.

  Definition Classes

  l1

• AnIntegerValueLike
• ConcreteIntegerValue
• DomainIllegalValue
• DomainReferenceValue
• DomainReturnAddressValue
• DomainReturnAddressValues
• DomainTypedValue
• DomainValue
• IllegalValue
• IntegerLikeValue
• IntegerRangeLike
• RETValue
• ReferenceValue
• ReturnAddressValue
• ReturnAddressValues
• TypedValue
• Value

org.opalj.ai.domain.l1.IntegerRangeValues

IntegerRangeLike

Companion class IntegerRangeLike