# Wednesday, May 19, 2004

This article will try to give an rough overview of the MbUnit vision of tests, and consequently it's architecture. It's contains some material of a previous CodeProject article.

Why MbUnit ?

Unit testing is a great tool for ensuring an application quality and frameworks like NUnit or csUnit have made it very simple to implement. However, as the number of tests begins to grow, the need for more functionalities begin to show up. The above frameworks are based on the Simple Test Pattern which is basically the sequence of SetUp, Test, TearDown actions. Although highly generic, this solution lets a lot of work to be done by the test writer. Sadely, there is no easy way to derive and include a new "fixture" type in those frameworks.

MbUnit is simply born from the fact that I wanted a new fixture and integrating it into existing frameworks was nearly impossible (I was also resting from a knee surgery at hospital with nothing else to do than coding).

Illustrating example

In order to make things clear, I will refer to an example while explaining how MbUnit works. Let me consider the Simple Test Pattern which is implemented by most test unit framework available. This is the classic way of writing unit test as described in the figure below. A TestFixture attribute tags the test class, one SetUp method, tests are done in the Test tagged method and clean up is performed in TearDown tagged method. This is illustrated in the left of the figure.

Attribute -> Run -> Invoker

The kernel of MbUnit is  composed of different components that work in a serial way. The first component is the fixture attribute

The fixture attribute is used to tag the classes that contain unit tests (TestFixtureAttribute is a fixture attribute). The new thing in MbUnit is that each fixture attribute contains the execution logic of the fixture which is returned at run-time under the form of a Run (IRun interface). In the case of the example, the TestFixtureAttribute is defined as a sequence of SetUp, Test and TearDown:

public class TestFixtureAttribute : TestFixturePatternAttribute 
{
     public override IRun GetRun()
     {
          SequenceRun runs = new SequenceRun();
            
          // setup
          OptionalMethodRun setup = new
                              OptionalMethodRun(typeof(SetUpAttribute),false);
          runs.Runs.Add( setup );
            
          //tests
          MethodRun test =new MethodRun(typeof(TestPatternAttribute),true,true);
          runs.Runs.Add(test);
            
          // tear down
          OptionalMethodRun tearDown = new
                           OptionalMethodRun(typeof(TearDownAttribute),false);
          runs.Runs.Add(tearDown);
            
          return runs;                        
     }
}

where

  • TestFixturePatternAttribute is the abstract base class for all new fixture attribute in MbUnit,
  • the GetRun method is called by the MbUnit core to know what is the execution path of the fixture. The fixture can use built-in basic attributes to build it's execution path.
  • An IRun instance can represent the call to a method, or to a sequence of methods, etc...
  • SequenceRun is a sequence of IRun's,
  • MethodRun is a IRun instance that wraps a call to a method tagged by a predefined attribute.
  • OptionalMethodRun is inherited from MethodRun and describes optional methods.

The IRun object will create an execution tree  by exploring the tagged type. Each node of the tree contains a RunInvoker (IRunInvoker interface). The RunInvoker is in charge for calling the method, garding the execptions, loading data, etc... On our sample fixture, there are two tests that the Run will extract:

When the tree is built, we just extract all the possible path from the root node to the leaves to extract the different possible tests. Each of these path is called a Pipe (RunPipe class).

In the GUI, the RunPipe instances are attached to the TreeNode nodes so you can easily select and execute separately the tests. This ensures that the test execution are isolated.

This architecture brings a lot of flexibility (and complexity) on the kind of fixtures that can be defined. Any user can define it's own fixture and use MbUnit to execute it.

posted on Wednesday, May 19, 2004 11:22:00 AM (Pacific Daylight Time, UTC-07:00)  #    Comments [4]

As most of you should know by now, PageRank is the ranking system used by Google to estimate the importance of a page (you can see it in the Google toolbar). Of course, since the basic idea of the algorithm was published they may have been some significant modifications. In this blog, I'll how we can use QuickGraph to compute the PageRank of a graph...

PageRank

The idea behind PageRank is simple and intuitive: pages that are important are referenced by other important pages, page importance is distributed to out-edges. There is an important literature on the web that explains PageRank:

The PageRank is computed by using the following iterative formula:

PR(A) = (1-d) + d (PR(T1)/C(T1) + ... + PR(Tn)/C(Tn)) 

where PR is the PageRank, d is a damping factor usually set to 0.85, C(v) is the number of out edges of v.

PageRank can also be expressed in terms of matrix algebra where it is shown that it is equivalent to finding the eigen values of a sparse matrix (see http://citeseer.ist.psu.edu/kamvar03extrapolation.html). And in fact, the above formula is equivalent to the Power Method (a slow method for finding eigen values).

Where's my LAPACK ?

Until C# (and .Net) has a real and free wrapper around LAPACK, we cannot attack PageRank using matrix algebra. As mentionned above, the formula is equivalent to the Power Method, which is a slow (potentially very slow) method for computing eigen values (convergence rate is |la_1 / la_2| where la_1 is the largest eigen value, la_2 is the second largest eigen value). There are other faster methods (like model reduction) that we could use to speed up things if we had LAPACK (I'm throwing a bottle in the .Net sea here).

Implementation in QuickGraph

Since we have no matrix algebra, the implementation is very basic and unefficient (I'm almost ashamed). This is almost a disclaimer: do not use this algorithm for big graphs, it is potentially slow. The main loop looks like this:

// temporay rank dictionary
VertexDoubleDictionary tempRanks = new VertexDoubleDictionary();
// create filtered graph that removes dangling links
FilteredBidirectionalGraph fg = new FilteredBidirectionalGraph(
    this.VisitedGraph,
    Preds.KeepAllEdges(),
    new InDictionaryVertexPredicate(this.ranks)
    ); 

int iter = 0;
double error = 0;
do
{
    // compute page ranks
    error = 0;
    foreach(DictionaryEntry de in this.Ranks) 
    {
        IVertex v = (IVertex)de.Key;
        double rank = (double)de.Value;

        double r = 0;
        foreach(IEdge e in fg.InEdges(v))
        {
            r += this.ranks[e.Source] / fg.OutDegree(e.Source);
        }
        // add sourceRank and store
        double newRank = (1-this.damping) + this.damping * r;
        tempRanks[v] = newRank;
        // compute deviation
        error += Math.Abs(rank - newRank);
    } 
    // swap ranks
    VertexDoubleDictionary temp = ranks;
    ranks = tempRanks;
    tempRanks = temp; 
    iter++;
// iterate until convergence, or max iteration reached
}while( error > this.tolerance && iter < this.maxIterations);

where ranks is the PR method, damping is the d factor. Note that because we use enumerators, we cannot modify the ranks as we iterate the dictionary, otherwize the enumerator would be invalidated, therefore we use 2 dictionaries and swap between them as we go along.

The results

As usual, we use GraphvizAlgorithm to output the results. Here are some graph sample with each corresponding page rank:

posted on Wednesday, May 19, 2004 9:22:00 AM (Pacific Daylight Time, UTC-07:00)  #    Comments [5]
# Tuesday, May 18, 2004

This is the first episode of an article serie on graphs and databases. The menu of today will be

  1. how to extract the structure of a database,
  2. create a graph representation using QuickGraph
  3. draw it using NGraphviz

To make things user friendly, I will use the PropertyGrid to set up things.

Extracting the database schema

At first sight this seemed to be a tedious and boring task but hopefully a light poped up on the back of my head saying "you have already seen that in CodeSmith". In fact CodeSmith comes with an assembly, SchemaExplorer, whose purpose is to extract database schema. Even better, the main class, DatabaseSchema, comes with a custom type editor (DatabaseSchemaTypeEditor) so that integration in the PropertyGrid is straightforward. 

public class DataGraphProperties
{
    private DatabaseSchema schema = null;
    [Category("Data")]
    [TypeConverter(typeof(DatabaseSchemaTypeConverter))]
    public DatabaseSchema Schema
    {
        get
        {
            return this.schema;
        }
        set
        {
            this.schema = value;
        }
    }
}

In the PropertyGrid, the Schema property will let the user to select a data source.

Database and graphs

It is straightforward to see that a database is a graph where the tables are the vertex and the foreign keys are the edges. The DatabaseSchema class contains the collection of tables (TableSchema instances), each table containing a collection of foreign keys (TableKeySchema instance). So we have all we need to populate the graph.

Custom Vertex and Edges

The first step for creating a representation of the database as a QuickGraph graph is to create the custom vertex (that implements IVertex) and edge classes (that implement IEdge). This task is straigtforward by using two default classes, Vertex and Edge, available in the QuickGraph assembly. This is illustrated for TableSchemaVertex:

public class TableSchemaVertex : Vertex
{
    private TableSchema table = null;
    public TableSchemaVertex(int id)
    :base(id)
    {}

    public TableSchema Table
    {
        get
        {
            if (this.table==null)
                throw new InvalidOperationException("table not initialized");
            return this.table;
        }
        set
        {
            this.table = value;
        }
    }
}

The TableSchemaVertex instance are to be created by a vertex provider:

public class TableSchemaVertexProvider : TypedVertexProvider
{
    public TableSchemaVertexProvider()
    :base(typeof(TableSchemaVertex))
    {}
}

The same thing is done again for the edges, which is called TableKeySchemaEdge.

Custom Graph

The custom graph is generated using the CodeSmith template AdjacencyGraph.cst. The class is called DatabaseSchemaGraph.

Populating the graph

Once the data structure is ready, populating the graph with the tables and the keys is straightforward:

DatabaseSchema schema = ...;
DatabaseSchemaGraph graph = ...;
// add tables;
foreach(TableSchema table in schema.Tables)
{
    graph.AddVertex(table);
}
// foreach table, add all relations (out-edges)
foreach(TableSchema table in schema.Tables)
{
    foreach(TableKeySchema key in table.ForeignKeys)
    {
        graph.AddEdge(key);
    }
}

That's it :)

Let's do some drawing

Now that we have a graph of the database, the Graphviz "machinery" can be used to output a number of different drawings (refer to this post for a detailled tutorial on using Graphviz). Bundled that with the PropertyGrid and we get a nice and simple database grapher. I have applied DbGrapher on the database that MbUnit uses to store test results:

Next episode

In the next episode, we will see how to improve the (poor) quality of the drawing and how to detect cascade cycles (on delete cycles etc...).

posted on Tuesday, May 18, 2004 9:37:00 PM (Pacific Daylight Time, UTC-07:00)  #    Comments [2]
# Monday, May 17, 2004

The Abstract Test Pattern (ATP)

I have received a few comments on my blog entry on Composite Unit Testing (CUT) arguying that this was the Abstract Test Pattern . Here's a snapshot of the definition from the definition form http://c2.com/cgi/:

A Testing Pattern describing a way to reuse test cases for multiple implementations of an Interface.
Problem
How to write a Test Suite against an Interface (or Abstract Class) that can be used to test all implementations of the interface.

Solution

  • Write an AbstractTest  for every Interface and Abstract Class). The AbstractTest should have an abstract FactoryMethod that creates an object with the type of the Interface.
  • Write a ConcreteTest for every implementation of the Interface. The ConcreteTest? should be a descendant of the AbstractTest and override the FactoryMethod to construct an instance of the implementation class.

Functional Compliance

Eric George's article gives a more detailled description of the pattern and describes it as functional compliance. It is easy enough for the compiler to tell whether a class is syntactically compliant with an interface. It applies a check to see if all required methods have been implemented with the correct signatures (syntaxic compliance), but the compiler cannot check functional compliance of a class with its interface. Here's the formal definition given by Eric George:

Functional Compliance is a module's compliance with some documented or published functional specification. The specification can be purely documentational, or it can be partially enforced through Interfaces or Abstract Classes. Interfaces and Abstract Classes along with their associated documentation represent a contract between the implementation code and the client (or user) code. It is this contract that needs to be fully tested. The Liskov Substitution Principle (LSP) tells us that all modules that honor a contract (usually by implementing an interface), should behave the same from the perspective of the client code. A module's functional compliance is really the degree to which it obey's the LSP.

So what about Composite Unit Testing ?

The remarks from the readers were right. Composite Unit Testing is

  • an enhanced form of the Abstract Test Pattern,
  • is a tool to test functional compliance

There is, however, a major difference between ATP and CUT: separation of the test code and the factory methods. In AUT, you create a ConcreteTest that inherits AbstractTest and implements a factory method, so the code that generates the tested entity is "hard-coded" into concrete test. In CUT, the framework takes care of retreiving and feeding you AbstractTest using user-specified factories (you can easily have multiple factories):

// AUT
// abstract method
public abstract class AbstractEnumerableTest
{
    public IEnumerable Create();
    public void GetEnumeratorTest()
    {
        IEnumerable en = this.Create();
        ...
    }
}

// concrete implementation
[TestFixture]
public class ArrayListEnumerableTest
{
    public override IEnumerable Create()
    { return new ArrayList();}
}

The same test as above, using CUT:

// the fixture
public class EnumerableFixture
{
    public void GetEnumeratorTest(IEnumerable en)
    {
        ...
    }
}

// the factories
public class ArrayListFactory
{
    public ArrayList Emtpy
    { get{ return new ArrayList();}}
}

// link the fixture with the factories
[CompositeFixture(typeof(EnumerableFixture), typeof(IEnumerable))]
[ProviderFactory(typeof(ArraListFactory),typeof(IEnumerable))]
public class EnumerableTest
{}
posted on Monday, May 17, 2004 7:42:00 AM (Pacific Daylight Time, UTC-07:00)  #    Comments [3]
# Friday, May 14, 2004

The following article on CodeProject talks about Scarified Treemaps, an interresting tree visualization. I wonder what it would look like in MbUnit...

Demo application - treemaps.png

posted on Friday, May 14, 2004 2:17:00 PM (Pacific Daylight Time, UTC-07:00)  #    Comments [0]
# Thursday, May 13, 2004

Here's a preview of a new fixture that will be available in the next release of MbUnit. The fixtures loads XML data from files and feeds it to the different methods, hence it's Data Driven Unit Testing. I'll make a real article on this when the fixture stabilizes but here's a first example.

[DataFixture]
[XmlDataProvider("../../sample.xml","//User")]
[XmlDataProvider("../../sample.xml","DataFixture/Customers")]
public class DataDrivenTests
{
    [ForEachTest("//User")]
    public void ForEachTest(XmlNode node)
    {
        Assert.IsNotNull(node);
        Assert.AreEqual("User",node.Name);
        Console.WriteLine(node.OuterXml);
    }

    [ForEachTest("//User",DataType = typeof(User))]
    public void ForEachTestWithSerialization(User user)
    {
        Assert.IsNotNull(user);
        Console.WriteLine(user.ToString());
    }
}

The file sample.xml looks like this:

<DataFixture>
  <Employees>
    <User Name="Mickey" LastName="Mouse" />
  </Employees>
  <Customers>
    <User Name="Jonathan" LastName="de Halleux" />
    <User Name="Voldo" LastName="Unkown" />
  </Customers>
</DataFixture>

and the User class like this:

[XmlRoot("User")]
public class User
{
    private string name;
    private string lastName;
    public User()
    {}
    [XmlAttribute("Name")]
    public String Name
    {
    get{ return this.name;}
    set{ this.name = value;}
    }
    [XmlAttribute("LastName")]
    public String LastName
    {
    get{ return this.lastName;}
    set{ this.lastName = value;}
    }
}

How does it work ?

  • The XmlDataProvider attributes are used to specify an XML filename that contains the data (of course, you can put several of those).
  • Then a first XPath expression is applied to the data.
  • Each selected node is then feeded to the ForEachTest which applies a second XPath expression on the node. This allows a fine grained selection of the data.
  • You can also specify the desired output data type. XmlSerializer is then used to convert XmlNode into the desired object.

Screenshot

Here's a screenshot that shows have MbUnit loads the XML and creates the different test case.

posted on Thursday, May 13, 2004 10:28:00 PM (Pacific Daylight Time, UTC-07:00)  #    Comments [13]

This is a copy of an upcoming CodeProject article (number 36)

Introduction

This article presents a new way of creating unit tests. Rather than creating a fixture for each class, we split the testing effort by class functionality. Taking advantage of interface composition, we use split the unit tests for each interface and we feed those fixtures using factories. This is why I call this technique: Composite Unit Testing.

There are several advantages of using this approach:

  1. Expressing requirements: interfaces are a natural place for expressing requirements, which later on translate into unit tests,
  2. Test reusability: once you have design a test suite for an interface, you can apply it to any class that implements this interface,
  3. Test Driven Developement: this approach fits nicely and naturally into the TDD paradigm since execution path is interface -> interface fixture -> implementation(s).
  4. Separation of tests and tested instance generation: the code that generates the tested instances is located in factories that can be reused for each fixtures.

In the rest of the article, I will illustrate this technique on ArrayList and Hashtable.

Test Case

Let us consider illustrate the process with two classes of the System.Collections namespace: ArrayList and Hashtable.

The two classes belong to different families of containers: ArrayList is a sequential container, while Hashtable is an associative container. However, as the interface diagram below shows, they share a lot of  functionalities (enumeration, cloneable, serialization, etc...). These functionalities are usually represented by interface composition: ICloneable, ISerializable, etc... The interface define functionalities and requirements on those functionalities.

ArrayList and Hashtable

If you take the usual unit testing methodology, you will need to write two (huge) fixture to test the two classes. Since they share functionality, you will end up duplicating testing code, maintenance problem will increase, etc...

Composite unit testing provides a flexbile solution to those problems. In the following, I will illustrate how it is implemented in MbUnit.

Composite Unit Testing Methodology

As mentionned in the introduction, composite unit testing fits naturally in the TDD idea. The principal steps of the process are:

  1. create the interface and express requirements,
  2. create the interface fixture and translate the requirements into unit tests,
  3. implement the interface,
  4. create a class factory that provides instances of the interface,
  5. link fixture to factories and run...

Step 1: Create the interface

This is where you define the functionalities and the requirements. If the documentation is clear enough, it should translate naturally into unit tests. (In this example, the job is already done).

Step 2: Create the interface fixture (for IEnumerable)

MbUnit defines a new custom attribute TypeFixture that is used to create fixture for types (classes, structs or interface). TypeFixture constructor take the type that is tested as argument. Let us start with the fixture of IEnumerable:

EnumerableTest

using System;
using System.Collections;
using MbUnit.Core.Framework;
using MbUnit.Framework;
[TypeFixture(typeof(IEnumerable))]
public class EnumerableTest
{}

The test case in EnumerableTest will receive an instance of the tested type (IEnumerable here) as argument. Therefore, the correct signature of those methods is as follows:

[TypeFixture(typeof(IEnumerable))]
public class EnumerableTest
{
    [Test]
    public void EmptyTest(IEnumerable en)
    {...}
}

The argument is the only difference with the "classic" unit test. You can use test decorators like ExpectedException, Ignore, etc... as usual. IEnumerable defines one method, GetEnumerator. The only requirement is that the IEnumerator instance is not a null reference:

[TypeFixture(typeof(IEnumerable))]
public class EnumerableTest
{
    [Test]
    public void GetEnumeratorNotNull(IEnumerable en)
    {
        Assert.IsNotNull(en.GetEnumerator());
    }
}

That's pretty short but there is nothing else to test. If you want to test the enumeration, you need to write another fixture for IEnumerator. By defining fixtures for each interface you quickly increase the coverage of the tested code.

Composition of tests

Step 4: Create the factories

(We have skipped step 3, the implementation step)

A factory is simply a class that defines public properties or method (with no arguments) that return an object to be tested. You can use factories to provide different flavor of the same class: an empty ArrayList, randomly filled, ordered filled, etc... For example, a possible factory for ArrayList is:

public class ArrayListFactory
{
    public ArrayList Empty
    {
        get
        {
            return new ArrayList();
        }
    } 
    public ArrayList RandomFilled()
    {
        ArrayList list = new ArrayList();
        Random rnd = ...;
        for(int i=0;i<15;++i) 
            list.Add(rnd.Next());
        return list;
    } 
}

Note that a factory does not need any particular attributes. Similarly we can define HashtableFactory.

Step 5: Linking the fixtures to the factories

Linking the factories to the fixtures is simply done by using another custom attribute: ProviderFactory.

[TypeFixture(typeof(IEnumerable))]
[ProviderFactory(typeof(ArrayListFactory),typeof(IEnumerable))]
[ProviderFactory(typeof(HashtableFactory),typeof(IEnumerable))]
public class EnumerableTest
{...}

ProviderFactory takes the type of the factory, and the tested type as argument. The framework will take care of exploring by reflection the factories, select the suitable properties and feed the fixtures with created test instances.

Factories

Step 6: Running the tests

The full source of the example is available in the demo project. You need to create a new C# assembly project and add the reference to MbUnit.Core.dll and MbUnit.Framework.dll, which you can download from the MbUnit web site: http://mbunit.tigris.org. /

Launch MbUnit.GUI and load the assembly (right click -> Assemblies -> Add Assemblies...). Here are some screenshots of the application:

MbUnit GUI

MbUnit report

Conclusion

This article has presented composite unit testing, a new strategy for designing and implementing unit testing. Awaiting comments :)

 

posted on Thursday, May 13, 2004 11:52:00 AM (Pacific Daylight Time, UTC-07:00)  #    Comments [16]
# Wednesday, May 12, 2004

Just finished a new CodeSmith template for generating a MockObject out of an interface. The template is pretty basic: it loads the type, create the methods/properties of the different interfaces. You can also specify a number of expected values and the template will generate the SetXXX methods.

Located in the MbUnit CVS (Templates folder).

posted on Wednesday, May 12, 2004 5:05:00 PM (Pacific Daylight Time, UTC-07:00)  #    Comments [0]

Here's a sample C# application that implements structured numbers. (Practial use of this number is unknown :) )

The maths

The StrutucredNumbers sample contains an implementation of the binary tree operation developped by V. Blondel in Structured Numbers, Properties of a hierarchy of operations on binary trees, Acta Informatica, 35, 1-15, 1998.

We introduce a hierarchy of operations on (finite and infinite) binary trees. The operations are obtained by successive repetition of one initial operation. The ¯rst three operations are generalizations of the operations of addition, multiplication and exponentiation for positive integers.

In the paper, the author defines countably many internal operations on binary trees. The first operation, which he denote by .1. , is obtained by forming the binary tree whose left and right subtrees are equal to the operands. This operation is not associative. The second operation .2. is defined as follows: From the binary trees a and b we construct the binary tree a.2.b by repeating the operation .1. on the tree a with the structure dictated by b. In the same way, we define an operation .3. by repeating .2. , an operation .4. by repeating .3. , etc. We eventually obtain countably many internal operations ( .k. for k>= 1) with the definition

a .k. b = a^(k-1).a^(k-1).a^(k-1)...a^(k-1).a,

where there are b factors.

The code

The code is mainly composed of the BtNode that implements the algebra defined above and some valuators for the tree. At last, NGraphviz is used to render to trees :)

The results

Suppose that we have

x = ..
y = ..
, then
  • x:
  • x.1.y:
  • x.2.y:
  • x.3.y:
  • x.4.y:
posted on Wednesday, May 12, 2004 12:16:00 PM (Pacific Daylight Time, UTC-07:00)  #    Comments [0]
# Tuesday, May 11, 2004

In this "episode", we are going to build an application that generates and renders the exection graph of a method using QuickGraph and Lutz Reoder's IlReader. The execution graph is directed graph where each vertex is an IL instruction and each edge represent the transition between two instructions, it could be a jump or a simple move to the next instruction. The graph represents the different path that the application can take into your method. If you have access to the content of a method, you can potentially build.

In a future blog, I will use the execution graph to make "smart" test fixture generator by applying basic graph algorithms on the execution graph. In the following of the blog, I assume you have some basic knowledge of IL, Reflection and Reflection.Emit.

Step 1: creating the graph structures

We begin by creating a specialized type of vertex InstructionVertex that will hold a System.Reflection.Emit.Instruction instance:

using System.Reflection.Emit;
public class InstructionVertex : QuickGraph.Vertex
{
    private Reflector.Disassembler.Instruction instruction=null;
    public InstructionVertex(int id):base(id)
    {}
    public Reflector.Disassembler.Instruction Instruction
    {
        get
        {
            if (this.instruction==null)
                throw new InvalidOperationException();
            return this.instruction;
        }
        set
        {
            this.instruction = value;
        }
    }
    public override string ToString()
    {...}
}

Note that the ToString method uses the code Example.cs in the IlReader source to render an Instruction to string.

To generate a strongly-typed BidirecitonalGraph, that we call InstructionGraph, we use the CodeSmith template called AdjacencyGraph.cst located in the Templates directory.

Step 2: Loading the IL instructions of the method

The building of the graph is done by the IlGraphBuilder class. This class takes the type of the method as an argument. It contains one public method BuildGraph that takes a MethodInfo instance as argument as outputs a InstructionGraph instance containing the execution graph.

The code to extract the MethodBody from the method is almost entirely copy pasted from the IlReader sample:

public class IlGraphBuilder
{
    private Type visitedType;
    private ModuleReader reader;
    public IlGraphBuilder(Type visitedType)
    {
        this.visitedType = visitedType;
        this.reader = new ModuleReader(this.visitedType.Module, new AssemblyProvider());
    }
    public InstructionGraph BuildGraph(MethodInfo mi)
    { 
        // get method body using IlReaderr
        MethodBody methodBody = reader.GetMethodBody(mi);
        ...
    }

    private sealed class AssemblyProvider : IAssemblyProvider
    {
        public Assembly Load(string assemblyName)
        {
            return Assembly.Load(assemblyName); 
        }
        public Assembly[] GetAssemblies()
        {
            throw new NotImplementedException(); 
        }
    }
}

Step 3: Building the graph

The MethodBody instance contains the list of Instruction instances and the list of exception handlers. The building of the graph is done in 3 steps:

  1. iterate the Instruction list and create the vertices in the graph. We also build a dictionary that associates the instruction offset to the corresponding vertex,
    // new field in the class
    Hashtable instructionVertices;
    ...
    foreach(Instruction i in methodBody.GetInstructions())
    {
        // avoid certain instructions
        if (i.Code.FlowControl == FlowControl.Phi || i.Code.FlowControl == FlowControl.Meta)
            continue;
        // add vertex
        InstructionVertex iv = g.AddVertex();
        iv.Instruction = i;
        // store in hashtable
        this.instructionVertices.Add(i.Offset,iv);
    }
    
  2. iterate again over the instructions and add the edges that represent the transitions. This is the tricky part, I use a recursive exploration of the instructions, (this part of the code is a bit heavy for the blog)
  3. iterate over the exception handlers to link the different sections: create the link to catch,finally handlers, etc...

Step 4: Drawing the graph

Drawing the graph is straight-foward using the GraphvizAlgorithm class:

GraphvizAlgorithm gv = new GraphvizAlgorithm(g); 
gv.Write(mi.Name);

Analysing some basic flow contructions:

As an application, I going to show the graph of some basic instruction flow like for, while, if, foreach, etc...

public void HelloWorld()
{
    Console.WriteLine("Hello World");
}

public void IfAlone(bool value)
{
    if (value)
        Console.Write("value is true");
}

public void IfThenElse(bool value)
{
    if (value)
        Console.Write("value is true");
    else
        Console.Write("value is false");
}

public void For()
{
    for(int i = 0;i!=10;++i)
    {
        Console.Write(i);
    }
}

public void While()
{
    int i = 0;
    while(i!=10)
    {
        i++;
    }
}

public void TryCatchFinally()
{
    try
    {
        Console.WriteLine("try");
    }
    catch(Exception)
    {
        Console.WriteLine("catch"); 
    }
    finally
    {
        Console.WriteLine("finally");
    }
}

public void TryMultiCatchFinally()
{
    try
    {
        Console.WriteLine("try");
    }
    catch(ArgumentException)
    {
        Console.WriteLine("catch(arg)"); 
    }
    catch(Exception)
    {
        Console.WriteLine("catch"); 
    }
    finally
    {
        Console.WriteLine("finally");
    }
}

public void ForEach(ICollection col)
{
    foreach(Object o in col)
    {
        Console.WriteLine(o);
    }
}

public void ForEachContinue(int[] col)
{
    foreach(int o in col)
    {
        Console.WriteLine(o);
        if (o == 0)
            continue;
    }
}

public void ForEachBreak(int[] col)
{
    foreach(int o in col)
    {
        Console.WriteLine(o);
        if (o == 0)
            break;
    }
}

public void SwitchIt(int value)
{
    switch(value)
    {
    case 0:
        Console.Write("0");
        break;
    case 1:
        Console.Write("1");
        break;
    default:
        Console.Write("default");
        break;
    }
}

posted on Tuesday, May 11, 2004 10:51:00 AM (Pacific Daylight Time, UTC-07:00)  #    Comments [9]