Extracting PDB signature from DLL file - c#

I have to extract PDB signature from both .pdb and .dll file.
That's the code I use to extract it from .pdb file. Unfortunately I haven't found similiar way of extracting it from a DLL.
public static string GetPdbSignature(string pdbFilePath)
{
using (var pdbFileStream = File.OpenRead(pdbFilePath))
{
var metadataProvider = MetadataReaderProvider.FromPortablePdbStream(pdbFileStream);
var metadataReader = metadataProvider.GetMetadataReader();
var id = new BlobContentId(metadataReader.DebugMetadataHeader.Id);
return $"{id.Guid.ToString("N")}ffffff";
}
}
I found out that a PeNet nuget package can be used to perform the extraction, yet I'd prefer to achieve that without installing external dependancies.
Also, I managed to find the desired data using a dotPeek (screen), but as I need to resolve the problem programatically it doesn't solve my issue either.
I'd apreciate any hint how to aproach that problem. Either by using some built in dotnet mechanism or by some smart low level byte extraction.

I managed to find an official Microsoft's package - Microsoft.Diagnostics.Tracing.TraceEvent that contains PEFile class allowing to extract the exact data I need.
public static string GetDllSignature(string dllFilePath)
{
var peFile = new PEFile.PEFile(dllFilePath);
peFile.GetPdbSignature(out string pdbName, out Guid pdbGuid, out int pdbAge);
return $"{pdbGuid.ToString("N")}ffffff";
}
UPDATE:
Actually there also is a PEReader class in System.Reflection.PortableExecutable namespace that makes the reading possible using only the system libraries. However it requires some knowledge of the portable executable format, as the PEReader does not provide an explicit, user-friendly method for extracting the signature, instead it just allows getting all kind of data that the PE file contains.
public static string GetDllSignatureV2(string dllFilePath)
{
using (var pdbStream = File.OpenRead(pdbPath))
using (var peReader = new PEReader(pdbStream))
{
var debugDirectory = peReader.ReadDebugDirectory().First(entry => entry.Type == DebugDirectoryEntryType.CodeView);
var codeViewData = peReader.ReadCodeViewDebugDirectoryData(debugDirectory);
return $"{codeViewData.Guid.ToString("N").Replace("-", string.Empty)}FFFFFFFF".ToUpper();
}
}

Related

How to programmatically create a class library DLL using reflection?

Suppose my code possesses the knowledge about the metadata of a
nonexistent class library "mytest.dll", such as the types in this library, the functions of the types, the parameters and return types of the functions, etc.
How does my code manufacture this DLL using techniques such as reflection?
I know my code can generate the "mytest.cs" text file, then execute the compiler to produce the DLL, then delete the "mytest.cs" file. Just want to know if there are "more advanced" or "cooler" ways to do it.
Thanks.
There are 4 main steps in the process to compile and execute dynamic .net scripts from your application, even really complex scenarios can be simplified in this way:
Generate the code
Compile the script
Load the assembly
Execute the code
Lets generate a simple Hello Generated C# World App right now!:
Create a method that will generate an assembly that has 1 class called HelloWorldApp, this class has 1 method called GenerateMessage it will have X input parameters that will be integers, it will return a CSV string of the arguments that were passed in to it.
This solution requires the following package to be installed:
PM> Install-Package 'Microsoft.CodeAnalysis.CSharp.Scripting'
And will require the following using statements:
using Microsoft.CodeAnalysis;
using Microsoft.CodeAnalysis.CSharp;
using Microsoft.CodeAnalysis.Emit;
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Reflection;
using System.Text;
Orchestration
The following method encapsulates the above steps:
private static void GenerateAndExecuteApp(int numberOfParameters)
{
string nameSpace = "Dynamic.Example";
string className = "HelloWorldApp";
string methodName = "GenerateMessage";
// 1. Generate the code
string script = BuildScript(nameSpace, className, methodName, numberOfParameters);
// 2. Compile the script
// 3. Load the Assembly
Assembly dynamicAssembly = CompileScript(script);
// 4. Execute the code
int[] arguments = Enumerable.Range(1, numberOfParameters).ToArray();
string message = ExecuteScript(dynamicAssembly, nameSpace, className, methodName, arguments);
Console.Out.WriteLine(message);
}
Generate the code
You say you already have item 1 sorted out, you can use StringBuilder, T4 templates or other mechanisms to generate the code files.
generating the code itself is its own question if you need help with that.
However, for our demo app, the following would work:
private static string BuildScript(string nameSpace, string className, string methodName, int numberOfParameters)
{
StringBuilder code = new StringBuilder();
code.AppendLine("using System;");
code.AppendLine("using System.Linq;");
code.AppendLine();
code.AppendLine($"namespace {nameSpace}");
code.AppendLine("{");
code.AppendLine($" public class {className}");
code.AppendLine(" {");
var parameterNames = Enumerable.Range(0, numberOfParameters).Select(x => $"p{x}").ToList();
code.Append($" public string {methodName}(");
code.Append(String.Join(",", parameterNames.Select(x => $"int {x}")));
code.AppendLine(")");
code.AppendLine(" {");
code.Append(" return $\"");
code.Append(String.Join(",", parameterNames.Select(x => $"{x}={{{x}}}")));
code.AppendLine("\";");
code.AppendLine(" }");
code.AppendLine(" }");
code.AppendLine("}");
return code.ToString();
}
For an input value of 3, the following code is generated:
using System;
using System.Linq;
namespace Dynamic.Example
{
public class HelloWorldApp
{
public string GenerateMessage(int p0,int p1,int p2)
{
return $"p0={p0},p1={p1},p2={p2}";
}
}
}
Compile the script (and Load it)
These are two discrete steps, however it is easiest to code them together in the same method, for this example we will ignore the generated dll and load the assembly directly into memory, that is generally the more likely use case for this type of scripting scenario anyway.
The hardest element of this is usually the referencing of the relevant dlls. There are a number of ways to achieve this, including loading all the dlls that are in the current executing context, I find a simple way to do this is to access the Assembly reference from the Type reference for the types we want to use inside the dynamic script:
List<string> dlls = new List<string> {
typeof(object).Assembly.Location,
typeof(Enumerable).Assembly.Location
};
Cut a long story short, this method compiles and loads the assembly into memory. It includes some crude compilation error handling, just to demonstrate how to do it:
private static Assembly CompileScript(string script)
{
SyntaxTree syntaxTree = CSharpSyntaxTree.ParseText(script);
// use "mytest.dll" if you want, random works well enough
string assemblyName = System.IO.Path.GetRandomFileName();
List<string> dlls = new List<string> {
typeof(object).Assembly.Location,
typeof(Enumerable).Assembly.Location
};
MetadataReference[] references = dlls.Distinct().Select(x => MetadataReference.CreateFromFile(x)).ToArray();
CSharpCompilation compilation = CSharpCompilation.Create(
assemblyName,
syntaxTrees: new[] { syntaxTree },
references: references,
options: new CSharpCompilationOptions(OutputKind.DynamicallyLinkedLibrary));
// Now we actually compile the script, this includes some very crude error handling, just to show you can
using (var ms = new MemoryStream())
{
EmitResult result = compilation.Emit(ms);
if (!result.Success)
{
IEnumerable<Diagnostic> failures = result.Diagnostics.Where(diagnostic =>
diagnostic.IsWarningAsError ||
diagnostic.Severity == DiagnosticSeverity.Error);
List<string> errors = new List<string>();
foreach (Diagnostic diagnostic in failures)
{
//errors.AddDistinct(String.Format("{0} : {1}", diagnostic.Id, diagnostic.Location, diagnostic.GetMessage()));
errors.Add(diagnostic.ToString());
}
throw new ApplicationException("Compilation Errors: " + String.Join(Environment.NewLine, errors));
}
else
{
ms.Seek(0, SeekOrigin.Begin);
return Assembly.Load(ms.ToArray());
}
}
}
Execute the code
Finally, we can use reflection to instantiate an instance of the new app and then we can obtain a reference to the method and it. The name of the parameters is irrelevant, as long
we pass them through in the correct order:
for this demo the order is sort of irrelevant to, given they are all the same type ;)
private static string ExecuteScript(Assembly assembly, string nameSpace, string className, string methodName, int[] arguments)
{
var appType = assembly.GetType($"{nameSpace}.{className}");
object app = Activator.CreateInstance(appType);
MethodInfo method = appType.GetMethod(methodName);
object result = method.Invoke(app, arguments.Cast<object>().ToArray());
return result as string;
}
Output
The final output from all this for our method with 3 passed into it is:
p0=1,p1=2,p2=3
So that was super crude, you can bypass most of the indirect reflection aspects through the use of Interfaces. If your generated script inherits from types or interfaces that the calling code also has a strong reference to, then ExecuteScript in the above example might look like this:
private static string ExecuteScript(Assembly assembly, string nameSpace, string className)
{
var appType = assembly.GetType($"{nameSpace}.{className}");
object app = Activator.CreateInstance(appType);
if (app is KnownInterface known)
{
return known.GenerateMessage(1,2,3);
}
throw new NotSupportedException("Couldn't resolve known type");
}
The major benefit to using an interface or base class reference is that you can natively set properties or call other methods without having to reflect references to them all or to resort to using dynamic which would work, but becomes a bit harder to debug.
Of course the interface solution is hard to implement when we had a variable number of parameters, so that's not the best example, usually with dynamic scripts you would construct a known environment, say a known class and methods, but you might want to inject custom code into the body of the method.
It's a bit of fun in the end, but this simple example shows that C# can be used as a runtime scripting engine without too much trouble.

Cannot find the embedded schemas in the assembly

I have DefaultSchemaSet.xsd. Now I'm getting FileNotFoundException for the codes below. Give me any suggestion, please? May I know how to solve this?
public static void GetDefaultSchemas(string path, XmlSchemaSet schemas, ValidationEventHandler schemaValidationEventHandler)
{
using (var stream = Assembly.GetExecutingAssembly().GetManifestResourceStream(path))
{
if (stream == null)
{
throw new FileNotFoundException("Cannot find the embedded schemas in the assembly!");
}
var schema = XmlSchema.Read(stream, schemaValidationEventHandler);
schemas.Add(schema);
}
}
Check the format of the resource name:
DefaultNamespace[.Subfolder][...MoreSubfolers].FileName[.extension]
You need to set Build Action to Embedded Resource in project's file's properties.
Also, you need to check the namespace you use for your project:
Try to examine the available resources, so you can find if a particular one present:
var executingAssembly = Assembly.GetExecutingAssembly();
var resourceNames = executingAssembly.GetManifestResourceNames();
foreach (var resourceName in resourceNames)
{
Console.WriteLine("Resource: " + resourceName);
Console.WriteLine("Contents:");
using (var sr = new StreamReader(executingAssembly.GetManifestResourceStream(resourceName)))
{
Console.WriteLine(sr.ReadToEnd());
}
}
Output:
Resource: EmbeddingTests.TextFile1.txt
Contents:
Hello
Resource: EmbeddingTests.NewFolder1.TextFile2.txt
Contents:
Hello 2
In order to make sure you can access it from your code you need to ensure that the file's build action is set to "Embedded Resource"
To help further we really need to see where the file lies in your solution (to give you an exact answer), however in the mean time if you ensure that your parameter "path" follows the pattern:
[DefaultNamespace].[AnySubFolders].[filename.fileextension]
note without the square brackets

Roslyn / Find References - Can't properly load Workspace

I'm trying to write some code to find all method invocations of any given method as I am looking to create an open source UML Sequence Diagramming tool. I'm having trouble, however, getting past the first few lines of code :/
The API appears to have changed drastically and I can't seem to infer proper usage by looking at the code.
When I do:
var workspace = new CustomWorkspace();
string solutionPath = #"C:\Workspace\RoslynTest\RoslynTest.sln";
var solution = workspace.CurrentSolution;
I find that workspace.CurrentSolution has 0 Projects. I figured this would be the equivalent to what was previously Workspace.LoadSolution( string solutionFile ) which would then supposedly contain any Projects in the Solution, but I am not finding any success with this path.
I am terribly confused 0.o
If someone could offer some additional guidance as to how I can use the FindReferences API to identify all invocations of a particular method, it would be very much appreciated!
Alternatively, would I be better off taking a static-analysis approach? I would like to support things like lambdas, iterator methods and async.
====================================================================
Edit -
Here is a full example based on the accepted answer:
using System.Linq;
using Microsoft.CodeAnalysis.CSharp;
using Microsoft.CodeAnalysis.CSharp.Syntax;
using Microsoft.CodeAnalysis.MSBuild;
using Microsoft.CodeAnalysis.FindSymbols;
using System.Diagnostics;
namespace RoslynTest
{
class Program
{
static void Main(string[] args)
{
string solutionPath = #"C:\Workspace\RoslynTest\RoslynTest.sln";
var workspace = MSBuildWorkspace.Create();
var solution = workspace.OpenSolutionAsync(solutionPath).Result;
var project = solution.Projects.Where(p => p.Name == "RoslynTest").First();
var compilation = project.GetCompilationAsync().Result;
var programClass = compilation.GetTypeByMetadataName("RoslynTest.Program");
var barMethod = programClass.GetMembers("Bar").First();
var fooMethod = programClass.GetMembers("Foo").First();
var barResult = SymbolFinder.FindReferencesAsync(barMethod, solution).Result.ToList();
var fooResult = SymbolFinder.FindReferencesAsync(fooMethod, solution).Result.ToList();
Debug.Assert(barResult.First().Locations.Count() == 1);
Debug.Assert(fooResult.First().Locations.Count() == 0);
}
public bool Foo()
{
return "Bar" == Bar();
}
public string Bar()
{
return "Bar";
}
}
}
CustomWorkspace is
A workspace that allows manual addition of projects and documents.
Since you're trying to load a solution, you should use the MSBuildWorkspace, which is
A workspace that can be populated by opening MSBuild solution and project files.
You can create a new MSBuildWorkspace and call OpenSolutionAsync with your solutionPath. For the reference finding part, take a look at the SymbolFinder.
Solutions are an MSBuild concept.
You need to create an MSBuildWorkspace and call OpenSolutionAsync().
string solutionPath = #"C:\Workspace\RoslynTest\RoslynTest.sln";
creates a local variable. It has no influence on your CustomWorkspace object.

HowTo read actual C# *.cs source code files into objects that can be queried

I need to write a windows app that reads an actual C# .cs file and then tries to identify what procedures on that source file do NOT have a parameter named "int userID" (it's kind of a Code Checker).
Is there something out there that already reads C# files into some sort of objects that have a property named "procedureName" and another property named "parameterList" ...that can be queried ??
If not, what would be the best way to accomplish that?
I would recommend using NRefactory which should be able to provide the solution.
The following is an example which will find all the Methods in a file with the Name of bob.
CSharpParser parser = new CSharpParser();
SyntaxTree syntaxTree = parser.Parse(#"namespace Test
{
public class TestClass
{
public void Method(string bob)
{
}
}
}");
var result = syntaxTree.Descendants.OfType<MethodDeclaration>().Where(x => x.Descendants.OfType<ParameterDeclaration>().Any(y => y.NameToken.Name == "bob"));
if (result.Any())
{
Console.WriteLine("We Win");
}

How to release COM handle in .NET

I am using the following code under ASP.NET 4.0 framework to obtain the version of MSI file from a web app:
string strVersion = "";
try
{
Type InstallerType;
WindowsInstaller.Installer installer;
InstallerType = Type.GetTypeFromProgID("WindowsInstaller.Installer");
installer = (WindowsInstaller.Installer)Activator.CreateInstance(InstallerType);
WindowsInstaller.Database db = installer.OpenDatabase(strMSIFilePath, 0);
WindowsInstaller.View dv = db.OpenView("SELECT `Value` FROM `Property` WHERE `Property`='ProductVersion'");
WindowsInstaller.Record record = null;
dv.Execute(record);
record = dv.Fetch();
strVersion = record.get_StringData(1).ToString();
dv.Close();
//db.Commit();
System.Runtime.InteropServices.Marshal.FinalReleaseComObject(dv);
System.Runtime.InteropServices.Marshal.FinalReleaseComObject(db);
}
catch
{
//Failed
strVersion = "";
}
It works fine except that when the code finishes running it holds an internal MSI file handle so when I try to move or rename the MSI file I get the error that the file is still in use. This continues until I actually navigate away from the ASPX page that calls the method above.
My question is, I obviously didn't close some handle or object in the code above. But what could that be?
PS. I'm testing it in a development IDE from VS2010.
EDIT: Edited the code like it should be after Adriano's suggestion. Thanks!
The COM object has not been released (it should be auto-released when it goes out of scope but in .NET this doesn't work really well). Because it does not implement the IDisposable interface you can't call its Dispose() method and you can't use it inside an using statement. You have to explicitly call Marshal.FinalReleaseComObject. For example:
try
{
// Your stuffs
}
finally
{
dv.Close();
Marshal.FinalReleaseComObject(dv);
Marshal.FinalReleaseComObject(db);
}
Moreover note that you do not really need a call to the Commit() method because you didn't make any change but just a query.
FWIW, you should be using Windows Installer XML (WiX) Deployment Tools Foundation (DTF). It's an FOSS project from Microsoft that can be found on CodePlex. It has MSI interop libraries with classes that are very similar to the COM classes but implement IDisosable and use P/Invoke instead of COM behind the scenes. There is even support for Linq to MSI if you want. And the full source code is available.
DTF is the gold standard for MSI interop in a .NET world. Here are two examples:
using System;
using System.Linq;
using Microsoft.Deployment.WindowsInstaller;
using Microsoft.Deployment.WindowsInstaller.Linq;
namespace ConsoleApplication3
{
class Program
{
const string DATABASE_PATH = #"C:\FOO..MSI";
const string SQL_SELECT_PRODUCTVERSION = "SELECT `Value` FROM `Property` WHERE `Property`='ProductVersion'";
static void Main(string[] args)
{
using (Database database = new Database(DATABASE_PATH, DatabaseOpenMode.ReadOnly))
{
Console.WriteLine(database.ExecuteScalar(SQL_SELECT_PRODUCTVERSION).ToString());
}
using (QDatabase database = new QDatabase(DATABASE_PATH, DatabaseOpenMode.ReadOnly))
{
var results = from property in database.Properties where property.Property == "ProductVersion" select property.Value;
Console.WriteLine(results.AsEnumerable<string>().First());
}
}
}
}
try to Dispose the Objects.
dv.Dispose();
db.Dispose();

Categories

Resources