XSockets Plug-in Framework AddLocation - plugins

Composable.AddLocation doesn't works for me, even dll is loaded (i can see it in output window), but GetExport(s) return null always.
I used standard example from http://xsockets.net/docs/the-plugin-framework
So this works:
Composable.LoadAssembly(Path.Combine(Helper.PluginsDirectory, "testplugin.dll"));
But this doesn't:
Composable.AddLocation(Helper.PluginsDirectory, SearchOption.AllDirectories, false);
All other code is same.
P.S. Here is solution: Composable.AddLocation begins to work when I deleted XSockets Plug-in Framework dll and dll, which describes plugin interface from Plugins directory.

My guess is this:
You have files in "Helper.PluginsDirectory" that is already loaded by the plugin framework.
If you load one of them twice you will not be able to get the export.
A workaround...
class Program
{
static void Main(string[] args)
{
Composable.RegisterExport<IAnimal>();
//Helper that fix your issue...
Helpers.AddLocation(#"C:\Users\Uffe\Desktop\DynamicAssemblies\Implementation\bin\Debug", SearchOption.AllDirectories);
Composable.ReCompose();
var a = Composable.GetExports<IAnimal>();
foreach (var animal in a)
{
animal.Says();
}
Console.ReadLine();
}
}
public static class Helpers
{
public static void AddLocation(string location, System.IO.SearchOption searchOption)
{
foreach (var assembly in Directory.GetFiles(location, "*.dll", searchOption))
{
AssemblyName verifyName = AssemblyName.GetAssemblyName(assembly);
if(!Composable.LoadedAssemblies.Any(p => p.FullName == verifyName.FullName))
Composable.LoadAssembly(assembly);
}
}
}

Related

Eclipse JDT ListRewrite inserts new node at wrong places

I'm trying to add an annotation to some selected fields using Eclipse JDT infrastructure. However, this is not run as a plugin. I added all the required dependencies to a separate project so this can be run in batch mode. However I found out that, the ListRewrite is not inserting my annotation at the right place. I have given the code below. I initially get all the field declarations in a map using a visitor and then add them one by one using the code below.
FieldDeclaration fld = lVrblDet.listStringVarMap.get(propName);
final MarkerAnnotation autoWiredAnnotate = ast.newMarkerAnnotation(); autoWiredAnnotate.setTypeName(ast.newName("MyAnnot"));
lrw = rewriter.getListRewrite(fld, FieldDeclaration.MODIFIERS2_PROPERTY);
lrw.insertLast(autoWiredAnnotate, null);
Document document = new Document(cu.toString());
try {
TextEdit edits = rewriter.rewriteAST(document, null);
edits.apply(document);
} catch (MalformedTreeException | IllegalArgumentException | BadLocationException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
However the expected output is sometimes offset by 1 character.The input and output classes have been given below.
Input Class:
#SuppressWarnings("unchecked")
public class SampleClassA {
public SampleClassB classB;
public SampleClassB getClassB() {
return classB;
}
public void setClassB(SampleClassB classB) {
this.classB = classB;
}
#Deprecated
public void printNameFromSmapleClassB() {
System.out.println(this.classB.name);
}
}
Output Class:
#SuppressWarnings("unchecked") public class SampleClassA {
p #MyAnnot
ublic SampleClassB classB;
public SampleClassB getClassB(){
return classB;
}
public void setClassB( SampleClassB classB){
this.classB=classB;
}
#Deprecated public void printNameFromSmapleClassB(){
System.out.println(this.classB.name);
}
}
As you can see in the code above, the Annotation messed with the modifier. I have tried multiple combinations of insertFirst,insertLast.Examples on the net are incomplete. Can somebody point me the mistake/the right resource ?
I just couldn't get it to work with ListRewrite. I don't know what I was doing wrong. So I wrote a visitor to store all the FieldDeclarations in a map.
#Override
public boolean visit(FieldDeclaration node) {
for (Object obj : node.fragments()) {
listStringVarMap.put(((VariableDeclarationFragment) obj).getName().toString(), node);
}
return false;
}
I looped through the map and inserted the annotation nodes as a modifiers, for the declarations that met my criteria. Please do remember to turn on recormodifications for the compilation unit you are modifying.
CompilationUnit cu = jFileAst.getEquivCompilUnit();
cu.recordModifications();
FieldDeclaration fldDecl = lVrblDet.listStringVarMap.get(propName);
importVo = (JavaAnnotImportVo) javaAstNodeCreator
.createASTNode(SpringAnnotationEnum.AutowireAnnotation, ast);
cu.imports().add(importVo.getImpDecl());
fldDecl.modifiers().add(0, importVo.getAnnotNode());
Finally write to file on disk/save back. Formatting(optional) before saving is a good idea, because the node insertions mess up with the code formatting.

Trouble with Include extension on IObjectSet not working

Could anyone help me in assessing why the code below doesn't work. I'm using the common extension method for implementing Include when using IObjectset. In our repositories we were seeing this not returning correctly so I've isolated the code in test app as below. I've also included the interface based Context if this may prove relevant and a screenshot of the relevant model section. This occurs for all Includes on IObjectSet properties not just the DPASelections one I've chosen for this example.
If I update the context to return ObjectSet (still using the POCO entities) rather than IObjectSet it all works fine. When using IObjectSet and the extension method and step through the code I see that the extension method is completing correctly with a call to the ObjectQuery we're casting to but the included entities are never returned on the graph. As said, this works perfectly when I don't interface out the Context and return ObjectSet properties hence calling Include directly on ObjectSet.
I'm not getting any errors on executing the query so this isn't the same as several other questions on SO which refer to compiled queries.
Has anyone else experienced problems with this extension method implementation or can anyone spot what I'm doing wrong here?
Any help very much appreciated.
static void Main(string[] args)
{
using (var context = new AssocEntities())
{
context.ContextOptions.LazyLoadingEnabled = false;
Candidate candidate = context.Candidates
.Include("DPASelections.DPAOption")
.SingleOrDefault(c => c.Number == "N100064");
//Count is 0 when using ext. method and IObjectSet through AssocContext but correct when using Include
//on ObjectSet through AssocContext
Console.WriteLine("DPASelection count = {0}",candidate.DPASelections.Count);
//This is always null when using IObjectSet and ext. method but populated
//when using Include on ObjectSet
var option = candidate.DPASelections.First().DPAOption;
Console.WriteLine("First DPAOption = {0} : {1}",option.Id,option.Text);
}
Console.ReadLine();
}
}
public static class Extensions
{
public static IQueryable<TSource> Include<TSource>(this IQueryable<TSource> source, string path)
{
var objectQuery = source as ObjectQuery<TSource>;
if (objectQuery != null)
{
objectQuery.Include(path);
}
return source;
}
}
//Subset of custom context implementing IObjectSet as returns.
//Works fine when I return ObjectSet rather than IObjectSet and use
//the Include method directly
public partial class AssocEntities : ObjectContext
{
public const string ConnectionString = "name=AssocEntities";
public const string ContainerName = "AssocEntities";
#region Constructors
public AssocEntities()
: base(ConnectionString, ContainerName)
{
this.ContextOptions.LazyLoadingEnabled = true;
}
public AssocEntities(string connectionString)
: base(connectionString, ContainerName)
{
this.ContextOptions.LazyLoadingEnabled = true;
}
public AssocEntities(EntityConnection connection)
: base(connection, ContainerName)
{
this.ContextOptions.LazyLoadingEnabled = true;
}
#endregion
#region IObjectSet Properties
public IObjectSet<Address> Addresses
{
get { return _addresses ?? (_addresses = CreateObjectSet<Address>("Addresses")); }
}
private IObjectSet<Address> _addresses;
public IObjectSet<Answer> Answers
{
get { return _answers ?? (_answers = CreateObjectSet<Answer>("Answers")); }
}
private IObjectSet<Answer> _answers;
public IObjectSet<Candidate> Candidates
{
get { return _candidates ?? (_candidates = CreateObjectSet<Candidate>("Candidates")); }
}
}
And the model...
I needed to replace objectQuery.Include(path); with objectQuery = objectQuery.Include(path);
In .Net framework 4.0 there is a build-in Extentionmethod for Include
just add the System.Data.Entity namespace.
It uses reflection - here is how it works:
private static T CommonInclude<T>(T source, string path)
{
MethodInfo method = source.GetType().GetMethod("Include", DbExtensions.StringIncludeTypes);
if (!(method != (MethodInfo) null) || !typeof (T).IsAssignableFrom(method.ReturnType))
return source;
return (T) method.Invoke((object) source, new object[1]
{
(object) path
});
}

Custom IronPython import resolution

I am loading an IronPython script from a database and executing it. This works fine for simple scripts, but imports are a problem. How can I intercept these import calls and then load the appropriate scripts from the database?
EDIT: My main application is written in C# and I'd like to intercept the calls on the C# side without editing the Python scripts.
EDIT: From the research I've done, it looks like creating your own PlatformAdaptationLayer is the way you're supposed to to implement this, but it doesn't work in this case. I've created my own PAL and in my testing, my FileExsists method gets called for every import in the script. But for some reason it never calls any overload of the OpenInputFileStream method. Digging through the IronPython source, once FileExists returns true, it tries to locate the file itself on the path. So this looks like a dead end.
After a great deal of trial and error, I arrived at a solution. I never managed to get the PlatformAdaptationLayer approach to work correctly. It never called back to the PAL when attempting to load the modules.
So what I decided to do was replace the built-in import function by using the SetVariable method as shown below (Engine and Scope are protected members exposing the ScriptEngine and ScriptScope for the parent script):
delegate object ImportDelegate(CodeContext context, string moduleName, PythonDictionary globals, PythonDictionary locals, PythonTuple tuple);
protected void OverrideImport()
{
ScriptScope scope = IronPython.Hosting.Python.GetBuiltinModule(Engine);
scope.SetVariable("__import__", new ImportDelegate(DoDatabaseImport));
}
protected object DoDatabaseImport(CodeContext context, string moduleName, PythonDictionary globals, PythonDictionary locals, PythonTuple tuple)
{
if (ScriptExistsInDb(moduleName))
{
string rawScript = GetScriptFromDb(moduleName);
ScriptSource source = Engine.CreateScriptSourceFromString(rawScript);
ScriptScope scope = Engine.CreateScope();
Engine.Execute(rawScript, scope);
Microsoft.Scripting.Runtime.Scope ret = Microsoft.Scripting.Hosting.Providers.HostingHelpers.GetScope(scope);
Scope.SetVariable(moduleName, ret);
return ret;
}
else
{ // fall back on the built-in method
return IronPython.Modules.Builtin.__import__(context, moduleName);
}
}
Hope this helps someone!
I was just trying to do the same thing, except I wanted to store my scripts as embedded resources. I'm creating a library that is a mixture of C# and IronPython and wanted to distribute it as a single dll. I wrote a PlatformAdaptationLayer that works, it first looks in the resources for the script that's being loaded, but then falls back to the base implementation which looks in the filesystem. Three parts to this:
Part 1, The custom PlatformAdaptationLayer
namespace ZenCoding.Hosting
{
internal class ResourceAwarePlatformAdaptationLayer : PlatformAdaptationLayer
{
private readonly Dictionary<string, string> _resourceFiles = new Dictionary<string, string>();
private static readonly char Seperator = Path.DirectorySeparatorChar;
private const string ResourceScriptsPrefix = "ZenCoding.python.";
public ResourceAwarePlatformAdaptationLayer()
{
CreateResourceFileSystemEntries();
}
#region Private methods
private void CreateResourceFileSystemEntries()
{
foreach (string name in Assembly.GetExecutingAssembly().GetManifestResourceNames())
{
if (!name.EndsWith(".py"))
{
continue;
}
string filename = name.Substring(ResourceScriptsPrefix.Length);
filename = filename.Substring(0, filename.Length - 3); //Remove .py
filename = filename.Replace('.', Seperator);
_resourceFiles.Add(filename + ".py", name);
}
}
private Stream OpenResourceInputStream(string path)
{
string resourceName;
if (_resourceFiles.TryGetValue(RemoveCurrentDir(path), out resourceName))
{
return Assembly.GetExecutingAssembly().GetManifestResourceStream(resourceName);
}
return null;
}
private bool ResourceDirectoryExists(string path)
{
return _resourceFiles.Keys.Any(f => f.StartsWith(RemoveCurrentDir(path) + Seperator));
}
private bool ResourceFileExists(string path)
{
return _resourceFiles.ContainsKey(RemoveCurrentDir(path));
}
private static string RemoveCurrentDir(string path)
{
return path.Replace(Directory.GetCurrentDirectory() + Seperator, "").Replace("." + Seperator, "");
}
#endregion
#region Overrides from PlatformAdaptationLayer
public override bool FileExists(string path)
{
return ResourceFileExists(path) || base.FileExists(path);
}
public override string[] GetFileSystemEntries(string path, string searchPattern, bool includeFiles, bool includeDirectories)
{
string fullPath = Path.Combine(path, searchPattern);
if (ResourceFileExists(fullPath) || ResourceDirectoryExists(fullPath))
{
return new[] { fullPath };
}
if (!ResourceDirectoryExists(path))
{
return base.GetFileSystemEntries(path, searchPattern, includeFiles, includeDirectories);
}
return new string[0];
}
public override bool DirectoryExists(string path)
{
return ResourceDirectoryExists(path) || base.DirectoryExists(path);
}
public override Stream OpenInputFileStream(string path)
{
return OpenResourceInputStream(path) ?? base.OpenInputFileStream(path);
}
public override Stream OpenInputFileStream(string path, FileMode mode, FileAccess access, FileShare share)
{
return OpenResourceInputStream(path) ?? base.OpenInputFileStream(path, mode, access, share);
}
public override Stream OpenInputFileStream(string path, FileMode mode, FileAccess access, FileShare share, int bufferSize)
{
return OpenResourceInputStream(path) ?? base.OpenInputFileStream(path, mode, access, share, bufferSize);
}
#endregion
}
}
You would need to change the constant ResourceScriptsPrefix to whatever your base namespace is where you stored the python scripts.
Part 2, The custom ScriptHost
namespace ZenCoding.Hosting
{
internal class ResourceAwareScriptHost : ScriptHost
{
private readonly PlatformAdaptationLayer _layer = new ResourceAwarePlatformAdaptationLayer();
public override PlatformAdaptationLayer PlatformAdaptationLayer
{
get { return _layer; }
}
}
}
Part 3, finally, how to get a Python engine using your custom stuff:
namespace ZenCoding.Hosting
{
internal static class ResourceAwareScriptEngineSetup
{
public static ScriptEngine CreateResourceAwareEngine()
{
var setup = Python.CreateRuntimeSetup(null);
setup.HostType = typeof(ResourceAwareScriptHost);
var runtime = new ScriptRuntime(setup);
return runtime.GetEngineByTypeName(typeof(PythonContext).AssemblyQualifiedName);
}
}
}
It would be easy to change this to load scripts from some other location, like a database. Just change the OpenResourceStream, ResourceFileExists and ResourceDirectoryExists methods.
Hope this helps.
You can re-direct all I/O to the database using the PlatformAdaptationLayer. To do this you'll need to implement a ScriptHost which provides the PAL. Then when you create the ScriptRuntime you set the HostType to your host type and it'll be used for the runtime. On the PAL you then override OpenInputFileStream and return a stream object which has the content from the database (you could just use a MemoryStream here after reading from the DB).
If you want to still provide access to file I/O you can always fall back to FileStream's for "files" you can't find.
You need to implement import hooks. Here's an SO question with pointers: PEP 302 Example: New Import Hooks

Using IoC container as a service locator for HttpHandler

This question relates to my other post.
Ok so after a bit more messing around I decided to do it this way. Which seems to work fine when I run it, although I'm getting the following error in NUnit: Could not load file or assembly 'Castle.Core, Version=1.0.3.0, Culture=neutral, PublicKeyToken=407dd0808d44fbdc' or one of its dependencies. The located assembly's manifest definition does not match the assembly reference. (Exception from HRESULT: 0x80131040) So not sure what is happening there???
Just wanted to know what others thought about the design and if there are any obvious 'no no's' or improvements. I.e. Is the constructor of the base handler a good place to instantiate the windsor component or is there a better place to do this? As I said in the original post the idea behind doing things this way was to keep the components nicely decoupled and to make unit testing easy. I should also add I'm new to unit testing, mocking. Thanks!
public abstract class BaseHttpHandler : IHttpHandler
{
private HttpContext _httpContext;
private ILogger _logger;
private IDataRepository _dataRepository;
protected HttpRequest Request { get { return _httpContext.Request; } }
protected HttpResponse Response { get { return _httpContext.Response; } }
protected bool IsRequestFromUAD { get { return Request.UserAgent == null ? false : Request.UserAgent.Equals("UAD"); } }
protected ILogger Logger { get { return _logger; } }
protected IDataRepository DataRepository { get { return _dataRepository; } }
public virtual bool IsReusable { get { return false; } }
public BaseHttpHandler()
{
var container = new WindsorContainer(new XmlInterpreter(new ConfigResource("castle")));
_logger = container.Resolve<ILogger>();
_dataRepository = container.Resolve<IDataRepository>();
}
public void ProcessRequest(HttpContext context)
{
_httpContext = context;
ProcessRequest(new HttpContextWrapper(context));
}
public abstract void ProcessRequest(HttpContextBase context);
}
public class UADRecordHttpHandler : BaseHttpHandler
{
public override void ProcessRequest(HttpContextBase context)
{
if (IsRequestFromUAD)
{
using (var reader = new StreamReader(context.Request.InputStream))
{
string data = reader.ReadToEnd();
if (Logger != null)
Logger.Log(data);
if(DataRepository != null)
DataRepository.Write(data);
context.Response.Write(data);
}
}
else
ReturnResponse(HttpStatusCode.BadRequest);
}
}
That's a very bad thing to do, what you're doing here. You should have one instance of the container per application, while with this code you will have one per each request.
About the error in NUnit: make sure you don't have other versions of Castle assemblies in the GAC. If so, uninstall them.
About your BaseHttpHandler: the problem with this implementation is that you're creating a new container. Instead, use a single container per application, like Krzysztof said. Use a static service locator, e.g. CommonServiceLocator. (I never recommend this but it's one of the few places where it does make sense).

Does Castle-Windsor support ForwardedTypes via XML configuration

I have a class that implements multiple interfaces.
I would like to register these interfaces via XML.
All I've found is documentation for the new Fluent Interface.
Is this option supported via XML?
What would be involved in adding this feature?
[Update] This is now possible in Windsor 2.1 or newer. See the documentation for syntax here.
This feature has not been implemented in the XML interpreter as of yet.. however it is not difficult to add support for it via a facility (obviously this technique is also useful when wanting to add other features absent from the existing configuration parser).
So first off we add a facility which will detect when a handler is created for a type, and at the same time will register any of the forwarded services so they point to the existing handler:
public class HandlerForwardingFacility : AbstractFacility
{
IConversionManager conversionManager;
protected override void Init()
{
conversionManager = (IConversionManager)Kernel.GetSubSystem(SubSystemConstants.ConversionManagerKey);
Kernel.HandlerRegistered += new HandlerDelegate(Kernel_HandlerRegistered);
}
void Kernel_HandlerRegistered(IHandler handler, ref bool stateChanged)
{
if (handler is ForwardingHandler) return;
var model = handler.ComponentModel;
if (model.Configuration == null) return;
var forward = model.Configuration.Children["forward"];
if (forward == null) return;
foreach (var service in forward.Children)
{
Type forwardedType = (Type)conversionManager.PerformConversion(service, typeof (Type));
Kernel.RegisterHandlerForwarding(forwardedType, model.Name);
}
}
}
And then of course we need to make use of this in code, for this example I'm going to have a mutant duck/dog component that supports two separate services - IDuck and IDog:
public interface IDog
{
void Bark();
}
public interface IDuck
{
void Quack();
}
public class Mutant : IDog, IDuck
{
public void Bark()
{
Console.WriteLine("Bark");
}
public void Quack()
{
Console.WriteLine("Quack");
}
}
Now to actually configure the container:
<castle>
<facilities>
<facility id="facility.handlerForwarding" type="Example.Facilities.HandlerForwardingFacility, Example" />
</facilities>
<components>
<component id="mutant" service="Example.IDog, Example" type="Example.Mutant, Example">
<forward>
<service>Example.IDuck, Example</service>
</forward>
</component>
</components>
</castle>
And now we can happily execute a test like this:
WindsorContainer container = new WindsorContainer(new XmlInterpreter());
var dog = container.Resolve<IDog>();
var duck = container.Resolve<IDuck>();
Assert.AreSame(dog, duck);
Hope this helps.