I am using the open source Saxon XSLT processor for .NET to execute some 2.0 transforms.
I reference the saxon9api.dll as I would any other dll, and can compile code against this. However Visual Studio does not show any intellisense making the IDE as useful as notepad.
The saxon9api.dll is using the IKVM Java for .NET platform, and I wonder if this is the causing VS a problem. Reflector can inspect the DLL without issue, but I suspect VS is not happy for some reason.
Any ideas?
EDIT:
Surprised that no one else has encountered this behaviour seeing as Microsoft recommends (link is now dead) the use of Saxon in the absense of built in functionality in the framework.
I think I will reword the question to be about assemblies running under IKVM not showing intellisense although I will need to find another IKVM based project to prove that this is the case first...
To make compiling, running and intellisense work in Visual Studio, you need to do the following:
Reference saxon9api.dll, as you already did
Reference IKVM.OpenJDK.Core.dll
Reference IKVM.Runtime.dll (not sure this is needed, but I always include it)
If you are also referencing vjslib, you may run into issues, because it uses a lot of the same namespaces and classnames, leading to ambiguities that can cause Visual Studio's intellisense to get into problems. Perhaps other libraries exist that show the same namespace clashes. In that case, try this on a fresh project first and add the references one by one, starting with Saxon's dependencies first.
After you do this, at least in Visual Studio 2010, 2012 and 2013, you will find that the context-sensitive help is working (image is of VS 2012 with R#):
Note: since this post is old, it may have only applied to Visual Studio 2008 at the time, I have not tested that as I am not using it anyore.
Related
I'm new to Visual Studio Code (on Mac). We are using it to develop ASP.NET Core MVC applications in C#. When I hit F12 on a symbol, like classes, interfaces or variables, VS Code will jump to the definition of that symbol if the definition exists in the source files, sweet. However nothing will happen if the definition is from an external dependency. If I did this with Visual Studio on Windows, in this situation the default behaviour is jumping to the definition provided by its metadata. Is there a similar way in VS Code to do this, rather than doing nothing in this case? It would be very helpful during development.
The new OmniSharp C# extension for VS Code for >=.NET Core 1.0 now has this feature. If there is no source code available then jump to the definition provided by its metadata. (Since OmniSharp C# extension 1.6.0 for VS Code which I used. But the legacy OmniSharp C# extension for DNX doesn't have this feature)
I have a solution in Visual Studio 2013 (c# code) that has a reference to a third-party assembly that I added with NuGet (ICSharpCode.SharpZipLib). When I run the "Code Analysis" on the solution I get lots of warnings coming from the third-party assembly. How do I tell VS that I only want code analysis advice from my code?
I'm pretty sure the errors are coming from the library itself and not my usage of it. There are a total of 32 issues e.g.
CA2237: Mark ISerializable types with SerializableAttribute.
I am using ILMerge to bundle into a single executable do you think that has confused the code analysis?
There is no way to tell FxCop to ignore namespaces or types in your code if it lives in the same binaries. When using ILMerge, the SharpZipLib essentially becomes part of your code, as it's merged into your assembly.
To get proper results, run FxCop on the assembly that's generated before the merge.
Other solutions will mention to put a [GeneratedCode] attribute on classes you want excluded, but IlMerge has no option to to that while merging, so that won't help you out.
In Visual Studio 2015 Code Analysis works on the source code using the new Roslyn Engine, which will solve the issue probably, as it does the analysis even before compilation and as such will not be impacted by your use of IlMerge.
I'm trying to create a SSIS project with a Script Task that uses Visual Basic Code.
In it, I'm trying to use the Microsoft Scripting Runtime COM object. My VS 2010 has installed the .NET 4.5 Framework but whenever I try to add the Reference it comes with this problem:
*Edit #2: Visual Studio actually pops up an error:
After checking the References Properties of the Project, the Path says: "The System cannot find the reference path". Even then you can manually reach it yourself.
Upon further investigation I found that ALL COM objects encounter the same issue:
I have seen a number of solutions pointing to a re-install of Visual Studio and the like. Which I have done with no success.
Does anyone have an idea of what could be causing the problem?
*Edit: I think it bears to point, i have tried some funny business with the Interop (setting it to true and to false does nothing). I saw some answers online but this made no difference.
I've browsed around StackOverflow but couldn't find any pratical solution to something that would seem to have such an easy solution: I had a bunch of .Net projects that were developed back in VS2005 or VS2008 and I imported them into VS2010. One of them is a C++ project, which currently targets framework 4.0 (not by my choice). One of our clients is having a problem running this application, the lack of a MSVCP100D.dll. I checked this thread what is MSVCP100D.dll? and the most accepted answer is simple: having the client install Microsoft Visual C++ 2010 Redistributable Package. However, the client is stubborn enough not to install it and I know for a fact that they have Microsoft Visual C++ 2008 Redistributable installed. So, if I'm correct in my conclusions and MSVCP100D.dll is new to VS2010, I could just target a previous framework version, rebuild the project in VS2010 and I'd be good to go. The problem is: how do I chance a VC++ target framework? I could find several guidelines to change C# and VB projects, but none about VC++. Any pointers?
Edit: To you guys who suggested that I compile it in Release mode: I am! It's been pointed out that the "D" stands for debug, which is rather strange.
Actually, you are using the debug version of the runtime (That's the "D"). Did you try compiling with a non-debug version?
Another possibility to consider would be to statically link with the runtime library. Your program will be larger, but will not have the DLL dependency.
Go into the project's properties.
On the lefthand side, go into Configuration Properties > General.
Look at the Platform Toolset value. Select v90 from the drop down list to target 2008.
See here for further details:
Visual C++ 2010 compatibility with VC 2008
See here for yet even more details:
http://blogs.msdn.com/b/vcblog/archive/2009/12/08/c-native-multi-targeting.aspx
Also as Dark Falcon mentions, that's the debug version of the runtime.
I rebuilt F# with few syntax additions. I was able to tell VS2010 to use them for F# Interactive and for building, but the code editor still claims I'm making errors (even though the project builds just fine).
Is it even possible to change this? If it is - how?
As far as I know, there is no way to change the IntelliSense service used by F# in Visual Studio. The Visual Studio component isn't a part of the open-source release and it has a reference to strongly-signed FSharp.Compiler.dll. I'm not sure if there is some way to trick .NET to load a different assembly instead (specifying bindingRedirect in app.config requires the same strong name). If you find some way, let me know :-) I'd like to use it with my modified version of F# too.
If you want to get some IntelliSense for your modified F#, you can use the F# MonoDevelop plugin. The plugin is open-source and you can change the F# version used in the Services/FSharpCompiler.fs file (see the source on GitHub).