This is literally just some testing I've done with hosting .NET Core in native applications.
Again, this is just me trying to invoke a managed method. This is a full-on mess, even the outputs aren't organized.
What this does, all in all, is it contains two projects - Native and TestModule:
- Native is a C++ project that will prompt the user to provide the path to the hostfxr.dll and the compiled TestModule's root.
- TestModule is a C# Code Library that only has one function, called Test, inside class Module, in namespace TestModule, that prints to stdout.
Important
This is for educational purposes only.
The code is badly written and messy, very static and hard-coded and generally only shows how to approach accessing the .NET Core. I've only shared it as I believe it could potentially help someone in the future but it is not a good source. A great example of doing this properly would be Godot's implementation, and even that one has plenty that can be improved!
So, since this was really a concentrated test on just invoking a managed method from C++, there's plenty of things that should probably be done and are not addressed at all in here (although they are surprisingly easy in theory):
- Automatically finding hostfxr.dll (and for the love of god - do not use
get_hostfxr_path!!!) - Calling functions with different parameters and return types
- Remove all hardcoding from there, possibly load data from the user or JSON or something else
- Allow loading DLLs without a runtimeconfig (this is actually really easy, just pretend the DLL is a console app, apparently!)