Add a keyword to C# with code generation? - c#

I have a domain specific language that I would like to interact with C# by adding new keywords (or some keyword-like syntax). Using attributes would be insufficient (I can't use them in method bodies), and shoehorning it into 'valid' C# notation that gets compiled into something else would be ugly and ruin the analogy with the DSL (and the translation from DSL-like notation to C# is nontrivial, so just writing the C# each time is out of the question).
I already have a way to parse the .cs file and transform it into legitimate, nontrivial, C# code which can be compiled.
The problem is, even through I can do all the work of defining the DSL, parsing it, and translating it into valid C#, Visual Studio won't let me use notation it doesn't understand; it just adds red squiggles, emits an error "cannot resolve symbol", and then often fails to properly parse things after it.
Is there a way to to force visual studio to ignore specific strings in its analysis? I've looked at visual studio plugins but it looks like, although I can do syntax highlighting and other stuff, I can't force it to ignore something it doesn't know how to parse (unless I'm missing some way to do that in the extension API, which is certainly possible).
I've skimmed through the Roslyn stuff and don't see offhand a way to do this there, either. (Again, may be missing something, it doesn't seem to have great documentation.)

Take a look at PowerLanguages.E: http://visualstudiogallery.msdn.microsoft.com/a512e0d0-f4f3-4435-bad4-8d5efbb1db4a
No english docs yet, sorry

Related

What's the best way to highlight C# syntax using Roslyn?

I'm trying to make a Xamarin.Android app that highlights the syntax of many different languages. I plan to use ANTLR to deal with most of them, but for C# I want to use Roslyn as that will undoubtedly be faster and less buggy than ANTLR.
What is the best way to implement syntax highlighting with Roslyn? For highlighting Java syntax, the approach I took was parsing the text into a parse tree, and using a visitor to color the text associated with each terminal. You can view my code here. Is this also a good idea for Roslyn, or does Roslyn provide APIs for syntax highlighting? (e.g. Does the code behind syntax highlighting in Visual Studio live in the dotnet/roslyn repo?) I'd really prefer not to reinvent the wheel, but I will if I have to.
edit: I have accepted Tamas' answer because his solution is the most practical for my use case; I do not have access to the full solution to build a semantic model with, so I will have to do some of my own analysis. However, if your app supports more broad C# integration and can build a semantic model, take a look at the Roslyn Classification APIs which are used in Jonathon Marolf's answer.
the ConsoleClassifier project in the roslyn Samples should be a good starting place.
Did you have a look at SourceBrowser? If you can do a full solution build, then I would use the same approach.
If your context doesn't allow a full build, then you can implement something relatively good based on syntax token types. However you might have to handle some corner cases, like contextual keywords, var, implicitly declared local variables (like value), etc. Have a look at what SonarQube is using.
Similarly, you can look for other tools that you know are Roslyn based, like OmniSharp. I'm not sure if that uses regex or Roslyn to do the highlighting. But in any case you could get quite far with Regex too.

How to add new operator in C# using Roslyn

I am trying to implement a DSL like feature in C#. It may look something similar to LINQ queries. I am wondering if it is possible to implement new unary or binary operators using Roslyn.
I have been googling last few days without much success. It would be great if someone could point me to some samples or Roslyn documentations.
There are two ways how you could use Roslyn to implement a new C#-based language.
Use the Roslyn API to parse the source code into a syntax tree, then transform the syntax tree into actual C# and compile that.
This is ideal if your language is actually syntactically valid C# code, but the semantics are different. For example, you could implement await this way, if you forced await to look like a function call (e.g. await(x) would be valid, but not await x).
If you want to introduce new syntax (like a new operator), it might work, since Roslyn does support parsing “broken” code. But it most likely won't work that well, because then the syntax tree might not look the way you want. Worse, the results might not be consistent (sometimes, your new syntax will be parsed one way, sometimes another).
Since Roslyn is now open source, you can actually modify the source code of the compiler in any way you want, including adding a new operator.
But doing that is most likely not going to be simple. And I think the workflow is also going to be more complicated: you need to compile your own version of the compiler, not just use a library from NuGet like in option 1.

How to parse simple statement into CodeDom object

I need to parse a simple statement (essentially a chain of function calls on some object) represented as a string variable into a CodeDom object (probably a subclass of CodeStatement). I would also like to provide some default imports of namespaces to be able to use less verbose statements.
I have looked around SO and the Internet to find some suggestions but I'm quite confused about what is and isn't possible and what is the simplest way to do it. For example this question seems to be almost what I want, unfortunately I can't use the solution as the CodeSnippetStatement seems not to be supported by the execution engine that I use (the WF rules engine).
Any suggestions that could help me / point me into the right direction ?
There is no library or function to parse C# code into CodeDOM objects as part of the standard .NET libraries. The CodeDOM libraries have some methods that seem to be designed for this, but none of them are actually implemented. As far as I know, there is some implementation available in Visual Studio (used e.g. by designers), but that is only internal.
CodeSnippetStatement is a CodeDOM node that allows you to place any string into the generated code. If you want to create CodeDOM tree just to generate C# source code, than this is usually fine (the source code generator just prints the string to the output). If the WF engine needs to understand the code in your string (and not just generate source code and compile it), than CodeSnippetStatement won't work.
However, there are 3rd party tools that can be used for parsing C# source code. In one project I worked on, we used NRefactory library (which is used in SharpDevelop) and it worked quite well. It gives you some tree (AST) representing the parsed code and I'm afraid you'll need to convert this to the corresponding CodeDOM tree yourself.
I have found a library implementation here that seems to cover pretty much everything I need for my purposes. I don't know if it's robust enough to be used in business scenarios, but for my unit tests it's pretty much all I can ask for.

Reflect and Load code at design time in Visual Studio

I have an XML file that lists a series of items, and the items are often referenced by their name in code.
Is there any way to use reflection within Visual Studio to make this list 'accessible' sort of like intellisence? I've seen it done before - and I'm pretty sure it's ridiculously difficult, but I figure it can't hurt to at least ask.
I would recommend against using reflection for this.
Apart from the added complexity in the code you are also opening the code up to abuse from somebody modifying your XML to get your code to do what they want (think injection attack).
You would be better off parsing the XML as usual but using a big if / switch statement to define what how the code runs. That way you have more chance of catching any problems and validating the input.
From string to function call sounds great but will bite you in the bum.
I think he wants to access xml from c# code with intelligences.
My guess is that you would have to build some sort of code generator that would generate c# class that has the properties of you xml field... kind of how visual studio generates code for resourcefiles.

Using reflection for code gen?

I'm writing a console tool to generate some C# code for objects in a class library. The best/easiest way I can actual generate the code is to use reflection after the library has been built. It works great, but this seems like a haphazard approch at best. Since the generated code will be compiled with the library, after making a change I'll need to build the solution twice to get the final result, etc. Some of these issues could be mitigated with a build script, but it still feels like a bit too much of a hack to me.
My question is, are there any high-level best practices for this sort of thing?
Its pretty unclear what you are doing, but what does seem clear is that you have some base line code, and based on some its properties, you want to generate more code.
So the key issue here are, given the base line code, how do you extract interesting properties, and how do you generate code from those properties?
Reflection is a way to extract properties of code running (well, at least loaded) into the same execution enviroment as the reflection user code. The problem with reflection is it only provides a very limited set of properties, typically lists of classes, methods, or perhaps names of arguments. IF all the code generation you want to do can be done with just that, well, then reflection seems just fine. But if you want more detailed properties about the code, reflection won't cut it.
In fact, the only artifact from which truly arbitrary code properties can be extracted is the the source code as a character string (how else could you answer, is the number of characters between the add operator and T in middle of the variable name is a prime number?). As a practical matter, properties you can get from character strings are generally not very helpful (see the example I just gave :).
The compiler guys have spent the last 60 years figuring out how to extract interesting program properties and you'd be a complete idiot to ignore what they've learned in that half century.
They have settled on a number of relatively standard "compiler data structures": abstract syntax trees (ASTs), symbol tables (STs), control flow graphs (CFGs), data flow facts (DFFs), program triples, ponter analyses, etc.
If you want to analyze or generate code, your best bet is to process it first into such standard compiler data structures and then do the job. If you have ASTs, you can answer all kinds of question about what operators and operands are used. If you have STs, you can answer questions about where-defined, where-visible and what-type. If you have CFGs, you can answer questions about "this-before-that", "what conditions does statement X depend upon". If you have DFFs, you can determine which assignments affect the actions at a point in the code. Reflection will never provide this IMHO, because it will always be limited to what the runtime system developers are willing to keep around when running a program. (Maybe someday they'll keep all the compiler data structures around, but then it won't be reflection; it will just finally be compiler support).
Now, after you have determined the properties of interest, what do you do for code generation? Here the compiler guys have been so focused on generation of machine code that they don't offer standard answers. The guys that do are the program transformation community (http://en.wikipedia.org/wiki/Program_transformation). Here the idea is to keep at least one representation of your program as ASTs, and to provide special support for matching source code syntax (by constructing pattern-match ASTs from the code fragments of interest), and provide "rewrite" rules that say in effect, "when you see this pattern, then replace it by that pattern under this condition".
By connecting the condition to various property-extracting mechanisms from the compiler guys, you get relatively easy way to say what you want backed up by that 50 years of experience. Such program transformation systems have the ability to read in source code,
carry out analysis and transformations, and generally to regenerate code after transformation.
For your code generation task, you'd read in the base line code into ASTs, apply analyses to determine properties of interesting, use transformations to generate new ASTs, and then spit out the answer.
For such a system to be useful, it also has to be able to parse and prettyprint a wide variety of source code langauges, so that folks other than C# lovers can also have the benefits of code analysis and generation.
These ideas are all reified in the
DMS Software Reengineering Toolkit. DMS handles C, C++, C#, Java, COBOL, JavaScript, PHP, Verilog, ... and a lot of other langauges.
(I'm the architect of DMS, so I have a rather biased view. YMMV).
Have you considered using T4 templates for performing the code generation? It looks like it's getting much more publicity and attention now and more support in VS2010.
This tutorial seems database centric but it may give you some pointers: http://www.olegsych.com/2008/09/t4-tutorial-creatating-your-first-code-generator/ in addition there was a recent Hanselminutes on T4 here: http://www.hanselminutes.com/default.aspx?showID=170.
Edit: Another great place is the T4 tag here on StackOverflow: https://stackoverflow.com/questions/tagged/t4
EDIT: (By asker, new developments)
As of VS2012, T4 now supports reflection over an active project in a single step. This means you can make a change to your code, and the compiled output of the T4 template will reflect the newest version, without requiring you to perform a second reflect/build step. With this capability, I'm marking this as the accepted answer.
You may wish to use CodeDom, so that you only have to build once.
First, I would read this CodeProject article to make sure there are not language-specific features you'd be unable to support without using Reflection.
From what I understand, you could use something like Common Compiler Infrastructure (http://ccimetadata.codeplex.com/) to programatically analyze your existing c# source.
This looks pretty involved to me though, and CCI apparently only has full support for C# language spec 2. A better strategy may be to streamline your existing method instead.
I'm not sure of the best way to do this, but you could do this
As a post-build step on your base dll, run the code generator
As another post-build step, run csc or msbuild to build the generated dll
Other things which depend on the generated dll will also need to depend on the base dll, so the build order remains correct

Categories

Resources