Scheme for your tuples
This small library implements a Scheme-inspired Lisp dialect allowing you to write some lisp directly in your C# code.
In other words, by exploiting certain C# language features λTuple aims to to provide Lisp's glorious syntax and semantics from within the C#/.NET environment.
Among other things, it is:
- 🏎️ Highly dynamic, as any Lisp shall be: data and code are the same.
- ⚡ Pretty fast, because it's not an interpreter: all your code is compiled using
System.Linq.Expressions. - 🤝 Well integarted with C# and .NET: mix and match your normal C# and Lisp code as you like
- 🌀 Optimizing tail calls when possible.
- ❓ Work in progress, more goodies to come...
First things first, let's add λTuple to the usings section of your program.
using yTuple;
using static yTuple.Elementary;This should make it possible to write the following code.
var fib = Lisp.Parse(n =>
(begin,
(define, _.loop, (lambda, (_.i, _.curr, _.prev),
(cond,
((eq, _.i, 0), 0),
((eq, _.i, 1), _.curr),
(@else,
(_.loop, (sub, _.i, 1), (add, _.curr, _.prev), _.curr))))),
(_.loop, n, 1, 0))
).Compile();
Console.WriteLine(fib(42));This creates a tail-recursive lambda function fib(), that computes n-th number in the Fibonacci sequence and then uses it to output the result for n = 42.
Another good place to get a look and feel of it would be in the test code. More examples are coming soon.
- More C# interop features
- Better static typing to avoid unnecessary boxing and dynamic type casts
- Serialization, so that it's possible to load code from external source
- Support macros
- Use FastExpressionCompiler for even better performance
If you have a question, or think you found a bug, or have a good idea and don't mind sharing it, please open an issue and I would be happy to discuss it.