schmu

Strongly typed, compiled pet programming language.

Disclaimer schmu is a passion project which I develop for fun. Please don't use it for anything too serious. Also, the name will most likely change.

schmu is the language I'd like to program in: A strongly typed, type-inferred compiled language that can be programmed in a functional way (see below). It prefers stack- over heap allocations, and can easily interface with C code. Think OCaml, but slightly more control over allocations and data layout.

Here's what it looks like:

Fibonacci example

-- variable binding
let number = 35

-- calculate fibonacci number
fun rec fib(n) {
  match n {
    0 | 1: n
    _: fib(n - 1) + fib(n - 2)
  }
}

-- and print it
fib(number).fmt().print()

More examples can be found in the skeleton std library or the test directory.

Features

• Functional schmu is a functional language based on a Hindley-Milner type system. This means all the basic features one might expect from an ML style language are (will be) present, like

  • Parametric polymorphism
  • Higher order functions and automatic closures
  • Algebraic data types and pattern matching
  • Module system
  • Full type inference within a module, but interfaces between modules
  • Focus on recursion

• Mutable Value Semantics schmu implements mutable value semantics a la hylo/val. This means references are second-class citizens and cannot be stored in records or returned from functions. To make this feasible, schmu has move semantics and a simple borrow checkers for downward borrows, such as arguments to functions.

• Practical schmu aims to be a practical language. Data types are unboxed to make it straightforward to use C code. It doesn't try to compete with the fastest languages out there, but should be reasonably fast thanks to LLVM. It also doesn't try to be a system programming language. Some low-level access necessary to interface with C code, but that's not what schmu excels at.

Current focus

• Escape analsyis for closures
• Backtraces