It has been said that C# can be regarded as a functional programming language, even though it is widely recognized as a OO programming language.
So, what feature set makes C# a functional programming language?
I can only think of:
Anything else?
There being no rigourous definition of "OO Language", "Functional Language", "Procedural Language", one can make arguments that any language fits mostly any classification; one can write procedural Java, object oriented C and functional C++. I typically use a classification based around what the main semantic features support, along with common development practice. A good way of looking at this is to examine builtin and popular frameworks, and see what style they use.
Functional languages are mostly defined as those with first class function primitives, with development styles that use these to reduce complexity with idioms like "map". One other common feature is pattern matching, but I don't see this as exclusively functional. "Pure" functional languages also have no side effects, but that's not mandatory (see how fuzzy these concepts are? :).
So, what's C#? Well, it has first class function style primitives, with delegates (and has gained better syntactic support for the style with anonymous delegates and lambdas). Does this make it functional? Perhaps, if one writes in a functional style. Does the Framework use this style? No, not really.
As such, I wouldn't class C# as functional in general discussion - it is, at best, multi-paradigm, with some functional flavour.
C# has the some functional language features, closures, for example. The .NET libraries aren't written in a functional style, so in practice C# isn't a functional language. Almost everything is accomplished with mutation. The collection types are all mutable.
Well, delegates and closures allow it to operate in a largely functional way... for example:
var sum = data.Sum(x=>x.SomeProp);
etc
You can write most higher-order functions using lambdas / delegates. The type inference isn't quite the same as pure functional languages such as F#, but C# generic-type-inference is still pretty good (especially in C# 3.0).
This is especially true in .NET 3.5 and C# 3.0, where LINQ takes a highly-functional approach to many of the problems. But you can still use the functional aspects of C# with .NET 2.0 and C# 2.0. It is just easier with C# 3.0 and lambdas ;-p
Actually, C# is a pragmatic programming language. It aims to make it possible to use a number of paradigms, without punishing you hideously if you want to do something different.
Function pointers is another feature that C# has in the functional category.
I don't think C# is very widely regarded as a functional language, however. I do think it's important to point out that you can program in a functional style in many languages that aren't purely functional.
From Functional Programming:
In computer science, functional programming is a programming paradigm that treats computation as the evaluation of mathematical functions and avoids state and mutable data. It emphasizes the application of functions, in contrast to the imperative programming style, which emphasizes changes in state.
Using that definition, you can program in a functional style in almost any procedural language. Purely functional languages just enforce it.
I mostly agree with the others here who say that C# is better described as multi-paradigm than functional. But I'd add to the examples of functional features in C# LINQ, a first-class and relatively understandable system for writing monads. While purely functional languages don't require the use of monads, the example of Haskell has shown that they can be extremely useful. Yet they're one of the hardest things go grasp for many people new to Haskell. In C#, on the other hand, many people write LINQ queries these days without even realizing that they're writing monads.
C# has borrowed a lot of features from ML and Haskell for example:
C# 2.0 brought us parametric polymorphism (or "generics"). I've heard that Dom Syme, one of the creators of F#, was largely responsible for implementing generics in the .NET BCL.
C# 2.0 also allows programmers to pass and returns functions as values for higher-order functions, and has limited support for anonymous delegates.
C# 3.0 and 3.5 improved support anonymous functions for true closures.
LINQ can be considered C#'s own flavor of list comprehensions.
Anonymous types look like an approximation of ML records
Type-inference is a given.
I don't know about you, but C# extension methods look an awful lot like Haskell type classes.
There's been a lot of talk about the "dynamic" keyword in C# 4.0. I'm not 100% sure of its implementation details, but I'm fairly sure its going to use structural typing rather than late binding to retain C#'s compile time safety. Structural typing is roughly equivalent to "duck typing for static languages", its a feature that Haskell and ML hackers have been enjoying for years.
This isn't to say that C# is a functional programming language. Its still missing important features such as pattern matching, tail-call optimization, and list and tuple literals. Additionally, idiomatic C# is fundamentally imperative with a heavy dependence on mutable state.
I wouldn't necessarily consider some of those features mentioned above as exclusive to functional programming languages, but its pretty clear that the C# developers have taken a lot of inspiration from functional programming languages in the past few years.
Hi,
you can find a great overview regarding language features in the presentation from Andrew Kennedy (from MS Research) called C# is a functional programming language. My article about functional programming in C# and F# gives an overview from a higher level perspecitve (especially towards the end).
T.