Haskell对Int和Double执行算术运算

Question

I've recently been writing an interpreter for a basic programming language and everything has gone ok. However can someone advise me on the best approach to add support for doubles. Currently only Ints are supported but I would like to add support for Doubles too.

Answer 1

type Env a = [[Var, a]]

This is a family of types. Env Int is an int environment, Env Double is a double environment, and so on. This is not a type that can hold both ints and doubles, which is probably what you need.

A type that can hold both integers and doubles may look like this:

data Val = IntNum Int | DoubleNum Double

and then you can have your environment as

type Env = [ (Var, Val) ]

(I don't know why you are using a list of lists here).

You need to define arithmetic operations separately for IntNum and DoubleNum cases (and perhaps for mixed operands too if your language supports that).

Adding booleans and lists is straightforward, just add another couple of cases to Val .

You will have to deal with type errors that will arise in your language as it begins to support more than one type. I don't think this simple design is well suited for a statically type-safe language. If you want that, the design has to be adapted significantly.

Answer 2

The simple answer would be to define an variant type that can hold ether Int s or Double s, eg

data Value = Int Int
           | Double Double

and modify the definition of Env and other types correspondingly:

type Env = [[(Var, Value)]]

data ValExpr = IVar Var
             | IVal Value
             ...

But if you plan on adding more types, I would consider using GADT s.

EDIT: typo