Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I agree. These functions have completely changed how I approach algorithms and data manipulation.

I was super-excited to learn Go, but it was a showstopper to discover that Go's lack of support for generics effectively made this pattern of programming impossible. Major disappointment.



Genuinely curious, what do generics have to do with this?


The type of map is `(a -> b) -> [a] -> [b]`, which means that map takes a function from any type `a` to any type `b` and a list of `a`s, and returns a list of `b`s. Generics, aka parametric polymorphism, are required for this because the definition of map makes no assumptions on what types `a` and `b` will be instantiated to.


What does "from any type 'a' to any type 'b'" mean? That sounds like a range of types, but I don't see how that makes sense. Did you just mean that the map converts from 'a' to 'b'?


a and b are placeholders (type variables). They can be any type, for example Integer, or a Float. It doesn't matter.

[a] is a list of a, [b] is a list of b. Map takes a function that takes an a and returns a b, and applies it to every element of [a] in order to produce the [b]. An example usage is

  map lessThanThree [1,2,3,4]
In this case, map takes a function where a is an integer, and b is a boolean, then takes a list of integers and produces a list of booleans.

The result is [True,True,False,False]


It's also illuminating to note that haskell curries function application. This allows us to think of map as a function of one argument.

map :: (a -> b) -> ([a] -> [b])

so instead of thinking that map takes a function and a list and applies that function to the list, we can think that map takes a function and "lifts" that functions argument and result types to the list type.

This thinking leads us directly to Functors and fmap




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: