Functions in Elm

Functions in Elm

Functions in Elm are “pure”.

But what is a pure function?

I think the easiest way to think about it is, pure functions can be completely represented as a table mapping inputs to outputs.

For example, a function that doubles its input is pure. The table would look like this:

Properties of pure functions

For any given input, you’ll always get the same output

In the lookup table, every possible input is listed exactly once. If you think about what this means - the function can’t rely on getting information from global or member variables (so there’s absolutely no reason to have a this keyword). It can’t query a database to get different results and it can’t ask the user for input. There’s no point as it can’t possibly affect the outcome.

There can’t be any side-effects

There’s no way for this table to represent things like output being written to the screen, or a file, or a database.

I think a lot of what scares people away from functional languages is the syntax, especially around functions. Compared to other languages like C and JavaScript, the lack of parentheses and braces can make it look quite unfamiliar. That said, if you’re familiar with more modern JavaScript, hopefully this code will make sense:

let addTwo = x => x + 2;

You can actually write something very similar in Elm (and in Haskell and F#):

addTwo = \x -> x + 2