Using generics

Strain
Strain in Go 
// Package strain is a small library for modifying slices of different types.
package strain

type Ints []int
type Lists [][]int
type Strings []string

type Slicer interface {
	int | string | []int
}

// Keep filters in elements of a slice.
func Keep[T Slicer](input []T, filter func(T) bool) []T {
	if input == nil {
		return input
	}
	output := make([]T, 0, len(input))
	for _, value := range input {
		if filter(value) {
			output = append(output, value)
		}
	}
	return output
}

// Discard filters out elements of a slice.
func Discard[T Slicer](input []T, filter func(T) bool) []T {
	return Keep(input, func(val T) bool { return !filter(val) })
}

This approach starts be defining some types that will be used in the package. It then defines an interface type which will be used to constain which types are to be used for the generics:

Generics are a way of writing code that is independent of the specific types being used. Functions and types may now be written to use any of a set of types.

To keep the code DRY, there will be one function that does most of the work.

The Keep() function takes a generic type parameter of [T Slicer]. T is the name of the parameter. It does not have to be T, but it is a common convention in Go (and other languages) to start with T, and if a second name is needed, to use U, and so on. The type T is constrained to be one of the types defined in the Slicer interface. The arguments to the function are a slice of the generic type T, and a function which takes a value of type T and returns a boolean value.

Note

If you haven't passed a function as a parameter before, a short explanation can be found here.

The Keep() function is defined to return a slice of type T.

If the input is nil it is simply returned. Otherwise an output slice is made of the same type as T, given a length of 0, and given a capacity that is the length of the input slice.

Note

Since the returned slice can be no bigger than the input slice, the output slice is given a capacity of the length of the input slice to prevent a lot of reallocating of the output slice. However, if the usual expected input was large and the usual expected output was small, then a lower capacity might be used to define the output slice to save allocating more memory than needed. Since this exercise uses a small length for the input slices, then it is okay to make the capacity of the output slice the same length. For more information on the length and capacity of slices, see here.

range is used to iterate the elements of the input slice. Each element is passed to the input filter function. If the function returns true, the the element is appended to the output slice. After the iteration of all the input elements is finished, the output slice is returned from the function. So, the Keep() function is intended to keep the elements for which the filter function returns true.

The Discard() function is defined with the same signature (i.e. the same generic type parameter, arguments and return type) as Keep(), but it is intended to leave out the elements for which the filter function returns true. The Keep() function can be re-used for this by logically negating the filter function passed in to Discard(). This is done by wrapping it in another function created in Discard() like so

func(val T) bool { return !filter(val) }

This defines an anonymous function which takes an argument of the same type as the filter function and also returns a boolean value. Inside the function, it passes the argument to the filter function and returns the result of the filter function logically negated by the logical NOT operator (!). So, if the filter function returns true for Discard() to leave out the element, the anonymous function will return false when passed in to the Keep() function, so the element will not be kept in by Keep(). Keep() can be called by Discard(), since not keeping an element in with Keep() is the same as leaving it out in Discard(). Discard() passes the input slice and the anonymous function to Keep() and returns the expected result.

Being able to use the same function for multiple types, instead of having to write a separate function for each type, saves a lot of work!

7th Aug 2024 · Found it useful?