Is there anything similar to a slice.contains(object) method in Go without having to do a search through each element in a slice?
18 Answers
Mostafa has already pointed out that such a method is trivial to write, and mkb gave you a hint to use the binary search from the sort package. But if you are going to do a lot of such contains checks, you might also consider using a map instead.
It's trivial to check if a specific map key exists by using the value, ok := yourmap[key] idiom. Since you aren't interested in the value, you might also create a map[string]struct{} for example. Using an empty struct{} here has the advantage that it doesn't require any additional space and Go's internal map type is optimized for that kind of values. Therefore, map[string] struct{} is a popular choice for sets in the Go world.
No, such method does not exist, but is trivial to write:
func contains(s []int, e int) bool { for _, a := range s { if a == e { return true } } return false } You can use a map if that lookup is an important part of your code, but maps have cost too.
8Starting with Go 1.18, you can use the slices package – specifically the generic Contains function: .
go get import "" things := []string{"foo", "bar", "baz"} slices.Contains(things, "foo") // true Note that since this is outside the stdlib as an experimental package, it is not bound to the Go 1 Compatibility Promise™ and may change before being formally added to the stdlib.
2The sort package provides the building blocks if your slice is sorted or you are willing to sort it.
input := []string{"bird", "apple", "ocean", "fork", "anchor"} sort.Strings(input) fmt.Println(contains(input, "apple")) // true fmt.Println(contains(input, "grow")) // false ... func contains(s []string, searchterm string) bool { i := sort.SearchStrings(s, searchterm) return i < len(s) && s[i] == searchterm } SearchString promises to return the index to insert x if x is not present (it could be len(a)), so a check of that reveals whether the string is contained the sorted slice.
Instead of using a slice, map may be a better solution.
simple example:
package main import "fmt" func contains(slice []string, item string) bool { set := make(map[string]struct{}, len(slice)) for _, s := range slice { set[s] = struct{}{} } _, ok := set[item] return ok } func main() { s := []string{"a", "b"} s1 := "a" fmt.Println(contains(s, s1)) } 2If the slice is sorted, there is a binary search implemented in the sort package.
With Go 1.18+ we could use generics.
func Contains[T comparable](s []T, e T) bool { for _, v := range s { if v == e { return true } } return false } 3func Contain(target interface{}, list interface{}) (bool, int) { if reflect.TypeOf(list).Kind() == reflect.Slice || reflect.TypeOf(list).Kind() == reflect.Array { listvalue := reflect.ValueOf(list) for i := 0; i < listvalue.Len(); i++ { if target == listvalue.Index(i).Interface() { return true, i } } } if reflect.TypeOf(target).Kind() == reflect.String && reflect.TypeOf(list).Kind() == reflect.String { return strings.Contains(list.(string), target.(string)), strings.Index(list.(string), target.(string)) } return false, -1 } You can use the reflect package to iterate over an interface whose concrete type is a slice:
func HasElem(s interface{}, elem interface{}) bool { arrV := reflect.ValueOf(s) if arrV.Kind() == reflect.Slice { for i := 0; i < arrV.Len(); i++ { // XXX - panics if slice element points to an unexported struct field // see if arrV.Index(i).Interface() == elem { return true } } } return false } 2If it is not feasable to use a map for finding items based on a key, you can consider the goderive tool. Goderive generates a type specific implementation of a contains method, making your code both readable and efficient.
Example;
type Foo struct { Field1 string Field2 int } func Test(m Foo) bool { var allItems []Foo return deriveContainsFoo(allItems, m) } To generate the deriveContainsFoo method:
- Install goderive with
go get -u - Run
goderive ./...in your workspace folder
This method will be generated for deriveContains:
func deriveContainsFoo(list []Foo, item Foo) bool { for _, v := range list { if v == item { return true } } return false } Goderive has support for quite some other useful helper methods to apply a functional programming style in go.
Not sure generics are needed here. You just need a contract for your desired behavior. Doing the following is no more than what you would have to do in other languages if you wanted your own objects to behave themselves in collections, by overriding Equals() and GetHashCode() for instance.
type Identifiable interface{ GetIdentity() string } func IsIdentical(this Identifiable, that Identifiable) bool{ return (&this == &that) || (this.GetIdentity() == that.GetIdentity()) } func contains(s []Identifiable, e Identifiable) bool { for _, a := range s { if IsIdentical(a,e) { return true } } return false } 1I created a very simple benchmark with the solutions from these answers.
It isn't a real benchmark because initially, I haven't inserted too many elements but feel free to fork and change it.
1I think map[x]bool is more useful than map[x]struct{}.
Indexing the map for an item that isn't present will return false. so instead of _, ok := m[X], you can just say m[X].
This makes it easy to nest inclusion tests in expressions.
The go style:
func Contains(n int, match func(i int) bool) bool { for i := 0; i < n; i++ { if match(i) { return true } } return false } s := []string{"a", "b", "c", "o"} // test if s contains "o" ok := Contains(len(s), func(i int) bool { return s[i] == "o" }) 1If you have a byte slice, you can use bytes package:
package main import "bytes" func contains(b []byte, sub byte) bool { return bytes.Contains(b, []byte{sub}) } func main() { b := contains([]byte{10, 11, 12, 13, 14}, 13) println(b) } Or suffixarray package:
package main import "index/suffixarray" func contains(b []byte, sub byte) bool { return suffixarray.New(b).Lookup([]byte{sub}, 1) != nil } func main() { b := contains([]byte{10, 11, 12, 13, 14}, 13) println(b) } If you have an int slice, you can use intsets package:
package main import "" func main() { var s intsets.Sparse for n := 10; n < 20; n++ { s.Insert(n) } b := s.Has(16) println(b) } There are several packages that can help, but this one seems promising:
var numbers = []int{1, 5, 4, 3, 2, 7, 1, 8, 2, 3} contains, _ := stream.Contains(7) fmt.Printf("stream.Contains(7): %v\n", contains) I created the following Contains function using reflect package. This function can be used for various types like int32 or struct etc.
// Contains returns true if an element is present in a slice func Contains(list interface{}, elem interface{}) bool { listV := reflect.ValueOf(list) if listV.Kind() == reflect.Slice { for i := 0; i < listV.Len(); i++ { item := listV.Index(i).Interface() target := reflect.ValueOf(elem).Convert(reflect.TypeOf(item)).Interface() if ok := reflect.DeepEqual(item, target); ok { return true } } } return false } Usage of contains function is below
// slice of int32 containsInt32 := Contains([]int32{1, 2, 3, 4, 5}, 3) fmt.Println("contains int32:", containsInt32) // slice of float64 containsFloat64 := Contains([]float64{1.1, 2.2, 3.3, 4.4, 5.5}, 4.4) fmt.Println("contains float64:", containsFloat64) // slice of struct type item struct { ID string Name string } list := []item{ item{ ID: "1", Name: "test1", }, item{ ID: "2", Name: "test2", }, item{ ID: "3", Name: "test3", }, } target := item{ ID: "2", Name: "test2", } containsStruct := Contains(list, target) fmt.Println("contains struct:", containsStruct) // Output: // contains int32: true // contains float64: true // contains struct: true Please see here for more details:
It might be considered a bit 'hacky' but depending the size and contents of the slice, you can join the slice together and do a string search.
For example you have a slice containing single word values (e.g. "yes", "no", "maybe"). These results are appended to a slice. If you want to check if this slice contains any "maybe" results, you may use
exSlice := ["yes", "no", "yes", "maybe"] if strings.Contains(strings.Join(exSlice, ","), "maybe") { fmt.Println("We have a maybe!") } How suitable this is really depends on the size of the slice and length of its members. There may be performance or suitability issues for large slices or long values, but for smaller slices of finite size and simple values it is a valid one-liner to achieve the desired result.
2