Justin Joyce

Practical tips and tutorials about software development.

Check if a key exists in a Golang map

Use an “index expression“. Buried in that official doc link is an example like this:

value, ok := someMap["someKey"]

if ok {
  fmt.Println("it exists!")
  doThings(value)
} else {
  fmt.Println("it does not exist")
}

In other languages, using index expressions (a term I just learned) could be risky if the key doesn’t exist:

# python raises an exception
obj["bad_key"] -> KeyError: 'bad_key'

# ruby returns nil
obj["bad_key"] -> nil

// javascript returns undefined
obj.badKey -> undefined

Instead of blowing up, or returning a generic (and possibly dangerous) value, Go gives back a second piece of information letting you know if the key you used was ok or not. I suspect this is due to how Golang uses zero values. I’ll explain with an example:

myMap := map[string]int{
	"a": 1,
  "b": 2,
}

// now let's try and access a key that doesn't exist
myMap["fakeKey"]
// 0

“Why would an invalid key return a value of 0?” you might be asking. Good question. That’s because 0 is the zero value of any integer type in Golang. Since this map is mapping string keys to integer values, when a non-existent key is passed the map returns the zero value of the map’s value type, which in this case is 0.

Let’s do one more example:

anotherMap := map[string]bool {
	"a": true,
  "b": false,
}

anotherMap["c"]
// false

This second example is mapping strings to boolean values, and the zero value for booleans in Golang is false.

Without checking if your map key was actually found, you could cause a hard-to-diagnose bug; zero values look like valid values because they are! To avoid those bugs, use the ok value returned from the index expression:

value, ok := anotherMap["c"]

if !ok {
	// log a message, maybe raise an error
} else {
	// all good, do what you came to do
}

Helpful Links