In the following snippet I create a func list that should, the way I see it, contain funcs that print the numbers 0 1 2.
package main
import "fmt"
func main() {
flist := make([]func(), 0)
for i := 0; i < 3; i++ {
flist = append(flist, func() { fmt.Printf("%d ", i) })
}
for j := 0; j < 3; j++ {
flist[j]()
}
}
However, all funcs in flist
are the same main.main.func1
and the output is
3 3 3
even though i
never reaches the value 3 in the loop. This would make me thing the func is simply taking the address of i
, creating the same func every time but that i
is out of scope when the funcs are called later.
What am I missing here?
If we think of what code is generated for those functions, it is identical indeed: the same variable is used every time. Even though i
is out of scope after the loop, it becomes the part of the closures and its latest value is seen by the function when they run. By the moment the loop exits, it's 3
. If you want to bind a closure to the particular value and guard it from further changes, you may create a narrow-scoped variable, like this:
package main
import "fmt"
func main() {
flist := make([]func(), 0)
for i := 0; i < 3; i++ {
j := i // a different variable for each iteration
flist = append(flist, func() { fmt.Printf("%d ", j) })
}
for j := 0; j < 3; j++ {
flist[j]()
}
}