Jenkins groovy for loops behaving differently - jenkins

Question

My jenkins job looks like this:
class Device {
String name
Device(name) {
this.name = name
}
}
class globals {
static List<Device> devices = new ArrayList<Device>()
}
pipeline {
stages {
stage('test for loop') {
steps {
script {
addDevices()
def testStages = [:]
echo "TESTING REGULAR LOOP"
for(index=0; index < globals.devices.size(); index++) {
Device device = globals.devices.get(index)
echo "BEFORE TEST STAGES $device.name" //shows correct name
testStages["$device.name"] = { //correct name here
echo "INSIDE TEST STAGE $device.name" //works correctly
}
}
parallel testStages
}
}
}
stage('test for each loop') {
steps {
script {
def testStages = [:]
echo "TESTING FOR EACH LOOP"
for(Device device: globals.devices) {
echo "BEFORE $device.name" //shows correct name
testStages["$device.name"] = { //correct name here
echo "INSIDE TESTSTAGES: $device.name" //device name is always the last name in the list.
}
}
parallel testStages
}
}
}
}
}
void addDevices() {
globals.devices.add(new Device("a"))
globals.devices.add(new Device("b"))
}
The first loop works fine, it's a different device.name every time. However the second loop does not work, it prints the last item in the list. I really don't understand why. Would be really greatful if anyone can provide some insight.

Answer 1

The core of the problem is this:
int x = 1
def cls = { y -> x + y }
x = 2
println cls(3)
Output:
5
So the closure changes when the variable x changes. If you are using a normal for loop or the each method, every iteration creates a new instance of the variable (device in you case), but since the for-each loop uses iterators internally, the same variable is reused which is then used in the created closures. So only the last value of device counts in this case.