I recently needed to parse a text file and convert it to a json file, all within a Jenkins pipeline. I ran into some interesting issues during that process that i’d like to share.

I had a file that looked like below, that I needed to group the apps by their environment

dev/app1
dev/app2
stage/app3
prod/app4
...

Desired result

{
  "dev": ["app1", "app2"],
  "stage": ["app3"],
  "prod": ["app4"]
}

I initially wrote some code that did the work.

def envMap = [:].withDefault{key -> return []}
fileContents.eachLine { line ->
    split = line.split('/')
    def env = split[0]
    def app = split[1]
    envMap[env].add(app)
}

I ran it through Jenkins and I got an error: java.io.NotSerializableException: groovy.lang.MapWithDefault. Not being too familiar with this error and groovy not being my strongest language, I did the next logical step, I Google’d the error. We’ll, I got a few hits but ultimately, it always ended up at the same location JENKINS-38186.

In short, using MapWithDefault cannot store the state of execution, so it can be paused/resumed (as i understand).

So, after a little trial and error, I change the logic to use a simpler approach, which worked!

def envMap = [:]
fileContents.eachLine { line ->
    split = line.split('/')
    def env = split[0]
    def app = split[1]
    if (!envMap.containsKey(env)) {
        envMap.put(env, [])
    }
    envMap.get(env).add(app)
}