Dynamically filling parameters from a file in a Jenkins pipeline - jenkins

Question

TL;DR:
I would like to use ActiveChoice parameters in a Multibranch Pipeline where choices are defined in a YAML file in the same repository as the pipeline.
Context:
I have config.yaml with the following contents:
CLUSTER:
dev: 'Cluster1'
test: 'Cluster2'
production: 'Cluster3'
And my Jenkinsfile looks like:
pipeline {
agent {
dockerfile {
args '-u root'
}
}
stages {
stage('Parameters') {
steps {
script {
properties([
parameters([
[$class: 'ChoiceParameter',
choiceType: 'PT_SINGLE_SELECT',
description: 'Select the Environemnt from the Dropdown List',
filterLength: 1,
filterable: false,
name: 'Env',
script: [
$class: 'GroovyScript',
fallbackScript: [
classpath: [],
sandbox: true,
script:
"return['Could not get The environemnts']"
],
script: [
classpath: [],
sandbox: true,
script:
'''
// Here I would like to read the keys from config.yaml
return list
'''
]
]
]
])
])
}
}
}
stage("Loading pre-defined configs") {
steps{
script{
conf = readYaml file: "config.yaml";
}
}
}
stage("Gather Config Parameter") {
options {
timeout(time: 1, unit: 'HOURS')
}
input {
message "Please submit config parameter"
parameters {
choice(name: 'ENV', choices: ['dev', 'test', 'production'])
}
}
steps{
// Validation of input params goes here
script {
env.CLUSTER = conf.CLUSTER[ENV]
}
}
}
}
}
I added the last 2 stages just to show what I currently have working, but it's a bit ugly as a solution:
The job has to be built without parameters, so I don't have an easy track of the values I used for each job.
I can't just built it with parameters and just leave, I have to wait for the agent to start the job, reach the stage, and then it will finally ask for input.
Choices are hardcoded.
The issue I'm currently facing is that config.yaml doesn't exist in the 'Parameters' stage since (as I understand) the repository hasn't been cloned yet. I also tried using
def yamlFile = readTrusted("config.yaml")
within the groovy code but it didn't work either.
I think one solution could be to try to do a cURL to the file, but I would need Git credentials and I'm not sure that I'm going to have them at that stage.
Do you have any other ideas on how I could handle this situation?