I have a declarative pipeline script for my multibranch project in which I would like to read a text file and store the result as a string variable to be accessed by a later step in the pipeline. Using the snippet generator I tried to do something like this:
filename = readFile 'output.txt'
For which filename
would be my string.
I get an error in the Jenkins console output:
org.codehaus.groovy.control.MultipleCompilationErrorsException: startup failed:
WorkflowScript: 30: Expected a step @ line 30, column 5.
filename = readFile 'output.txt'
Do I need to use a withEnv
step to set the output of readFile
to a Jenkins environment variable? If so, how?
Thanks
This question is related to
jenkins
groovy
jenkins-plugins
jenkins-pipeline
multibranch-pipeline
We got around this by adding functions to the environment
step, i.e.:
environment {
ENVIRONMENT_NAME = defineEnvironment()
}
...
def defineEnvironment() {
def branchName = "${env.BRANCH_NAME}"
if (branchName == "master") {
return 'staging'
}
else {
return 'test'
}
}
I can' t comment yet but, just a hint: use try/catch clauses to avoid breaking the pipeline (if you are sure the file exists, disregard)
pipeline {
agent any
stages {
stage("foo") {
steps {
script {
try {
env.FILENAME = readFile 'output.txt'
echo "${env.FILENAME}"
}
catch(Exception e) {
//do something, e.g. echo 'File not found'
}
}
}
}
Another hint (this was commented by @hao, and think is worth to share): you may want to trim like this readFile('output.txt').trim()
A complete example for scripted pipepline:
stage('Build'){
withEnv(["GOPATH=/ws","PATH=/ws/bin:${env.PATH}"]) {
sh 'bash build.sh'
}
}
According to the documentation, you can also set global environment variables if you later want to use the value of the variable in other parts of your script. In your case, it would be setting it in the root pipeline:
pipeline {
...
environment {
FILENAME = readFile ...
}
...
}
Source: Stackoverflow.com