[JENKINS-52142] Default values in closures are not work

Type: Bug
Resolution: Unresolved
Priority: Minor
Component/s: workflow-cps-plugin
Labels:
None
Environment:

Hide
Jenkins ver. 2.73.2
java.runtime.version 1.8.0_171-b10
groovy-2.4.6

pipeline-build-step 2.5.1 true
pipeline-graph-analysis 1.5 true
pipeline-input-step 2.8 true
pipeline-milestone-step 1.3.1 true
pipeline-model-api 1.2.5 true
pipeline-model-declarative-agent 1.1.1 true
pipeline-model-definition 1.2.5 true
pipeline-model-extensions 1.2.5 true
pipeline-rest-api 2.9 true
pipeline-stage-step 2.3 true
pipeline-stage-tags-metadata 1.2.5 true
pipeline-stage-view 2.9 true
pipeline-utility-steps 1.5.1 true

windows-slaves 1.3.1 true
workflow-aggregator 2.5 true
workflow-api 2.26 true
workflow-basic-steps 2.6 true
workflow-cps 2.45 true
workflow-cps-global-lib 2.9 true
workflow-durable-task-step 2.17 true
workflow-job 2.15 true
workflow-multibranch 2.16 true
workflow-scm-step 2.6 true
workflow-step-api 2.14 true
workflow-support 2.18 true

Job runs on node with RHEL

Show
Jenkins ver. 2.73.2 java.runtime.version 1.8.0_171-b10 groovy-2.4.6 pipeline-build-step 2.5.1 true pipeline-graph-analysis 1.5 true pipeline-input-step 2.8 true pipeline-milestone-step 1.3.1 true pipeline-model-api 1.2.5 true pipeline-model-declarative-agent 1.1.1 true pipeline-model-definition 1.2.5 true pipeline-model-extensions 1.2.5 true pipeline-rest-api 2.9 true pipeline-stage-step 2.3 true pipeline-stage-tags-metadata 1.2.5 true pipeline-stage-view 2.9 true pipeline-utility-steps 1.5.1 true windows-slaves 1.3.1 true workflow-aggregator 2.5 true workflow-api 2.26 true workflow-basic-steps 2.6 true workflow-cps 2.45 true workflow-cps-global-lib 2.9 true workflow-durable-task-step 2.17 true workflow-job 2.15 true workflow-multibranch 2.16 true workflow-scm-step 2.6 true workflow-step-api 2.14 true workflow-support 2.18 true Job runs on node with RHEL

Similar Issues:
Powered by SuggestiMate

Show

if (jobConfig.scm?.type in ['github', null]) {
  this.handleScmResponse = {
    // TODO: check why default values doesn't work
    script, state = '0', message = null ->
    logGithubTestStatus script, state, message
  }
} else {
  this.handleScmResponse = {
    jobConfig, script, nameOfTest, state = null, message = null ->
    def currentMessage = message ?: stateMap[state]
    echo "State: ${state}, Message: ${currentMessage}"
  }
}

def logGithubTestStatus(script, String state = '0', message = '') {
  script.echo " ${message}"
  if (state == '-1') script.error "test failed"
}

def handleScmFromShell(script, String command) {
  def result = script.sh(script: command, returnStatus: true)
  this.handleScmResponse(script, result ? '-1' : '+1')
}

Here we use default values for parameters in this.handleScmResponse closure. But if we call method handleScmFromShell which will call this.handleScmResponse closure which will call logGithubTestStatus method then we will get null values for state and message parameters in logGithubTestStatus.

Provided script will be loaded to main pipeline script using method load(). It appears defaults in pipelines that are loaded with the load() function with default parameters in closures never have the defaults resolved (so the default state = '0', for example, would actually be null).

Simple pipeline script to test an issue:

node 'master', {
   try {
      sh 'echo "def f = { x = [1] -> x += 1; x += 2; println x }\nf()\nf()\nprintln x\nreturn this" > test.groovy'
      load 'test.groovy'
   } finally {
      sh 'rm test.groovy'
   }
}

The example output from a testing job:

[workspace] Running shell script
+ echo 'def f = { x = [1] -> x += 1; x += 2; println x }
f()
f()
println x
return this'
[Pipeline] load
[Pipeline] { (test.groovy)
[Pipeline] }
[Pipeline] // load
[Pipeline] sh
[workspace] Running shell script
+ rm test.groovy
[Pipeline] }
[Pipeline] // node
[Pipeline] End of Pipeline
java.lang.NullPointerException: Cannot execute null+1

is blocked by

JENKINS-47166 new

Resolved

Some Guy created issue - 2018-06-22 12:34

Some Guy made changes - 2018-06-22 12:49

Description

Original: {code}
if (jobConfig.scm?.type in ['github', null]) {
this.handleScmResponse = {
    // TODO: check why default values doesn't work
    script, state = '0', message = null ->
    logGithubTestStatus script, state, message
}
} else {
this.handleScmResponse = {
    jobConfig, script, nameOfTest, state = null, message = null ->
    def currentMessage = message ?: stateMap[state]
    echo "State: ${state}, Message: ${currentMessage}"
}
}

def logGithubTestStatus(script, String state = '0', message = '') {
script.echo " ${message}"
if (state == '-1') script.error "test failed"
}

def handleScmFromShell(script, String command) {
def result = script.sh(script: command, returnStatus: true)
this.handleScmResponse(script, result ? '-1' : '+1')
}{code}
Here we use default values for parameters in *this.handleScmResponse* closure. But if we call method handleScmFromShell which will call *this.handleScmResponse* closure which will call *logGithubTestStatus* method then we will get null values for state and message parameters in *logGithubTestStatus*.

Provided script will be loaded to main pipeline script using method load(). It appears defaults in pipelines that are loaded with the *load()* function with default parameters in closures never have the defaults resolved (so the default state = '0', for example, would actually be null).

The example output from a job to test this:
{code:java}
[workspace] Running shell script
+ echo 'def f = { x = [1] -> x += 1; x += 2; println x }
f()
f()
println x
return this'
[Pipeline] load
[Pipeline] { (test.groovy)
[Pipeline] }
[Pipeline] // load
[Pipeline] sh
[workspace] Running shell script
+ rm test.groovy
[Pipeline] }
[Pipeline] // node
[Pipeline] End of Pipeline
java.lang.NullPointerException: Cannot execute null+1
{code}

New: {code:java}
if (jobConfig.scm?.type in ['github', null]) {
this.handleScmResponse = {
    // TODO: check why default values doesn't work
    script, state = '0', message = null ->
    logGithubTestStatus script, state, message
}
} else {
this.handleScmResponse = {
    jobConfig, script, nameOfTest, state = null, message = null ->
    def currentMessage = message ?: stateMap[state]
    echo "State: ${state}, Message: ${currentMessage}"
}
}

def logGithubTestStatus(script, String state = '0', message = '') {
script.echo " ${message}"
if (state == '-1') script.error "test failed"
}

def handleScmFromShell(script, String command) {
def result = script.sh(script: command, returnStatus: true)
this.handleScmResponse(script, result ? '-1' : '+1')
}{code}
Here we use default values for parameters in *this.handleScmResponse* closure. But if we call method handleScmFromShell which will call *this.handleScmResponse* closure which will call *logGithubTestStatus* method then we will get null values for state and message parameters in *logGithubTestStatus*.

Provided script will be loaded to main pipeline script using method load(). It appears defaults in pipelines that are loaded with the *load()* function with default parameters in closures never have the defaults resolved (so the default state = '0', for example, would actually be null).

Simple pipeline script to test an issue:

The example output from a job to test this:
{code:java}
[workspace] Running shell script
+ echo 'def f = { x = [1] -> x += 1; x += 2; println x }
f()
f()
println x
return this'
[Pipeline] load
[Pipeline] { (test.groovy)
[Pipeline] }
[Pipeline] // load
[Pipeline] sh
[workspace] Running shell script
+ rm test.groovy
[Pipeline] }
[Pipeline] // node
[Pipeline] End of Pipeline
java.lang.NullPointerException: Cannot execute null+1
{code}

Some Guy made changes - 2018-06-22 12:50

Description

Original: {code:java}
if (jobConfig.scm?.type in ['github', null]) {
this.handleScmResponse = {
    // TODO: check why default values doesn't work
    script, state = '0', message = null ->
    logGithubTestStatus script, state, message
}
} else {
this.handleScmResponse = {
    jobConfig, script, nameOfTest, state = null, message = null ->
    def currentMessage = message ?: stateMap[state]
    echo "State: ${state}, Message: ${currentMessage}"
}
}

def logGithubTestStatus(script, String state = '0', message = '') {
script.echo " ${message}"
if (state == '-1') script.error "test failed"
}

def handleScmFromShell(script, String command) {
def result = script.sh(script: command, returnStatus: true)
this.handleScmResponse(script, result ? '-1' : '+1')
}{code}
Here we use default values for parameters in *this.handleScmResponse* closure. But if we call method handleScmFromShell which will call *this.handleScmResponse* closure which will call *logGithubTestStatus* method then we will get null values for state and message parameters in *logGithubTestStatus*.

Provided script will be loaded to main pipeline script using method load(). It appears defaults in pipelines that are loaded with the *load()* function with default parameters in closures never have the defaults resolved (so the default state = '0', for example, would actually be null).

Simple pipeline script to test an issue:

The example output from a job to test this:
{code:java}
[workspace] Running shell script
+ echo 'def f = { x = [1] -> x += 1; x += 2; println x }
f()
f()
println x
return this'
[Pipeline] load
[Pipeline] { (test.groovy)
[Pipeline] }
[Pipeline] // load
[Pipeline] sh
[workspace] Running shell script
+ rm test.groovy
[Pipeline] }
[Pipeline] // node
[Pipeline] End of Pipeline
java.lang.NullPointerException: Cannot execute null+1
{code}

New: {code:java}
if (jobConfig.scm?.type in ['github', null]) {
this.handleScmResponse = {
    // TODO: check why default values doesn't work
    script, state = '0', message = null ->
    logGithubTestStatus script, state, message
}
} else {
this.handleScmResponse = {
    jobConfig, script, nameOfTest, state = null, message = null ->
    def currentMessage = message ?: stateMap[state]
    echo "State: ${state}, Message: ${currentMessage}"
}
}

def logGithubTestStatus(script, String state = '0', message = '') {
script.echo " ${message}"
if (state == '-1') script.error "test failed"
}

def handleScmFromShell(script, String command) {
def result = script.sh(script: command, returnStatus: true)
this.handleScmResponse(script, result ? '-1' : '+1')
}{code}
Here we use default values for parameters in *this.handleScmResponse* closure. But if we call method handleScmFromShell which will call *this.handleScmResponse* closure which will call *logGithubTestStatus* method then we will get null values for state and message parameters in *logGithubTestStatus*.

Provided script will be loaded to main pipeline script using method load(). It appears defaults in pipelines that are loaded with the *load()* function with default parameters in closures never have the defaults resolved (so the default state = '0', for example, would actually be null).

Simple pipeline script to test an issue:
{code:java}
node 'master', {
   try {
      sh 'echo "def f = { x = [1] -> x += 1; x += 2; println x }\nf()\nf()\nprintln x\nreturn this" > test.groovy'
      load 'test.groovy'
   } finally {
      sh 'rm test.groovy'
   }
}
{code}
The example output from a testing job:
{code:java}
[workspace] Running shell script
+ echo 'def f = { x = [1] -> x += 1; x += 2; println x }
f()
f()
println x
return this'
[Pipeline] load
[Pipeline] { (test.groovy)
[Pipeline] }
[Pipeline] // load
[Pipeline] sh
[workspace] Running shell script
+ rm test.groovy
[Pipeline] }
[Pipeline] // node
[Pipeline] End of Pipeline
java.lang.NullPointerException: Cannot execute null+1
{code}

karthikbabu KV made changes - 2018-06-26 06:09

Link

New: This issue is blocked by ~~JENKINS-47166~~ [ ~~JENKINS-47166~~ ]

Andrew Bayer made changes - 2018-08-07 22:26

Component/s		New: workflow-cps-plugin [ 21713 ]
Component/s	Original: pipeline [ 21692 ]

Jenkins

Details

Description

Attachments

Issue Links

Activity

People

Dates