How to use cobertura with parallel build ?

[ikus060]

I'm trying to understand how to use cobertura when we are compiling and testing on multiple configuration. We have create a build to run on multiple axis (debian version, python version, cherrypy version, etc.) Each of them are running tests and generating a coverage.xml report.

How should I use cobertura in this situation ?

Here the Jenkins file: https://github.com/ikus060/rdiffweb/blob/develop/Jenkinsfile

Right now, I've try to publish the coverage report for each build. But it adds an entry on the Job. Like this

[jeffpearce]

One thing I've done to combine them is to stash each report, using a unique name for each, then at the end of the job unstash them all into a folder and publish from there - the cobertura plugin will combine them into one report

[jeffpearce]

e.g.
First report:

dir('path-to-coverage) {
    stash name: "coverage-1", includes: "*.xml"
}

Second report:

dir('path-to-coverage') {
    stash name: "coverage-2", includes: "*.xml"
}

At the end of your job:

stage('publish code coverage') {
    // Publish combined code coverage results 
    unstash 'coverage-1';
    unstash 'coverage-2';

    step([$class: 'CoberturaPublisher', \
      coberturaReportFile: "*.xml"])
}

[damonZDM]

例如 第一份报告：

dir('path-to-coverage) {
    stash name: "coverage-1", includes: "*.xml"
}

第二份报告：

dir('path-to-coverage') {
    stash name: "coverage-2", includes: "*.xml"
}

在你的工作结束时：

stage('publish code coverage') {
    // Publish combined code coverage results 
    unstash 'coverage-1';
    unstash 'coverage-2';

    step([$class: 'CoberturaPublisher', \
      coberturaReportFile: "*.xml"])
}

Nice answer!