Jenkins特意为了更好使用pipeline,开发了一些工具类,方便我们更好地在step中处理各种需求。
一、Pipeline Utility Steps
Pipeline Utility Steps
这是个插件,需要在jenkins插件管理安装。参考来源:Jenkins高级篇之Pipeline方法篇-Pipeline Utility Steps-2-方法readJSON和writeJSON_pipeline readjson-****博客
1、findFiles (需要在项目下要有log文件)
2、readJSON (方法有两种参数,分别是文件路径和字符串;示例是文件路径;datas = readJSON text : json_string)
3、writeJSON (file入参是文件路径,json可以是readJSON的结果,也可以是转换为json的字符串;示例是writeJSON的方式)
4、readProperties (读取properties文件,得到是一个map对象)
5、readYaml (读取yaml文件,得到是一个map对象;有2种参数,同上面的readJson; datas = readYaml text : yaml_string)
6、writeYaml (入参是readYaml的结果,也可以是一个map)
pipeline { agent any stages { stage('env') { steps { script { println env.JOB_NAME println env.BUILD_NUMBER println env.WORKSPACE } } } stage('Utility Steps method---findFiles') { steps { script { files = findFiles(glob: '**/*.log') println files[0].name } } } stage('Utility Steps method---readJSON') { steps { script { file_path_read = env.WORKSPACE + "/package.json" rd_json = readJSON file : file_path_read println rd_json println rd_json.name } } } stage('Utility Steps method---writeJSON') { steps { script { file_path_write = env.WORKSPACE + "/test_jenkins.json" input_json = env.WORKSPACE + "/package.json" input = readJSON file : input_json writeJSON file: file_path_write, json: input } } } stage('Utility Steps method---readProperties') { steps { script { properties_file = env.WORKSPACE + "/test_jenkins.properties" props = readProperties interpolate: true, file: properties_file println props } } } stage('Utility Steps method---readYaml') { steps { script { yaml_file = env.WORKSPACE + "/test_jenkins.yaml" y = readYaml file : yaml_file println y } } } stage('Utility Steps method---writeYaml') { steps { script { yaml_path = env.WORKSPACE + "/test_jenkins_write.yaml" input_path = env.WORKSPACE + "/test_jenkins.yaml" input_map = readYaml file : input_path writeYaml file: yaml_path , data: input_map } } } } }
二、pipeline basic steps
这是一个最基础的pipeline组件,这个也是一个独立的插件,在安装pipeline的时候默认会自动安装
官网:Pipeline: Basic Steps
1、deleteDir()
这个是默认递归删除WORKSPACE下的文件和文件夹
使用场景:做完了每一个stage里面的事情,我们需要在post{...}里面写一些clean up操作
2、dir()
这个方法是改变当前的工作目录
3、error()
遇到这个方法,该stage和后面的stage都会显示失败
4、fileExists
判断一个文件是否存在,返回值是布尔类型
5、isUnix()
判断当前运行的Jenkins node环境是不是linux环境
6、pwd()
linux是pwd,windows上是dir。所以这个插件就干脆支持一个方法,统称为pwd()
pipeline{ agent any stages{ stage("dir") { steps{ // 如果没有这个目录,会自动创建这个目录 dir("${env.WORKSPACE}/testdata"){ sh "pwd" } } } stage("fileExists") { steps{ script { json_file = "${env.WORKSPACE}/test_jenkins.log" if(fileExists(json_file) == true) { echo("log file is exists") }else { error("here haven't find log file") } } } } stage("isUnix") { steps{ script { if(isUnix() == true) { echo("this jenkins job running on a linux-like system") }else { error("the jenkins job running on a windows system") } } } } stage("pwd()") { steps{ script { sh("pwd") println "===========" println pwd() } } } } post { always { script { //写相关清除/恢复环境等操作代码 //deleteDir() println "结束" } } } }
三、pipeline basic steps(邮件)
pipeline{ agent any environment { to_email = "yin921125@qq.com" to_email_address_list = "yin921125@qq.com,yin921125@163.com" } stages{ stage("发邮件demo 1") { steps{ script { mail to: to_email, subject: "Running Pipeline: ${currentBuild.fullDisplayName}", body: "Something is wrong with ${env.BUILD_URL}" } } } stage("发邮件demo 2") { steps{ script { mail to: to_email, cc: 'xxxxx@qq.com', // 抄送 charset:'UTF-8', // or GBK/GB18030 mimeType:'text/plain', // or text/html subject: "Running Pipeline: ${currentBuild.fullDisplayName}", body: "Something is wrong with ${env.BUILD_URL}, just for test send mail via pipeline code" } } } stage("发邮件demo 3") { steps{ script { subject = "Jenkins Job : " + env.JOB_NAME + "/" + env.BUILD_ID result_url = env.BUILD_URL + "console" text = """ <!DOCTYPE html> <html> <head> <meta charset="utf-8"> <title></title> </head> <body> <div> <h1>Summary</h1> <div> <h2>Jenkins Build</h2> <ul> <li>Job URL : <a href='${env.BUILD_URL}'>${env.BUILD_URL}</a></li> <li>Build Result URL : <a href='${result_url}'>${result_url}</a></li> </ul> </div> <div> <h2>GIT Branch</h2> <ul> <li>这是一个测试分支</li> </ul> </div> </div> </body> </html> """ mail body: text, subject: subject, mimeType: 'text/html', to: to_email_address_list } } } } post{ always{ script { module_test = load env.WORKSPACE + "/test_groovy.groovy" } } failure { script { module_test.send_email_results("Failed","Master",to_email_address_list) } } success { script { module_test.send_email_results("Success","Master",to_email_address_list) } } } }
上面引用的test_groovy.groovy如下
def send_email_results(status,GITBranch,to_email_address_list) { subject = "Jenkins Job : " + env.JOB_NAME + "/" + env.BUILD_ID + " has " + status result_url = env.BUILD_URL + "console" text = """ <!DOCTYPE html> <html> <head> <meta charset="utf-8"> <title></title> </head> <body> <div> <h1>Summary</h1> <div> <h2>Jenkins Build</h2> <ul> <li>Job URL : <a href='${env.BUILD_URL}'>${env.BUILD_URL}</a></li> <li>Build Result URL : <a href='${result_url}'>${result_url}</a></li> </ul> </div> <div> <h2>GIT Branch</h2> <ul> <li>这是测试引用其他groovy文件的邮件${GITBranch}</li> </ul> </div> </div> </body> </html> """ mail body: text, subject: subject, mimeType: 'text/html', to: to_email_address_list } return this
四、pipeline basic steps(readFile、retry、sleep)
1、readFile,读取文件内容(两个参数,第一个是file的path,第二个是encoding,默认是根据你当前平台的编码)
如果不知道文件是什么格式(格式例如txt,log,json,properties,config,yaml),则推荐用readFile
如果拿来读一个图片或者音频视频文件,读取二进制文件会采用Base64转码的字符串输出。
2、retry和sleep
retry只能在出现异常的地方才能使用,否则只跑一遍就结束
retry(3)这个3是一个整数,表示尝试的次数,sleep中的2表示2秒,注意这单位是秒
pipeline{ agent any stages{ stage("init") { steps{ script { json_file = "${env.WORKSPACE}/package.json" file_contents = readFile json_file println file_contents } } } stage("retry and sleep") { steps{ script{ try { retry(3) { println "here we are test retry fuction" sleep 5 println 10/0 } }catch (Exception e) { println e } } } } } }
五、pipeline basic steps(timeout 、waitUntil 、withEnv )
1、timeout (超时则aborted)
2、waitUntil (直到返回true才跳出,否则一直重试)
3、withEnv (添加环境变量)
pipeline{ agent any stages{ stage("init") { steps{ script { // 设置超时5s,默认单位是分钟 timeout(time: 50, unit: 'SECONDS') { //initialRecurrencePeriod :设置重试之间的初始等待时间(以毫秒为单位)。 // 默认为 250 毫秒。每次失败都会将尝试之间的延迟减慢至最多 15 秒 //quiet :如果为真,则每次检查条件时该步骤都不会记录消息。默认为假。类型: boolean println "是unix系统吗" + isUnix() waitUntil(initialRecurrencePeriod: 1000,quiet: true){ script { isUnix() } } } } } } stage("env") { steps{ script { withEnv(['java_home=/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.181-3.b13.el6_10.i386/jre']) { sh("$java_home/bin/java -version") println "test with withEnv feature" } } } } } }
六、pipeline basic steps(writeFile)
println env.JOB_NAME println env.BUILD_NUMBER pipeline{ agent any stages{ stage("writeFile demo") { steps{ script { write_file_path = "${env.WORKSPACE}/test_jenkins.txt" file_contents = "Hello Anthony!! 这是一个测试例子" // 写文件 writeFile file: write_file_path, text: file_contents, encoding: "UTF-8" // 读文件 fileContents = readFile file: write_file_path, encoding: "UTF-8" println fileContents } } } } }
参考:jenkins流水线学习二(工具) - whitewall - 博客园 (cnblogs.com)