jenkins:实现Jenkinsfile与Json的转换

实现Jenkinsfile与Json的转换

最近在做个需求,需要支持Jenkinsfile和json的转换。

方法1:使用现有的jenkins插件

参考的是这篇文章。下面介绍一下将插件打包成镜像的步骤:

  • 本地安装jdk和maven,jdk建议采用的版本为8(该工程会依赖一个名为tools.jar的包,jdk 9之后移除了该包)

    如果本地没有找到tools.jar,可以下载一个1.8版本的jdk,然后在pom.xml中增加如下依赖

    <dependency>
        <groupId>jdk.tools</groupId>
        <artifactId>jdk.tools</artifactId>
        <version>1.8.0</version>
        <scope>system</scope>
        <systemPath>/root/jdk1.8.0_271/lib/tools.jar</systemPath>
    </dependency>
    
  • clone pipeline-model-definition-plugin工程

  • 在/root/.m2/目录下创建settings.xml,内容来自Jenkins官方

    <settings>
      <pluginGroups>
        <pluginGroup>org.jenkins-ci.tools</pluginGroup>
      </pluginGroups>
     
      <profiles>
        <!-- Give access to Jenkins plugins -->
        <profile>
          <id>jenkins</id>
          <activation>
            <activeByDefault>true</activeByDefault> <!-- change this to false, if you don't like to have it on per default -->
          </activation>
          <repositories>
            <repository>
              <id>repo.jenkins-ci.org</id>
              <url>https://repo.jenkins-ci.org/public/</url>
            </repository>
          </repositories>
          <pluginRepositories>
            <pluginRepository>
              <id>repo.jenkins-ci.org</id>
              <url>https://repo.jenkins-ci.org/public/</url>
            </pluginRepository>
          </pluginRepositories>
        </profile>
      </profiles>
      <mirrors>
        <mirror>
          <id>repo.jenkins-ci.org</id>
          <url>https://repo.jenkins-ci.org/public/</url>
          <mirrorOf>m.g.o-public</mirrorOf>
        </mirror>
      </mirrors>
    </settings>
    
  • 执行 mvn install进行编译

  • 由于主要用到的是Jenkinsfile和json之间的转换关系,因此主要用的是如下两个REST API:

    • Conversion to JSON representation from Jenkinsfile

      • URL: JENKINS_URL/pipeline-model-converter/toJson
      • Parameters: jenkinsfile - the Jenkinsfile contents
      • Info: Takes a Jenkinsfile and converts it to the JSON representation for its pipeline step.
      • Returns: JSON with a result field that will either be success or failure. If success, the JSON representation will be in the json field. If failure, there'll be an additional array in the errors field of the error messages encountered.
    • Conversion to Jenkinsfile from JSON representation

      • URL: JENKINS_URL/pipeline-model-converter/toJenkinsfile
      • Parameters: json - the JSON representation of the model
      • Info: Takes the JSON representation of the model and converts it to the contents for a Jenkinsfile invoking the pipeline step.
      • Returns: JSON with a result field that will either be success or failure. If success, the Jenkinsfile contents will be in the jenkinsfile field. If failure, there'll be an additional array in the errors field of the error messages encountered.

    上述两个API在pipeline-model-definition-plugin/pipeline-model-definition目录下,因此在该目录下直接运行:mvn hpi:run -Dhost=0.0.0.0 -Djetty.port=8080即可。

  • 将json转换为Jenkinsfile的操作如下:

    完整的返回值如下:

    {
        "status": "ok",
        "data": {
            "result": "success",
            "json": {
                "pipeline": {
                    "stages": [
                        {
                            "name": "Hello",
                            "branches": [
                                {
                                    "name": "default",
                                    "steps": [
                                        {
                                            "name": "echo",
                                            "arguments": [
                                                {
                                                    "key": "message",
                                                    "value": {
                                                        "isLiteral": true,
                                                        "value": "Hello World"
                                                    }
                                                }
                                            ]
                                        }
                                    ]
                                }
                            ]
                        }
                    ],
                    "agent": {
                        "type": "any"
                    }
                }
            }
        }
    }
    
  • 将Jenkinsfile转换为json的操作如下

  • 制作容器镜像时,只需要将本地工程和/root/.m2上传到容器,生成对应的镜像即可,下面Dockerfile假设生成的镜像为pipeline-model-definition-plugin:latest

    FROM pipeline-model-definition-plugin:latest
    WORKDIR /usr/pipeline-model-definition-plugin/pipeline-model-definition
    ENV PATH=$PATH:/usr/local/bin/maven-3.6.3/bin
    ENTRYPOINT ["sh", "-c", "mvn hpi:run -Dhost=0.0.0.0"]
    

    我自己打包了一个镜像:docker pull quay.io/woodliu/pipeline-model-definition-plugin

需要注意的是,本插件提供的转换API toJenkinsfile和toJson并不是万能的,只能支持jenkins标准的参数类型,例如对于gitParameter这样的参数就无法解析(扩展功能),一种解决方式是独立解析扩展的参数,然后将其插入解析好的标准JenkinsFile中;另外一个方式就是写一个jenkinsfile的解析器。

参考

  • mvn hpi的命令可以参考官方文档

  • 可以运行mvn hpi:hpi生成对应的hpi文件,如:

    /pipeline-model-definition-plugin/pipeline-model-definition/target/pipeline-model-definition.hpi

方法2:解析原生的jenkinsfile文件

在GitHub上有一个支持jenkinsfile解析的项目,该项目使用rust的pest crate来编写jenkinsfile的语法,支持对jenkinsfile的格式验证。Pest官方文档中给出了一个非常好的对json语法的解析例子,主要是使用递归的方式来解析语法。

pest官方提供了一个编辑器,可以使用该编辑器查看经过pest解析之后的字段,对了解pest的工作方式非常有用。如,使用jdp项目提供的pest文件解析如下jenkinsfile:

pipeline {
    agent {
        docker {
            reuseNode true
            image 'maven:3-alpine'
            label 'my-defined-label'
            args  '-v /tmp:/tmp'
            registryUrl 'https://myregistry.com/'
            registryCredentialsId 'myPredefinedCredentialsInJenkins'
        }
    }

    stages {
        stage('Build') {
            steps { sh 'make' }
        }
    }
}

对应的解析结果如下:

- preceeding_junk: ""
- opening_brace: "{"
- agentDecl > agentBlock
  - opening_brace: "{"
  - dockerAgent
    - opening_brace: "{"
    - bool: "true"
    - string > single_quoted
      - single_quote: "\'"
      - inner_single_str: "maven:3-alpine"
      - single_quote: "\'"
    - string > single_quoted
      - single_quote: "\'"
      - inner_single_str: "my-defined-label"
      - single_quote: "\'"
    - string > single_quoted
      - single_quote: "\'"
      - inner_single_str: "-v /tmp:/tmp"
      - single_quote: "\'"
    - string > single_quoted
      - single_quote: "\'"
      - inner_single_str: "https://myregistry.com/"
      - single_quote: "\'"
    - string > single_quoted
      - single_quote: "\'"
      - inner_single_str: "myPredefinedCredentialsInJenkins"
      - single_quote: "\'"
    - closing_brace: "}"
  - closing_brace: "}"
- stagesDecl
  - opening_brace: "{"
  - stage
    - string > single_quoted
      - single_quote: "\'"
      - inner_single_str: "Build"
      - single_quote: "\'"
    - opening_brace: "{"
    - stepsDecl
      - opening_brace: "{"
      - step > simple_step
        - IDENT: "sh"
        - args > string > single_quoted
          - single_quote: "\'"
          - inner_single_str: "make"
          - single_quote: "\'"
      - closing_brace: "}"
    - closing_brace: "}"
  - closing_brace: "}"
- closing_brace: "}"
- ending_junk: ""
- EOI: ""

Pest语法重点标注:

  • 当使用静默规则时,解析结果中将不会出现该规则字段。当解析下面规则时,解析结果中将不会存在silent,即parsed.as_rule() 中不会存在silent

    silent = _{ ... }
    
  • 当使用原子语法时,整个规则体将视为一个规则,如double_quoted = ${ (quote ~ inner_double_str ~ quote) },在解析时会将quote ~ inner_double_str ~ quote视为一个规则,而不是三个。这有利于获取一段完整的字符串。

    atomic = @{ ... }
    compound_atomic = ${ ... }
    

我尝试使用该项目解析jenkinsfile,但发现实现起来太过复杂,且rust和jenkinsFile的语法也是一言难尽。如下,当step中带括号和不带括号混用时会导致解析错误。

steps {
    echo 'test'
    dir('command') {
		sh "sh ./saas.sh ${params.channel} ${params.buildType} "
	}     
}

有精力的大神可以在此基础上实现解析JenkinsFile的功能。

参考

posted @ 2021-02-09 15:52  charlieroro  阅读(1258)  评论(0编辑  收藏  举报