流水线间分支名称传递与多分支触发阶段配置问题

第一步：修改first_pipeline.yml，传递分支名称并触发主流水线 #

先调整触发规则，让它能捕获所有目标分支的标签推送，同时把当前分支名称传递给main_pipeline：

trigger:
  tags:
    include:
      - '*'
  branches:
    include:
      - master
      - dev/*
      - deploy/*
      - feature/*
      - hotfix/*

variables:
  agent_pool: 'DEVAGENT'
  # 获取当前触发流水线的分支名称
  current_branch: $(Build.SourceBranchName)

pool:
  name: $(dev_pool)

# 添加阶段，手动触发主流水线并传递分支参数
stages:
- stage: TriggerMainPipeline
  jobs:
  - job: Trigger
    steps:
    - task: AzureCLI@2
      inputs:
        azureSubscription: '<你的Azure订阅名称>'
        scriptType: 'bash'
        scriptLocation: 'inlineScript'
        inlineScript: |
          az pipelines run \
            --name 'main_pipeline' \
            --project '<你的项目名称>' \
            --variables current_branch=$(current_branch)

如果不想用AzureCLI，也可以用专门的流水线触发任务：

trigger:
  tags:
    include:
      - '*'
  branches:
    include:
      - master
      - dev/*
      - deploy/*
      - feature/*
      - hotfix/*

variables:
  agent_pool: 'DEVAGENT'
  current_branch: $(Build.SourceBranchName)

pool:
  name: $(dev_pool)

resources:
  pipelines:
    - pipeline: main_pipeline
      source: 'main_pipeline'
      project: '<你的项目名称>'
      trigger: none # 避免自动循环触发

stages:
- stage: TriggerMain
  jobs:
  - job: Trigger
    steps:
    - task: PipelineTrigger@1
      inputs:
        pipeline: 'main_pipeline'
        project: '<你的项目名称>'
        parameters: 'current_branch=$(current_branch)'

第二步：修改main_pipeline.yml，根据分支控制阶段运行 #

在主流水线中接收传递来的分支变量，通过条件判断决定哪些阶段执行：

trigger: none
pr: none

resources:
  repositories:
    - repository: Grand
      type: git
      name: Automotive/Grand
      ref: master
      endpoint: Grand
    - repository: CICD
      type: git
      name: pipelines
      ref: master
  pipelines:
    - pipeline: Grand
      source: 'Grand'
      project: Automotive
      trigger: true

variables:
  - template: variables.yml
  # 声明接收从first_pipeline传递的分支变量
  - name: current_branch
    value: ${{ variables.current_branch }}

stages:
    - stage: DEV
      displayName: 'Development'
      # 运行条件：master分支，或dev/deploy开头的分支
      condition: or(eq(variables.current_branch, 'master'), startsWith(variables.current_branch, 'dev/'), startsWith(variables.current_branch, 'deploy/'))
      variables:
        - name: environment
          value: ${{ variables.env_dev}}
      pool:
        name: ${{ variables.agent_dev }}
      jobs:
        - deployment: DEV
          displayName: 'Deploying to DEV'
          # 剩余代码
  - stage: PRD
    displayName: 'Production Stage'
    # 运行条件：master分支，或feature/hotfix开头的分支
    condition: or(eq(variables.current_branch, 'master'), startsWith(variables.current_branch, 'feature/'), startsWith(variables.current_branch, 'hotfix/'))
    variables:
      - name: environment
        value: ${{ variables.prd }}
    pool:
      name: ${{ variables.agent_prd }}
    jobs:
      - deployment: PRDDeploy
        displayName: 'Deploying to PRD'
        # 剩余代码

核心逻辑说明 #

1. 分支传递：通过$(Build.SourceBranchName)自动获取触发流水线的分支名，再通过触发任务将变量传递给主流水线。
2. 阶段控制：给每个阶段添加condition条件，根据分支名匹配规则决定是否执行：
   • DEV阶段：仅master、dev/*、deploy/*分支运行
   • PRD阶段：仅master、feature/*、hotfix/*分支运行
3. 触发范围：调整first_pipeline的trigger规则，确保所有目标分支的标签推送都能触发流水线。

内容的提问来源于stack exchange，提问作者scrapkowe