如何基于多条件为Ansible变量定义多种取值场景

看起来你想要在Ansible的变量文件中，根据外部传入的role和environment变量，动态生成patch_plan的值——核心是给patch_plan_week_and_day匹配对应的场景值。变量文件本身不支持直接用when条件，不过我们有几种优雅的实现方式，完全能满足你的需求（包括14种场景的维护）。

首先先明确你的基础变量和需求：

你当前的变量定义是：

patch_plan: 'foo-{{ patch_plan_week_and_day }}-bar'

需要匹配的核心场景示例：

• 当role = 'master' 且 environment = 'srvb' → patch_plan_week_and_day = 'Week1_Monday'
• 当role != 'master' 且 environment = 'srvb' → patch_plan_week_and_day = 'Week1_Tuesday'
• 当role = 'slave' 且 environment = 'pro' → patch_plan_week_and_day = 'Week3_Wednesday'

你期望的类when逻辑写法虽然在纯变量文件里不能直接用，但我们可以用以下几种方式实现：

方案1：字典映射（推荐，适合多场景维护） #

这种方式把所有场景映射到一个字典里，然后用Jinja2条件逻辑去匹配，可读性和维护性都很强，尤其适合14种场景的情况：

# 第一步：定义所有场景的映射字典，键可以自定义，清晰即可
patch_plan_scenarios:
  master_srvb: 'Week1_Monday'
  non_master_srvb: 'Week1_Tuesday'
  slave_pro: 'Week3_Wednesday'
  # 在这里继续添加剩下的11种场景，比如：
  # worker_pro: 'Week2_Thursday'
  # master_pro: 'Week1_Sunday'
  # ...

# 第二步：动态计算patch_plan_week_and_day的值
patch_plan_week_and_day: >-
  {% if role == 'master' and environment == 'srvb' %}
    {{ patch_plan_scenarios.master_srvb }}
  {% elif environment == 'srvb' and role != 'master' %}
    {{ patch_plan_scenarios.non_master_srvb }}
  {% elif role == 'slave' and environment == 'pro' %}
    {{ patch_plan_scenarios.slave_pro }}
  {% else %}
    # 可以设置一个默认值，比如空字符串或者通用值
    'Default_Plan'
  {% endif %}

# 最终生成的patch_plan
patch_plan: 'foo-{{ patch_plan_week_and_day }}-bar'

如果很多场景是基于environment+role的组合，还可以优化字典的键为"{role}_{environment}"，简化条件判断：

patch_plan_scenarios:
  master_srvb: 'Week1_Monday'
  slave_srvb: 'Week1_Tuesday'
  worker_srvb: 'Week1_Tuesday'
  slave_pro: 'Week3_Wednesday'
  # 剩余场景...

patch_plan_week_and_day: >-
  {% set combo_key = f"{role}_{environment}" %}
  {% if combo_key in patch_plan_scenarios %}
    {{ patch_plan_scenarios[combo_key] }}
  {% elif environment == 'srvb' and role != 'master' %}
    {{ patch_plan_scenarios.slave_srvb }} # 统一匹配非master的srvb场景
  {% else %}
    'Default_Plan'
  {% endif %}

patch_plan: 'foo-{{ patch_plan_week_and_day }}-bar'

方案2：在Playbook中用set_fact（接近你期望的when写法） #

如果你更习惯用when条件的写法，可以在Playbook里通过set_fact模块来逐个设置变量，这种方式逻辑直观，和你想要的代码结构几乎一致：

- name: Configure patch plan based on role and environment
  hosts: your_target_hosts
  vars:
    # 基础变量定义
    patch_plan: 'foo-{{ patch_plan_week_and_day }}-bar'
  tasks:
    - name: Set patch plan for master+srvb
      set_fact:
        patch_plan_week_and_day: 'Week1_Monday'
      when: role == 'master' and environment == 'srvb'

    - name: Set patch plan for non-master+srvb
      set_fact:
        patch_plan_week_and_day: 'Week1_Tuesday'
      when: role != 'master' and environment == 'srvb'

    - name: Set patch plan for slave+pro
      set_fact:
        patch_plan_week_and_day: 'Week3_Wednesday'
      when: role == 'slave' and environment == 'pro'

    # 继续添加剩下的11个场景的set_fact任务...

    - name: Print final patch plan
      debug:
        msg: "Final patch plan: {{ patch_plan }}"

如果14个场景写起来繁琐，还可以把场景数据整理成列表，用循环来处理：

- name: Configure patch plan with loop
  hosts: your_target_hosts
  vars:
    patch_plan: 'foo-{{ patch_plan_week_and_day }}-bar'
    patch_plan_scenarios:
      - value: 'Week1_Monday'
        condition: "role == 'master' and environment == 'srvb'"
      - value: 'Week1_Tuesday'
        condition: "role != 'master' and environment == 'srvb'"
      - value: 'Week3_Wednesday'
        condition: "role == 'slave' and environment == 'pro'"
      # 剩余场景...
  tasks:
    - name: Set patch plan from scenarios
      set_fact:
        patch_plan_week_and_day: "{{ item.value }}"
      when: item.condition | bool
      loop: "{{ patch_plan_scenarios }}"

    - name: Print final patch plan
      debug:
        msg: "Final patch plan: {{ patch_plan }}"

方案3：单行Jinja2条件表达式（简洁但适合少场景） #

如果场景较少，也可以直接用单行的Jinja2条件表达式写在变量里，不过当场景到14种时，可读性会下降：

patch_plan_week_and_day: >-
  {{ 'Week1_Monday' if role == 'master' and environment == 'srvb' else
     'Week1_Tuesday' if role != 'master' and environment == 'srvb' else
     'Week3_Wednesday' if role == 'slave' and environment == 'pro' else
     'Default_Plan' }}

patch_plan: 'foo-{{ patch_plan_week_and_day }}-bar'

综合来看，方案1的字典映射是最适合你的，因为14种场景需要良好的可维护性，字典+条件逻辑的组合既能清晰管理所有场景，又能在变量文件中直接完成动态赋值。

内容的提问来源于stack exchange，提问作者GBT55