合并云 yaml AWS 云形成模板

Question

我有多个文件，其中包含我的云形成template.yaml 文件。例如

base.yaml

AWSTemplateFormatVersion: 2010-09-09
Transform: AWS::Serverless-2016-10-31

Conditions:
  conFunctionInVPC: !And
    - !Not [!Equals [!Join ["", !Ref parFunctionSubnets], ""]]
    - !Not [!Equals [!Join ["", !Ref parFunctionSecurityGroups], ""]]
  conFunctionNotInVPC: !Not [!Condition conFunctionInVPC]
  conAwsXRaySdkLayer: !Not [!Equals [!Ref parAwsXRaySdkLayerArn, ""]]

应用程序.yaml

Description: >
  test
  A short description of the function purpose

我想生成最终的 yaml 文件。我尝试了一个简单的合并功能

with open('base.yaml') as f:
    base = yaml.load(f)
with open('application.yaml') as f:
    application = yaml.load(fp)
template = merge(base, template) # my own function not important here
yaml.dump(template, open('template.yaml', 'w'))

但是我得到了错误：

yaml.constructor.ConstructorError: 无法确定标签 '!And' 的构造函数

如何告诉 yaml 让节点保持简单？并且在dump的时候还能再次输出吗？

我试过了

yaml.add_multi_constructor('!', lambda loader, suffix, node: node)

但是当我转储文件时，我会在我的 yaml 中为每个 !ref 获取 SequenceNode 对象

例如：

conFunctionInVPC: !!python/object:yaml.nodes.SequenceNode
    end_mark: !!python/object:yaml.error.Mark
      buffer: null
      column: 2
      index: 721
      line: 24
      name: template/template.yaml
      pointer: null
    flow_style: false
    start_mark: !!python/object:yaml.error.Mark
      buffer: null
      column: 20
      index: 581
      line: 21
      name: template/template.yaml
      pointer: null
    tag: '!And'
    value:

Answer 1

看着YAML Error: could not determine a constructor for the tag

它似乎可以使用RoundTrip 和ruamel

template = load(f, Loader=RoundTripLoader)
yaml.dump(template, f, Dumper=RoundTripDumper)