Ansible replace module duplicates replace values

Question

PROBLEM: Ansible 2.9 will duplicate entries in the "replace: " field of the replace module.

FILE TO CHANGE (/etc/netplan/50-cloud-init.yaml):

network:
ethernets:
    ens160:
        addresses:
        - 10.10.8.112/24
        gateway4: 10.10.8.1
        nameservers:
            addresses:
            - 10.10.8.15
version: 2

CODE:

- name: Check yaml has correct DNS values
  hosts: guest
  gather_facts: yes
  become: yes

  tasks:
          - name: Adjust yaml file
            replace:
              path: /etc/netplan/50-cloud-init.yaml
              after: '                addresses:'
              before: '    version: 2'
              regexp: '^(.+)$'
              replace: '                - 10.10.8.110\n                - 10.10.8.111\n                search:\n                - server.com\n            optional: true'

RESULTS:

• If the original file only has one entry say "- 10.10.8.110", the sections will be replaced inbetween the "after" and "before". However if run again, expansible will duplicate the entries.

File contents will look like this after running the script again:

network:
ethernets:
    ens160:
        addresses:
        - 10.10.8.112/24
        gateway4: 10.10.8.1
        nameservers:
            addresses:
            - 10.10.8.110
            - 10.10.8.111
            search:
            - server.com
        optional: true
            - 10.10.8.110
            - 10.10.8.111
            search:
            - server.com
        optional: true
version: 2

Believe this has to do with idempotence, and can possibly be fixed with the regex. However I am not sure how to do this. All I want to do is to ALWAYS replace all values between after and before.

How can I accomplish this?

Answer 1

I would not count on before and after but build a regexp that matches exactly what I'm looking for and replace with back-references. Something like the following (not sexy but functional)

- name: Adjust yaml file
  replace:
    path: /etc/netplan/50-cloud-init.yaml
    regexp: >-
      (^\s*nameservers:\n\s*addresses:\n)[\S\s]*(version: 2)$
    replace: |-
      \g<1>           - 10.10.8.110
                  - 10.10.8.111
                  search:
                  - server.com
              optional: true\g<2>

Meanwhile, note that since your file is yaml, there are other ways to modify the data (combine objects, etc...) that might be simpler.

Answer 2

The Code that you used was right, you just need to remove or before or after, if you use them at the same time the text will be inserted 2 times.

You need to do something like that (remove before):

    - name: Check yaml has correct DNS values
      hosts: guest
      gather_facts: yes
      become: yes

      tasks:

         - name: Adjust yaml file
           replace:
             path: /etc/netplan/50-cloud-init.yaml
             after: '                addresses:'
             regexp: '^(.+)$'
             replace: '                - 10.10.8.110\n                - 10.10.8.111\n                search:\n                - server.com\n            optional: true'