CODEWITHSUNDEEP
ansiblejinja2
airdata
airdata

Reputation: 641

Ansible multiple regexp when state

I'm trying to replace name in file based on the artifacts that ansible has collected. It's there some simple way to achieve this task with ONE lineinfile

Something like this, 

lineinfile:
    path: '/somefile.conf'
    regexp: '{{ item.regexp }}'
    line: '{{ item.line }}'
  with_items:
    - { regexp: '^DB_TYPE=', line: "DB_TYPE=mysql" } when: mysql.stat.exists
    - { regexp: '^DB_TYPE=', line: "DB_TYPE=oracle" } when: oracle.stat.exists

Upvotes: 1

Views: 228

Answers (2)

franklinsijo
franklinsijo

Reputation: 18270

This can be done with if ... else ... if

  lineinfile:
    path: '/somefile.conf'
    regexp: '^DB_TYPE='
    line: "{{ 'DB_TYPE=mysql' if mysql.stat.exists else 'DB_TYPE=oracle' if oracle.stat.exists }}"

Upvotes: 2

n7s
n7s

Reputation: 409

I believe when conditions are tied to a task, and thus you cannot have different conditions for different loops in the same task (happy to be corrected here).

This leaves you a couple options:

  1. Split this into multiple lineinfile tasks, which it sounds like you want to avoid.
  2. Make use of set_fact module. Not necessarily better, but cool to be aware of.

An example of the second option:

- name: set database type for mysql deployment
  set_fact:
    _db_type: mysql
  when: mysql.stat.exists

- name: set database type for oracle deployment
  set_fact:
    _db_type: oracle
  when: oracle.stat.exists

- name: something something
  lineinfile:
    path: '/somefile.conf'
    regexp: '^DB_TYPE='
    line: "DB_TYPE={{ _db_type }}"

I find this helpful if there are multiple tasks where you'll need to refer to either mysql or oracle (as an example).

Upvotes: 0

Related Questions