
Let's say you are fetching JSON or a dictionary from something outside of Ansible. For example, perhaps you are using ansible.builtin.shell to run a homegrown script that returns JSON or a dictionary. For example, something like this.
---
- hosts: localhost
tasks:
- name: invoke homegrown script.py
ansible.builtin.shell: /path/to/homegrown/script.py
register: out
- ansible.builtin.debug:
var: out.stdout
Let's say the above playbook returns the following, which means that the homegrown script returned JSON or a dictionary.
ok: [localhost] => {
"out": {
\"result\": \"success\",
\"data\": \"hello world\",
}
}
The type_debug filter can be used to display the type of object that Ansible is registering.
- ansible.builtin.debug:
var: out.stdout | type_debug
Which should return something like this, in this scenario, almost always a str or dict.
ok: [localhost] => {
"out.stdout | type_debug": "str"
}
Depending on the flavor and version of Ansible that you are using, you may be able to print the values of the keys like this.
- ansible.builtin.debug:
var: out.stdout.result
- ansible.builtin.debug:
var: out.stdout.data
On the other hand, some flavors or versions of Ansible will require you to use the from_json filter because Ansible is detecting that the out variable is a string and not a dictionary, thus from_json is needed to convert the out variable from a string to a dictionary.
Similarly, to_json can be used to convert a dictionary to a string.
- ansible.builtin.debug:
var: (out.stdout | from_json).result
- ansible.builtin.debug:
var: (out.stdout | from_json).data
Did you find this article helpful?
If so, consider buying me a coffee over at