Ansible Pluck Filter

This borrows heavily from this Ansible custom Jinja2 filters blog post by Jordan Bach from May 2016.

You can use this filter to easily extract a single key from a dictionary.

  - { name: "web", cluster: "gamma", url: 'http://www.example.com' }
  - { name: "worker", cluster: "alpha" }
  - { name: "relay", cluster: "omega" }

Usage

debug:
  var: "{{ your_dictionary | pluck('name') }}"

I've modified the filter so that it can cope with keys that are missing from some entries. For example, in the above dictionary, if you pluck 'url' you will get a list with a single item.

Returns

A list. (Run the test playbook to see some debug output.)

Installation:

• Copy pluck.py to your ansible/filter_plugins directory (if directory does not exist, create it)
• Run the following playbook to test it works:

ansible-playbook pluck-test.yml

Tested with:

• Ansible 2.5.0
• Python 2.7.14

Alternatives built-in to Ansible:

 your_dictionary | map(attribute='name') | list

In this case, where a key doesn't exist in all records, the output will look like this:

"urls": "[u'http://example.com', Undefined, Undefined]", 

Jinja Documentation