--column actually filters rows for openstack token issue

[Matthew Edmonds]

The --column actually acts on rows rather than columns for openstack token issue:

# openstack token issue -c project_id
+------------+----------------------------------+
| Field | Value |
+------------+----------------------------------+
| project_id | f866b93adc8247328d372367f448ed33 |
+------------+----------------------------------+

This is confusing and inconsistent from other commands, e.g. openstack project list:

# openstack project list -c ID
+----------------------------------+
| ID |
+----------------------------------+
| 35d7501ec6794eb6b13d2d897c603daf |
| c743108159c7483b8063a10c02d4a586 |
| f866b93adc8247328d372367f448ed33 |
+----------------------------------+

The table returned for openstack token issue should be structured as columns rather than as rows, like other subcommands.

[Karthik (kali-karthik)]

So u mean to say project ID should be on a column and Field Value rows should not be present?

[Steve Martinelli (stevemar)]

This is cliff's documented behaviour. When a single item it uses cliff's ShowOne class [1], which we use for create/show (and in this case, issue). When listing multiple items, the Lister class [2] is used.

Think of -c as a way to get either a column when listing lots of value or a field value when the command returns a single value.

Do you think the documentation / wording needs to be improved?

[1] https://github.com/openstack/cliff/blob/master/cliff/show.py
[2] https://github.com/openstack/cliff/blob/master/cliff/lister.py