Add support for image name history

[saschagrunert]

We leverage the containers/storage image history tracking feature to
show the previously used image names when running:
podman images --history

Closes #4566

When building two images locally in sequence, then the output looks like this now:

> sudo ./bin/podman images --history
REPOSITORY       TAG      IMAGE ID       CREATED          SIZE      HISTORY
localhost/test   latest   2ba60be0975e   1 second ago     2.07 kB   <none>
<none>           <none>   79af25cf9b71   21 seconds ago   2.07 kB   localhost/test:latest

We skip the first history entry in the first line, because it is already part of the first image name.

[vrothberg]

Just a couple of nits. Needs man page changes and bash completion. Other than that, LGTM

[rhatdan]

I think you need to update pkg/varlinkapi/images.go

[vrothberg]

Just one last nit. I really like that we don't need to change the API but do post-processing in the client 👍

[rhatdan]

/approve
/lgtm

[openshift-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: rhatdan, saschagrunert

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [rhatdan]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[edsantiago]

@saschagrunert the new test in 010-images.bats fails upon rerun because the history component is no longer empty. I'm willing to open a PR to fix that -- there are a couple other issues I want to address there -- but I'm not sure how to approach it. Is there a history-clear mechanism? Can you suggest a way to tackle this? TIA.

[saschagrunert]

@saschagrunert the new test in 010-images.bats fails upon rerun because the history component is no longer empty. I'm willing to open a PR to fix that -- there are a couple other issues I want to address there -- but I'm not sure how to approach it. Is there a history-clear mechanism? Can you suggest a way to tackle this? TIA.

Yeah, do you have a test output at hand somewhere in the CI where I can have a look at it?

[edsantiago]

It's on my own system, not CI (that's why I'm seeing this: CI only runs once, I'm running repeated iterations). This, I believe, demonstrates the problem:

# podman rmi -a                     ! get a fresh start
# podman pull -q alpine
965ea09ff2ebd2b9eeec88cd822ce156f6674c7e99be082c7efac3c62f3ff652
# podman images --format '{{.History}}'
                            <<<< empty, as expected
# podman tag alpine foo
# podman rmi foo
Untagged: localhost/foo:latest
# podman images --format '{{.History}}'
localhost/foo:latest        <<<<<<< no longer empty; and this persists to the next BATS run

[saschagrunert]

Ah I think I got it. Is it because the cleanup function does not remove the image completely after the run? Might this solve the issue already on recurring runs?

[edsantiago]

Is it because the cleanup function does not remove the image completely

Possibly, but that's intentional: it is critical that these tests pull only once, because quay.io is so unreliable: we need to minimize errors due to flakes, and I choose to do that by having one sane, trusted, unmodified image cached locally. Is there a way to clear/reset this tag history?

[saschagrunert]

No not now, but we could build the test images via an empty Dockerfile ad-hoc, maybe via a little helper function to prepare test images.