docs: fix pull --policy value from newer to ifnewer

[locker95]

Summary

The buildah pull man page documented --policy=newer, but the CLI only accepts ifnewer via define.PolicyMap (see cmd/buildah/pull.go and define/pull.go). Using --policy newer fails with:

Error: unsupported pull policy "newer"

Update docs/buildah-pull.1.md so the documented values match the CLI: always, missing, never, and ifnewer.

Note: buildah from / buildah build --pull docs that mention newer are intentionally left unchanged; those commands parse both newer and ifnewer via parse.PullPolicyFromOptions.

Fixes #6992

Test plan

• Confirmed CLI help: --policy string missing, always, ifnewer, or never.
• Confirmed define.PolicyMap keys are only missing, always, never, ifnewer
• Man page now documents ifnewer consistently in the flag synopsis and bullet list

[packit-as-a-service]

Ephemeral COPR build failed. @containers/packit-build please check.

[locker95]

The Packit COPR failures look infrastructure-side rather than related to this docs-only change:

Backend process error: Giving up waiting for copr_base repository, please try to manually regenerate the DNF repository

SRPM built successfully; the chroot never got past waiting for repomd.xml (no builder-live.log / mock log was produced). Other recent buildah PRs (e.g. #6961) show rpm-build green.

Change is only docs/buildah-pull.1.md (newer → ifnewer to match define.PolicyMap).

/packit build

[locker95]

/packit build

[nalind]

We do want to accept aliases for some of these, though. Pitching #7000 to do that.

[locker95]

@nalind Makes sense — accepting aliases on the CLI side is the better fix. Glad #7000 covers that.

This PR only corrects the man page so it documents the canonical ifnewer value from define.PolicyMap (the previous newer string was wrong for the shipped binary). Fine to land independently of, or after, #7000 as you prefer.

[nalind]

On that note, can we describe the two as being equivalent? IIRC the set of values we document here is meant to be in line with podman pull, possibly back in #2754.

[leonardomoreira00]

Just a note: it's a little frustrating to spend time using and understanding the code, opening an issue, and working on a solution, only to later see someone use AI to duplicate the pull request.

Ofc, this is just a typo, but it is happening even to bigger PRs.

My friends who are more involved in open source complain about this all the time.
I feel sorry for the maintainers... what a time to be alive

[locker95]

@nalind Good call — updated the man page to state that buildah's ifnewer is the same policy podman documents as newer (podman pull --policy=newer), so the documented value sets stay aligned.

Agree on accepting aliases in the CLI map; happy to leave that to #7000 rather than expanding PolicyMap in this docs-only PR.

[nalind]

Hold up, why did this start as an exact duplicate of #6993?

[locker95]

@nalind yeah, bad timing on my side. #6993 already had the ifnewer fix when i opened this — i should've checked and pointed at that instead of stacking another PR on the same typo.

closing this in favor of #6993. sorry for the noise @leonardomoreira00.