Skip to main content

apply_bump

Function apply_bump 

Source
pub fn apply_bump(
    ctx: &EffectCtx<'_>,
    bump: &BumpPlan,
    effective_date: &str,
) -> Result<BumpOutcome, BumpExecError>
Expand description

Apply the sealed bump inside ctx.repo_root (the coordinator’s clean checkout), committing the edits and returning the BumpOutcome. effective_date is the YYYY-MM-DD CHANGELOG date (freshly computed on a first run, or the journalled date on resume so it never re-dates).

The bump commit is created on the checkout’s detached HEAD; its object lives in the shared store, so the coordinator’s tagger (running against the real repo) can point the release tag at it within the same cut. Advancing the real repo’s branch to the bump commit is deliberately not done here (a best-effort pre-publish push would be an externally-visible effect before the build barrier, and a failed one leaves a split-brain remote) — it is a documented follow-up (release-rust-workspace-multicrate).

§Errors

BumpExecError on any failed edit, lockfile refresh, hook, post-hook validation, or git step — always before the build barrier, so nothing external happened.