Function incremental_import_skips

atomic-cli/src/commands/git/parallel.rs:206–220

Whether an incremental import would skip the git commit identified by `sha` with the given commit `message`, for a target view that already contains `imported_shas` and `known_states`. This mirrors exactly what the real importer does — the SHA skip in `collect_commit_oids` (already-imported commits) plus the self-push skip in `phase2_write` (commits whose carried view state is already present). I

(
    sha: &str,
    message: &str,
    imported_shas: &HashSet<String>,
    target_view: &str,
    known_states: &HashSet<Merkle>,
)

/// the single source of truth used by `atomic git import --dry-run` to forecast
/// the real import count. Callers must only invoke it for incremental imports.
pub(crate) fn incremental_import_skips(
    sha: &str,
    message: &str,
    imported_shas: &HashSet<String>,
    target_view: &str,
    known_states: &HashSet<Merkle>,
) -> bool {
    if imported_shas.contains(sha) {
        return true;
    }
    match parse_push_trailer(message) {
        Some(trailer) => self_push_state_known(&trailer, target_view, known_states),
        None => false,
    }
}

/// Metadata extracted from a git commit.
#[derive(Debug, Clone)]