The self-push rule in one place: a commit created by `atomic git push` carries the target view's Merkle state, so if that view already has the state, the commit adds nothing. Shared by [`should_skip_self_push`] and [`incremental_import_skips`] so the real import and the `--dry-run` forecast can never disagree on it.
(
trailer: &PushTrailer,
target_view: &str,
known_states: &HashSet<Merkle>,
)
| 187 | /// [`incremental_import_skips`] so the real import and the `--dry-run` forecast |
| 188 | /// can never disagree on it. |
| 189 | fn self_push_state_known( |
| 190 | trailer: &PushTrailer, |
| 191 | target_view: &str, |
| 192 | known_states: &HashSet<Merkle>, |
| 193 | ) -> bool { |
| 194 | trailer.view == target_view && known_states.contains(&trailer.state) |
| 195 | } |
| 196 | |
| 197 | /// Whether an incremental import would skip the git commit identified by `sha` |
| 198 | /// with the given commit `message`, for a target view that already contains |
no test coverage detected