Struct MaintenanceRequestWriteService 

pub struct MaintenanceRequestWriteService { /* private fields */ }

Implementations

impl MaintenanceRequestWriteService

pub fn new(pool: PgPool) -> Self

pub async fn create_request( &self, r: NewMaintenanceRequest, ) -> Result<Uuid, RequestWriteError>

File a request. The stage defaults to the first visible stage (the shared seed’s “New Request”); entering a done stage stamps close_date via the trigger.

pub async fn update_request( &self, request_id: Uuid, u: MaintenanceRequestUpdate, ) -> Result<(), RequestWriteError>

Change a request’s descriptive/recurrence fields. The guards run on the MERGED view (old row overlaid with the update), so — the widened G-MT2 fire-rule — moving schedule_date past an existing schedule_end is caught here and by the table CHECK.

pub async fn transition_request( &self, request_id: Uuid, target_stage_id: Uuid, kanban_state_override: Option<RequestKanbanState>, events: &dyn MaintenanceEventSink, ) -> Result<TransitionOutcome, RequestWriteError>

Move a request to another stage — the only sanctioned way a request’s stage changes.

Inside one transaction: set the managed-transition marker, CAS the stage move (the G-MT5 trigger applies the kanban reset + close_date on the same statement), and — when a non-done request CLOSES onto a done stage and is preventive+recurring — spawn the successor (claim the slot first; at most one successor per source). Stages the outbox events and commits; the successor exists iff the close committed.