Struct NewRequestRow 

pub struct NewRequestRow<'a> {

    pub id: Uuid,
    pub name: &'a str,
    pub description: Option<&'a str>,
    pub schedule_date: Option<DateTime<Utc>>,
    pub schedule_end: Option<DateTime<Utc>>,
    pub duration: Decimal,
    pub owner_user_id: Option<Uuid>,
    pub user_id: Option<Uuid>,
    pub asset_id: Option<Uuid>,
    pub stage_id: Uuid,
    pub kanban_state: &'a str,
    pub priority: &'a str,
    pub maintenance_type: &'a str,
    pub recurring: bool,
    pub repeat_interval: i32,
    pub repeat_unit: &'a str,
    pub repeat_type: &'a str,
    pub repeat_until: Option<NaiveDate>,
    pub successor_of_request_id: Option<Uuid>,
}

The exact row a request create (or successor spawn) writes.

Mirrors the raw column shape rather than the MaintenanceRequest entity: the enum columns bind as &str and are cast at the DB ($n::request_kanban_state etc.), so a bad value fails as a DB error rather than a deserialize panic.

Fields

id: Uuid

name: &'a str

description: Option<&'a str>

schedule_date: Option<DateTime<Utc>>

schedule_end: Option<DateTime<Utc>>

duration: Decimal

owner_user_id: Option<Uuid>

user_id: Option<Uuid>

asset_id: Option<Uuid>

stage_id: Uuid

kanban_state: &'a str

priority: &'a str

maintenance_type: &'a str

recurring: bool

repeat_interval: i32

repeat_unit: &'a str

repeat_type: &'a str

repeat_until: Option<NaiveDate>

successor_of_request_id: Option<Uuid>

Set only on the clone-on-done successor (the backref to its source).