Struct SeedBinder 

pub struct SeedBinder { /* private fields */ }

Verifies and remembers bindings.

Implementations

impl SeedBinder

pub async fn bind_v2( &self, b: &StewardBind<'_>, ) -> Result<BindingRecord, BindError>

Steward-profile bind of a v2 record. See the module docs.

pub fn unbind_v2( &self, signed: &SignedBinding, store: &CheckoutGrantStore, ) -> Result<UnbindOutcome, BindError>

Withdraw the binding with an operator-signed unbound record. Needs no Seed (it may be gone). Any operator-Admin node may do it. If the store could not save it, the unbind is still applied in memory (it only restricts), the event is chained, and the outcome says the save is pending (the store’s tick retries it).

impl SeedBinder

pub fn new(operators: Vec<[u8; 32]>, chain: Arc<ChainManager>) -> Self

A binder trusting operators (pinned operator public keys).

pub fn with_state_file( self, path: impl Into<PathBuf>, ) -> Result<Self, BindError>

Persist each device’s latest bound_at to path (mode 0600, atomic) and restore it now, so a replayed or older record is still refused after a restart. A missing file starts empty; a malformed, oversized or non-regular one is refused (nothing is overwritten).

pub fn with_max_age_secs(self, secs: u64) -> Self

Accept records for at most secs after bound_at.

pub fn device_of(&self, node_id: &str) -> Option<String>

The device id bound to node_id, if any.

pub async fn bind( &self, signed: &SignedBind, rt: &SeedApiRuntime, now: u64, ) -> Result<Binding, BindError>

Verify signed for rt at now (unix seconds) and chain the outcome (workload.node.bind, or workload.refuse on any failure).