Struct FileQuerySnapshotReader 

pub struct FileQuerySnapshotReader { /* private fields */ }

A long-lived read-only SQLite sidecar for latency-critical native views. Network rounds stay serialized on the mutable core owner, while atomic query snapshots use this independent connection and therefore never queue behind HTTP/WebSocket latency.

Implementations

impl FileQuerySnapshotReader

pub fn new(path: impl Into<String>) -> Self

pub fn snapshot_read( &mut self, request: &SnapshotReadRequest<'_>, ) -> Result<SnapshotRead, QueryReadFailure>

One atomic read over the shared primitive: every statement, coverage, catch-up, and delivery under one SAVEPOINT and one revision.

pub fn local_revision(&mut self) -> Result<String, QueryReadFailure>

The current durable local revision. A host can skip a read when it is unchanged; a corrupt or unreadable marker fails typed.

pub fn subscription_catchup( &mut self, id: &str, ) -> Result<SubscriptionCatchup, QueryReadFailure>

One subscription’s persisted catch-up state (see the §7.5 contract).

pub fn commit_delivery( &mut self, id: &str, ) -> Result<CommitDelivery, QueryReadFailure>

One client commit’s delivery status: pending, its retained outcome, or an explicit unknown (never held, or pruned by retention).

pub fn query_snapshot( &mut self, sql: &str, params: &[Value], coverage: &[WindowCoverage], ) -> Result<QuerySnapshot, QueryReadFailure>

Hosts report an owned failure, and the first success after it, to the owning core through SyncClient::record_query_read (§7.6).