v0.2.6-alpha.3

Swift: Inspect

inspect() answers the same five-field snapshot KizunaSyncInspector.snapshot() reads, decoded as a plain dictionary rather than the typed KizunaSyncInspectorSnapshot. Most apps call Inspector instead; this method is the one the typed reading is built on.

Examples#

Basic#

// TodoApp/TodoListView.swift (excerpt)
import KizunaSync

let raw = try await kizunasync.inspect()
let depth = raw["depth"] as? Int ?? 0

Parameters#

Name	Type	Required	Description
`(none)`	—	—	This method takes no arguments.

Returns#

[String: Any], decoded from the engine's JSON, with these keys.

Name	Type	Required	Description
`queued`	`[[String: Any]]`	—	The head of the outbox, oldest first.
`depth`	`Int`	—	The same count Outbox depth returns.
`last_mutation_id`	`String?`	—	The exactly-once push watermark: the mutation id of the last entry the server confirmed applied, not the last mutation queued. `nil` until a push lands.
`cursor`	`String`	—	The durable pull cursor, also on Checkpoint.
`client_id`	`String`	—	The client's current registered identity. It changes after Reset.

Errors#

Code	Condition
`ENGINE_UNAVAILABLE`	The bridge answered with bytes that do not decode to a JSON object.
`ENGINE_UNAVAILABLE`	The client has no engine because Initializing has not run.

Notes#

last_mutation_id is the watermark a completed push advances, not a record of what is queued. A device with ten mutations waiting in the outbox and none pushed answers nil here even though depth reads 10. Inspector wraps this call, adds the bounded verdict ring, and answers a typed KizunaSyncInspectorSnapshot. Reach for inspect() directly only when the raw dictionary is what the call site wants.

← Host scheduler Inspector →

Found a problem? Edit this page on GitHub