feat(flow): add whole-utterance batch ASR fallback (P1)

- Accumulate utterance PCM in FlowContinuousCapture for batch retry
- Run full-utterance transcribeChunk when stitched final lags partial
- Refactor Mac MLX tail drain to shared FlowUtteranceEndCoordinator
- Add FlowUtterancePCMStore, UtteranceBatchFallbackPolicy, and tests

Co-authored-by: Rocky <hkgood@users.noreply.github.com>
This commit is contained in:
Cursor Agent
2026-07-26 10:15:31 +00:00
parent c7ee891a90
commit f0240c3088
9 changed files with 256 additions and 3 deletions
@@ -0,0 +1,46 @@
// FlowUtterancePCMStore.swift
// OSGKeyboard · Shared
//
// Thread-safe rolling buffer of 16 kHz mono utterance PCM for whole-utterance
// batch ASR fallback when pipelined chunking drops weak tail segments.
import Foundation
public final class FlowUtterancePCMStore: @unchecked Sendable {
private let lock = OSAllocatedUnfairLock()
private var samples: [Float] = []
private let maxSampleCount: Int
public init(maxSampleCount: Int) {
self.maxSampleCount = max(1, maxSampleCount)
}
public func reset() {
lock.withLock {
samples.removeAll(keepingCapacity: false)
}
}
public func append(_ chunk: [Float]) {
guard !chunk.isEmpty else { return }
lock.withLock {
samples.append(contentsOf: chunk)
if samples.count > maxSampleCount {
samples.removeFirst(samples.count - maxSampleCount)
}
}
}
public var sampleCount: Int {
lock.withLock { samples.count }
}
/// Returns accumulated samples and clears the store.
public func consume() -> [Float] {
lock.withLock {
let out = samples
samples.removeAll(keepingCapacity: false)
return out
}
}
}