Pileup: work the strongest caller first (opt-in) + SNR on caller-queue chips

[patrickrb]

What & why

When you run a pileup — calling CQ and getting several answers in the same slot during a DXpedition, POTA activation, or contest — FT8AF queues the extra callers and works them one after another. Until now that queue was strict FIFO: the station that called earliest was always worked next, regardless of how well you were copying it. On a busy run that means strong, easy-to-complete stations sit waiting behind marginal ones, and you burn cycles on exchanges that may never finish.

This adds an opt-in operating option that lets the app work the strongest waiting caller next instead, so you complete the best-copy contacts first and keep the run moving.

What changed

• New Auto-Sequence setting — "Work strongest caller first" (Settings → CAT & Transmission → Auto-Sequence). Off by default, preserving the historic first-heard/FIFO order for anyone who prefers it.
• When on, the pileup auto-dequeue picks the highest-SNR waiting caller next. Ties break toward the earliest-queued caller, so equal signals still go in the order they called (fair, deterministic).
• The caller-queue chips in the active-QSO panel now show each waiting caller's SNR and sort strongest-first to match, so you can see at a glance who the engine will work next — and still tap any chip to pick someone else.

Implementation

• GeneralVariables.pileupStrongestFirst (persisted via DB key pileupStrongestFirst, default off; volatile — written on the UI thread, read on the transmit thread).
• Selection policy extracted to a pure, side-effect-free CallerQueueOrdering.pickNextIndex(queue, strongestFirst); FT8TransmitSignal.dequeueNextCaller() now uses it instead of remove(0).
• Display helpers orderCallerQueue / callerSnrLabel extracted from ActiveQsoPanel for testing.
• No change to the FT8 protocol, decoder, or waveform — this only reorders which already-queued caller is answered next.

Tests

• CallerQueueOrderingTest (Java, pure): FIFO vs. strongest-first selection, negative-SNR handling, tie-break to earliest, full-drain ordering, no-mutation, empty/null guards.
• CallerQueueDisplayTest (Kotlin, pure): display ordering mirrors the engine policy, stable-for-ties, SNR label formatting.

Verification

• ./gradlew :app:testDebugUnitTest — full suite green.
• ./gradlew :app:assembleDebug — APK builds clean.

[codecov]

Codecov Report

❌ Patch coverage is 8.33333% with 22 lines in your changes missing coverage. Please review.
✅ Project coverage is 36.68%. Comparing base (f8fc1f6) to head (a788b05).

Files with missing lines	Patch %	Lines
...o/ks3ckc/ft8af/ui/settings/TransmissionSettings.kt	0.00%	12 Missing ⚠️
...radio/ks3ckc/ft8af/ui/components/ActiveQsoPanel.kt	16.66%	10 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##                dev     #692      +/-   ##
============================================
- Coverage     36.70%   36.68%   -0.03%     
  Complexity      211      211              
============================================
  Files           222      222              
  Lines         27479    27500      +21     
  Branches       3395     3398       +3     
============================================
+ Hits          10087    10089       +2     
- Misses        17147    17166      +19     
  Partials        245      245              

Flag	Coverage Δ
android	15.78% <8.33%> (-0.01%)	⬇️
native	9.93% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...radio/ks3ckc/ft8af/ui/components/ActiveQsoPanel.kt	9.69% <16.66%> (+0.29%)	⬆️
...o/ks3ckc/ft8af/ui/settings/TransmissionSettings.kt	0.86% <0.00%> (-0.03%)	⬇️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.