Add loop antenna controls

[rfoust]

Summary

• add LoopA/LoopB support to panadapter status and ANT overlay controls
• show loop controls only for radios with FlexLib loop capabilities
• keep loop selections mutually exclusive and disable selection unless RX ANT is ANT1 or ANT2

Verification

• cmake --build build --parallel 8
• ctest --test-dir build -R model_capabilities_test --output-on-failure
• git diff --check
• merge-tree checks against PR 2860, PR 2862, and both combined

[Copilot]

Pull request overview

Adds LoopA/LoopB awareness to panadapter state and the spectrum ANT overlay, using model capabilities to expose loop controls only on supported radios.

Changes:

• Extends ModelCapabilities and tests with LoopA/LoopB flags.
• Tracks panadapter RX antenna and loop state from status updates.
• Adds loop controls to the spectrum overlay and wires commands through MainWindow.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
tests/model_capabilities_test.cpp	Adds LoopA/LoopB expectations to capability tests.
src/models/PanadapterModel.h	Adds RX antenna and loop state accessors/signals/storage.
src/models/PanadapterModel.cpp	Parses rxant, loopa, and loopb from panadapter status.
src/models/ModelCapabilities.h	Adds loop capability fields.
src/models/ModelCapabilities.cpp	Populates loop capability table per model.
src/gui/SpectrumOverlayMenu.h	Adds loop UI API, signals, and panadapter wiring state.
src/gui/SpectrumOverlayMenu.cpp	Adds loop buttons, visibility gating, and panadapter RX antenna synchronization.
src/gui/MainWindow.cpp	Wires overlay pan/model/capabilities and sends loop commands.

Comments suppressed due to low confidence (1)

src/gui/SpectrumOverlayMenu.cpp:384

• This mirrors the LoopA switching path: unchecking LoopA under QSignalBlocker suppresses loopAToggled(false), so switching from LoopA to LoopB only sends loopb=1 and does not explicitly clear LoopA on the radio. Emit/send the complete state transition so the radio and UI stay mutually exclusive.

        if (on && m_loopABtn && m_loopABtn->isChecked()) {
            QSignalBlocker blocker(m_loopABtn);
            m_loopABtn->setChecked(false);
        }
        emit loopBToggled(on);

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[rfoust]

Fixed in f281fc7 by mirroring the normal wirePanadapter initialization in the duplicate-pan path: the overlay now gets setPanId, setRadioModel, and setRadioCapabilities before that path returns. Verified with cmake --build build --parallel 8 and ctest --test-dir build -R model_capabilities_test --output-on-failure.

[rfoust]

Fixed in f281fc7. When switching LoopA/LoopB, the UI still blocks the opposite button signal while updating the check state, but now manually emits the opposite loop false signal so MainWindow sends loopb=0 or loopa=0 before the selected loop command. Verified with cmake --build build --parallel 8 and ctest --test-dir build -R model_capabilities_test --output-on-failure.

[aethersdr-agent]

Looks good — nicely scoped feature with capability gating, RAII connection management via QPointer<PanadapterModel>, and the test table extended to cover the two new flags. Both Copilot findings are addressed in f281fc7:

• The duplicate-panadapter path in MainWindow.cpp now mirrors wirePanadapter() by calling setPanId/setRadioModel/setRadioCapabilities before the early return, so loop button visibility is seeded correctly.
• The loop-toggle mutex in SpectrumOverlayMenu.cpp now manually emits loopBToggled(false) / loopAToggled(false) when unchecking the opposite button under QSignalBlocker, so the radio receives the complete desired state.

A couple of minor notes (non-blocking):

1. Implicit mutex in PanadapterModel::applyPanStatus (PanadapterModel.cpp +97): when the radio reports loopa=1, the parser proactively clears m_loopB (and vice versa) even if the status frame doesn't contain loopb. In practice FlexLib emits both keys, so this is defensive and harmless — but if a future firmware ever reported loopa=1 standalone with loopb unchanged-but-true, the client would silently drop the LoopB state without a radio echo. Worth a brief comment noting this is a UI-mirror assumption, not a protocol guarantee.

2. Scope overlap with #2862: this PR's rxant parsing in PanadapterModel and the "send via pan set when panadapter-authoritative" change in the RX ANT combo overlap with PR #2862's stated changes. You called out the merge-tree check in the description, so this is just a flag for the maintainer's attention on landing order.

Thanks for the FlexLib-mirroring discipline (capability table + matching test row) and for citing the FLEX-6500 LoopA / FLEX-6700 LoopA+LoopB hardware specifics in the model table comments. Appreciate the contribution! 🎯