Merge "Reuse AppTransitionProto fields for AppTransitionProto" into main

    /** Less duration for CHANGE type because it does not involve app startup. */

    private static final int CHANGE_TIMEOUT_MS = 2000;

    // State constants to line-up with legacy app-transition proto expectations.

    private static final int LEGACY_STATE_IDLE = 0;

    private static final int LEGACY_STATE_READY = 1;

    private static final int LEGACY_STATE_RUNNING = 2;

    private final ArrayList<TransitionPlayerRecord> mTransitionPlayers = new ArrayList<>();

    final TransitionMetricsReporter mTransitionMetricsReporter = new TransitionMetricsReporter();

    void dumpDebugLegacy(ProtoOutputStream proto, long fieldId) {

        final long token = proto.start(fieldId);

        int state = LEGACY_STATE_IDLE;

        int state = AppTransitionProto.APP_STATE_IDLE;

        if (!mPlayingTransitions.isEmpty()) {

            state = LEGACY_STATE_RUNNING;

            state = AppTransitionProto.APP_STATE_RUNNING;

        } else if ((mCollectingTransition != null && mCollectingTransition.getLegacyIsReady())

                || mSyncEngine.hasPendingSyncSets()) {

            // The transition may not be "ready", but we have a sync-transaction waiting to start.

            // Usually the pending transaction is for a transition, so assuming that is the case,

            // we can't be IDLE for test purposes. Ideally, we should have a STATE_COLLECTING.

            state = LEGACY_STATE_READY;

            // we can't be IDLE for test purposes. Ideally, we should have an APP_STATE_COLLECTING.

            state = AppTransitionProto.APP_STATE_READY;

        }

        proto.write(AppTransitionProto.APP_TRANSITION_STATE, state);

        proto.end(token);