[alex-so-3]

Stopping continuous ringing only stopped the app from re-triggering the sound. A DULT accessory (third-party Find My trackers) kept sounding until its own timeout.

AccessorySoundTrigger gets a stopSound, and the BLE implementation writes the DULT Sound_Stop opcode (0x0301) to the non-owner control point when the user stops ringing. AirTags have no stop command in the paths this app uses and stop by themselves, so nothing changes for them.

Verified: on a phone, with a commercial DULT tracker and with a DIY tag that implements DULT: the sound now stops when ringing is stopped. It stops with a short delay, because the app first has to connect to the tracker to send the command. ./gradlew testAll and add_strings.py --check pass.

[parawanderer]

Thanks for checking this + the fix & contribution. Checking this for issues and seeing if it makes sense to add a test to cover this...

[parawanderer]

Pushed 09f2abb on top:

• stopSound goes through an injected GattStopper seam instead of instanceof BluetoothDevice plus a static call, so BleAccessorySoundTriggerTest covers it: no permission, no candidates, not nearby, found, and a failing stop.
• The stop is only sent when the ping actually rang over DULT (BleSoundTriggerResult.hasStopCommand). Stopping an AirTag no longer costs a scan and a GATT connection.
• Starting a new ping for the same tag cancels a stop still in flight. Before, that stop could arrive after the new Sound_Start and silence it.
• Imports instead of fully qualified names, braces on the one-line ifs, and the unused async variable is gone.

Drafted by Claude Code, reviewed by @parawanderer