Add location topic

[brianmay]

Fixes #3660.

[netlify]

✅ Deploy Preview for teslamate ready!

Name	Link
🔨 Latest commit	00f3620
🔍 Latest deploy log	https://app.netlify.com/sites/teslamate/deploys/65f7ee2c80400700087eee41
😎 Deploy Preview	https://deploy-preview-3729--teslamate.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[brianmay]

Not tested yet :-)

[brianmay]

Not sure I understand why this test is failing:

 1) test publishes vehicle data (TeslaMate.Mqtt.PubSub.VehicleSubscriberTest)
Error:      test/teslamate/mqtt/pubsub/vehicle_subscriber_test.exs:27
     Unexpectedly received message {VehiclesMock, {:subscribe_to_summary, 0}} (which matched _)
     code: refute_receive _
     stacktrace:
       test/teslamate/mqtt/pubsub/vehicle_subscriber_test.exs:103: (test)

The test already has at the top:

    assert_receive {VehiclesMock, {:subscribe_to_summary, 0}}

[brianmay]

Test now works, but I don't understand why. Seems dodgy.

[brianmay]

Not sure how the tests were even passing before. All fixed now.

[brianmay]

Problem with this is that there are times when we have the latitude, longitude, but not the elevation. So as a result nothing gets published. Arrgh.

Think I will remove elevation. I think the Tesla API treats elevation is very different from latitude and longitude anyway.

[brianmay]

Seems to be working so far.

latitude/longitude are being encoded as quoted string values, not floats. Which is a bit unexpected.

I might try to get them as floats.

[brianmay]

teslamate uses Decimal type everywhere for latitude/longitude.

Suspect this is a misunderstanding here. Decimal is required for applications that keep track of numbers - like $ and cents, where it is vital that we represent values like 0.01 exactly.

But this is not one of those applications. It doesn't matter really if a latitude of 0.01 gets converted to 0.0099999997 (https://hexdocs.pm/elixir/1.13.4/Float.html). In fact it is likely the values came out of floating point numbers in Tesla's code anyway.

But will keep it as is for now.

Values are now coming out as floats. And tests appear to be passing.

Early on one of the tests passed, which I hadn't even updated. That was weird. Fixed now.

[JakobLichterfeld]

Thanks for your deditcation!

[brianmay]

Arggh. Looks like some parts of the code use the Decimal type, and other parts use the floating point type.

I like having compile time type checking...

I think I might get rid of all the Decimal values. For latitude/longitude. Just use floats everywhere.

Mar 12 08:01:51 iot2 teslamate[8982]: 2024-03-12 08:01:51.653 [error] GenServer #PID<0.176940.0> terminating
Mar 12 08:01:51 iot2 teslamate[8982]: ** (FunctionClauseError) no function clause matching in Decimal.to_float/1
Mar 12 08:01:51 iot2 teslamate[8982]:     (decimal 2.1.1) lib/decimal.ex:1498: Decimal.to_float(-37.906105)
Mar 12 08:01:51 iot2 teslamate[8982]:     (teslamate 1.28.4) lib/teslamate/mqtt/pubsub/vehicle_subscriber.ex:81: TeslaMate.Mqtt.PubSub.VehicleSubscriber.handle_info/2
Mar 12 08:01:51 iot2 teslamate[8982]:     (stdlib 5.2) gen_server.erl:1095: :gen_server.try_handle_info/3
Mar 12 08:01:51 iot2 teslamate[8982]:     (stdlib 5.2) gen_server.erl:1183: :gen_server.handle_msg/6
Mar 12 08:01:51 iot2 teslamate[8982]:     (stdlib 5.2) proc_lib.erl:241: :proc_lib.init_p_do_apply/3
Mar 12 08:01:51 iot2 teslamate[8982]: Last message: %TeslaMate.Vehicles.Vehicle.Summary{car: %TeslaMate.Log.Car{__meta__: #Ecto.Schema.Metadata<:loaded, "cars">, id: 1, name: nil, efficiency: 0.15006, model: "S", trim_badging: "90D", marketing_name: nil, exterior_color: "Red", wheel_type: "AeroTurbine19", spoiler_type: "None", eid: 1492931337154343, vid: 1318482982, vin: "5YJSB7E22GF153397", settings_id: 1, settings: %TeslaMate.Settings.CarSettings{__meta__: #Ecto.Schema.Metadata<:loaded, "car_settings">, id: 1, suspend_min: 21, suspend_after_idle_min: 15, req_not_unlocked: true, free_supercharging: true, use_streaming_api: true, car: #Ecto.Association.NotLoaded<association :car is not loaded>}, charging_processes: #Ecto.Association.NotLoaded<association :charging_processes is not loaded>, positions: #Ecto.Association.NotLoaded<association :positions is not loaded>, drives: #Ecto.Association.NotLoaded<association :drives is not loaded>, inserted_at: ~N[2021-01-07 00:28:25], updated_at: ~N[2024-02-09 21:57:09]}, display_name: nil, state: :online, since: ~U[2024-03-11 20:47:54.879000Z], healthy: true, latitude: -37.906105, longitude: 145.343587, heading: 19, battery_level: 69, charging_state: "Disconnected", usable_battery_level: 69, ideal_battery_range_km: 291.29, est_battery_range_km: 254.68, rated_battery_range_km: 363.15, charge_energy_added: 0.0, speed: 0, outside_temp: 24.5, inside_temp: 18.6, is_climate_on: true, is_preconditioning: false, locked: false, sentry_mode: nil, plugged_in: false, scheduled_charging_start_time: nil, charge_limit_soc: 90, charger_power: 0, windows_open: true, doors_open: false, odometer: 157541.71, shift_state: "P", charge_port_door_open: false, time_to_full_charge: 0.0, charger_phases: nil, charger_actual_current: 0, charger_voltage: 0, version: "2024.2.7", update_available: false, update_version: "", is_user_present: true, geofence: nil, model: "S", trim_badging: "90D", exterior_color: "Red", wheel_type: "AeroTurbine19", spoiler_type: "None", trunk_open: false, frunk_open: false, elevation: 228, power: 1, charge_current_request: 24, ...}

[brianmay]

Database values are stored in decimal, since 45d94d5. Supposedly this makes things more efficient.

Values from the TeslaApi are retrieved at floats.

I think I am will put all the Decimal types back and have this new code deal with both cases.

[brianmay]

We seem to have an unrelated intermittent test failure. https://github.com/teslamate-org/teslamate/actions/runs/8240221994/job/22535198738?pr=3729

[brianmay]

Sorry, didn't mean to close, pushed the wrong thing.

[brianmay]

Testing again. But next drive not scheduled until this afternoon.

[brianmay]

Just noticed we always publish this every time. We should be checking if the value has changed since the last time. Will fix.

[brianmay]

See #3736 concerning the test failures,

All the tests passed on my box :-)

[brianmay]

This seems to be working correctly now.

[JakobLichterfeld]

Test issue should be resolved by #3733. If you don't mind, update from master.

Edit: issue with async run still persist on master, "solved" by rerun action.

[brianmay]

Rebased against latest master.

Not 100% comfortable adding 30 extra lines to the handle_info function call, but I think to early to refactor.

Refactoring probably should happen if we are going to output other stuff as json blobs however.

[brianmay]

Had to retry CI tests three times to get a pass :-(

I think they work OK from my box however.

[JakobLichterfeld]

Will merge as soon as I came up with a good changelog for this, as it will break some third party integrations

[brianmay]

Shouldn't cause any breakage, the current topics remain unchanged.

Unless I missed something here?

[JakobLichterfeld]

Shouldn't cause any breakage, the current topics remain unchanged.

Unless I missed something here?

You are absolutely right, and I did not look close enough. Will double-check and merge tomorrow.

Thanks for your dedication!

[JakobLichterfeld]

lgtm, I added the doc entry for the new topic

[WesSec]

What a hero, thanks for the efforts <3

[vbarrier]

@tobiasehlert I think this new topic, generate a lot of warning on teslamateapi

[tobiasehlert]

@tobiasehlert I think this new topic, generate a lot of warning on teslamateapi

@vbarrier, noticed some errors as well at home. Going to add a fix very soon 👌