Fix energy and voltage for Aqara T2 relay

[abmantis]

Proposed change

SSIA.

Additional information

Device diagnostics

Checklist

• The changes are tested and work correctly
• pre-commit checks pass / the code has been formatted using Black
• Tests have been added to verify that the new code works
• Device diagnostics data has been attached

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.24%. Comparing base (aa7b32d) to head (e7661d2).

Additional details and impacted files

@@           Coverage Diff           @@
##              dev    #4458   +/-   ##
=======================================
  Coverage   92.24%   92.24%           
=======================================
  Files         369      369           
  Lines       12088    12090    +2     
=======================================
+ Hits        11150    11152    +2     
  Misses        938      938           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[puddly]

Would this work? It should avoid requiring any runtime-testable code:

    _CONSTANT_ATTRIBUTES = {
        MeteringCluster.AttributeDefs.divisor.id: 1000,
    }

[abmantis]

Yes! But its 1000000 instead since 1000 is already what we have by default for xiaomi: https://github.com/zigpy/zha-device-handlers/blob/dev/zhaquirks/xiaomi/__init__.py#L712

[puddly]

Ah. In that case you'd need to do _CONSTANT_ATTRIBUTES = {**MeteringCluster._CONSTANT_ATTRIBUTES, MeteringCluster.AttributeDefs.divisor.id: 1000000,}.

Is the device really sending reports scaled this much? Or are we hiding a * 1000 somewhere?

[abmantis]

Ah. In that case you'd need to do _CONSTANT_ATTRIBUTES = {**MeteringCluster._CONSTANT_ATTRIBUTES, MeteringCluster.AttributeDefs.divisor.id: 1000000,}.

Oh, duh! This is a sign that I should probably go to sleep... 😆

Is the device really sending reports scaled this much? Or are we hiding a * 1000 somewhere?

We scale it here: https://github.com/abmantis/zha-device-handlers/blob/xiaomi_t2_fix_energy/zhaquirks/xiaomi/__init__.py#L

[TheJulianJES]

Z2M also has a special case for this device:
https://github.com/Koenkk/zigbee-herdsman-converters/blob/ea780c92873ad254818f9bb7060f9417879893d6/src/lib/lumi.ts#L503-L507

The standard 1000 divisor is because of us needing to do the trip through zigpy cache/ZCL datatypes: https://github.com/abmantis/zha-device-handlers/blob/3916d9b9efa74e35a745c1c3d88cccb1645e86f2/zhaquirks/xiaomi/__init__.py#L329-L333
Z2M doesn't need to do this, but we'd just cut off all decimal points otherwise.

[TheJulianJES]

LGTM. Thanks!

[TheJulianJES]

Energy and voltage just didn't display for this device before at all, right?

[abmantis]

Energy and voltage just didn't display for this device before at all, right?

Power was already displayed and updated more frequently than the custom xiaomi attribute updates.
IIRC energy and voltage where shown as 0. If you think it is relevant, I can unpair a device and pair it to my dev env to check it out with a clean database.