Fix Electrical Measurement cluster readings

[manup]

No need to filter attributes per device, if attributes aren't available they won't be read a second time, as this is checked later on.

Fixes Blitzwolf BW-SHP13 ZHAPower sensor updates.
Issue: #3788

[Smanar]

@manup I don't remember for osram, but for legrand attributes are availables, just useless, values are locked to 0.
And from my memory it s same for Osram, the value is locked to 50 or something like that.

[SwoopX]

In general, you're right. However, a couple of months ago, I added some code to just query available attributes. Hence, if the attributes are non-existant, nothing will be queried.

Now with your example, I guess it all boils down to exposure of the resource items in question. E.g. the Osrams just report garbage as power, so power shouldn't be exposed for those. I suppose it should be double checked if the unwanted and previously blacklisted items aren't exposed via API...

[manup]

@manup I don't remember for osram, but for legrand attributes are availables, just useless, values are locked to 0.
And from my memory it s same for Osram, the value is locked to 50 or something like that.

I could only check for OSRAM plug, here the RMS attributes are not available.
But more importantly, the plug doesn't have the RStateVoltage and RStateCurrent ResourceItems so they won't be set to bogus values.

The add sensor node code doesn't add these items to Legrand devices either, I think they should be fine too?

[marithpl]

Is it released?
I have still problem with SHP13 and power consumption.

[Mimiix]

@marithpl Yes. Did you create a issue already?

[marithpl]

I didn't.
I found a bug report #3788 with same issue.
Here is my issue: #4629

[Mimiix]

I've reopened #3788