Create a sensor foreach attribute

[edenhaus]

Fixes #40
Fixes #55 long term statistics are available for sensors
Fixes #25

Fixes #51 Introducing a dependency model between each sensor/attribute and therefore we can only update some sensors/attributes.
This also gives the possibility, that we take only at most 60 seconds old values into the calculation.
Example use case:

1. The last state change of the weight and impedance sensor is a day old
2. A new state for the weight sensor is reported
3. Only sensor/attributes, that depend on the weight only will be updated. The impedance value is to old to take into account.
4. 30 seconds delay
5. A new state for the impdeance sensor is reported
6. Sensors who depends on impedance or weight and impedance will be updated, as the weight state is newer than 60 seconds.

[dckiller51]

I just tested your version. I have the sensors to create but they are in unknown.

[edenhaus]

@dckiller51 This PR is marked as draft as it is not ready yet :) I have to fix first the error, when not all data is available (which is your problem currently)
When it's ready for testing, I will request a review from you (Hopefully by the end of this week)

[edenhaus]

Fixes #51 Introducing a dependency model between each sensor/attribute and therefore we can only update some sensors/attributes. This also gives the possibility, that we take only at most 60 seconds old values into the calculation. Example use case:

1. The last state change of the weight and impedance sensor is a day old

2. A new state for the weight sensor is reported

3. Only sensor/attributes, that depend on the weight only will be updated. The impedance value is to old to take into account.

4. 30 seconds delay

5. A new state for the impdeance sensor is reported

6. Sensors who depends on impedance or weight and impedance will be updated, as the weight state is newer than 60 seconds.

@borpin @dckiller51 Do you think 60 seconds is enough?

P.S. This PR is still a draft, so not everything is working correctly (only be advised if you test this change)

[dckiller51]

To me that seems correct. Sometimes with our scale we have to go back on it because it was not at 0.

[edenhaus]

@dckiller51 Sorry my notebook was broken so it took a little bit longer. Can you please review and test it?

[dckiller51]

@dckiller51 Sorry my notebook was broken so it took a little bit longer. Can you please review and test it?
@edenhaus
Hello,
Overall it seems to me that it works well.

1. The weight is with many decimal.
2. When you create a new user you have to restart the system to update the data otherwise you get unknow.

Your again did an amazing job

[edenhaus]

@dckiller51 Thanks for your feedback :) I have fixed your findings.
Can you please retest it

[dckiller51]

@dckiller51 Thanks for your feedback :) I have fixed your findings. Can you please retest it

Just one last correction. The ideal weight is with many decimal.

[edenhaus]

I fixed your last finding too and will merge it now :)

[dckiller51]

@edenhaus
Hello, I just realized that we are losing information in the components bodymiscale.name