More information in CSV: add accuracy, filtered data and improve speed

[offtherailz]

• an additional field should be added to the CSV to discriminate if data has been filtered (by this custom build) or not (by the normal app). This field will discriminate between normal usage or unipi (not filtered data, with custom conf.json, see New build profile 'unipi'  #113)
• accuracy, retrieved from Location.getSpeedAccuracyMetersPerSecond should be added too in the CSV
• we should check if is possible to distinguish between 0 speed or NaN speed (in case it's not available, now it always results 0). [Location.hasSpeed]

Dev Notes:

• All the new fields must be added at the end of the CSV
• for the possible NaN speed value, we should add a new boolean column called has_speed with possible value 0 and 1.

[offtherailz]

@giohappy do you have any advice for handling this new CSV.
Can it be handled from the current back-end without issues? If yes, we don't have to do anything (event for future versions of SMB). If need some changes:

• Should we do some switch to create different CSV for different profiles(for profile unipi)?
• Should we improve back-end to support this format

[giohappy]

@offtherailz the current backend shouldn't complain if we add more fields, but it will fail if we change the current fields order. See here to check how the CSV rows are parsed.

[offtherailz]

sure, so I'll add in the specification that all the new fields should be added at the end.
What about the new NaN value? Do you think this new value can be handled correctly by the back-end?

[giohappy]

NaN shouldn't break it, but a test would be required to verify that nothing down the processing chain breask.
For safety I would create a new field ("speed_or_nan"?)

[offtherailz]

Updated, calling it "real_speed_value"

[giohappy]

let's go for "real_speed" ("value" suffix is redundant