Move timezone and utils to separate package

[pub-djedruszczak]

This PR moves the api.timeZone var and certain utilities from the api package into a new tz package. The GetFromTo function is exported for use in a future native implementation of the linearRegression function.

Some encapsulation of the api.timeZone var (now tz.TimeZone) is lost so that it can continue to be set from api.ConfigSetup, maintaining backwards compatibility. The api package is highly dependent on the expr package, making the export of api.getFromTo from the api package for use in native function implementations (the expr package) difficult due to import cycles.

[Dieterbe]

in master, the timezone is considered an api level configuration, as various of the api methods - before involving expr - call getFromTo, and thus need access to the default timezone configured.
I'm not sure why linearRegression would need to access that variable. Is it because from/to can be any 'at' style expression, which allows specifying a timezone? wouldn't that timezone then be scoped to that function call only? it has no bearing on the request-scoped from/to, correct? do you envision needing only read access or also writing to the timezone variable?

also, while i never particularly liked the idea of bringing all api models together in one 'models' package, that is how it is, and moving just a single model ( FromTo ) elsewhere is definitely counter to our current code structure.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[robert-milan]

I'll be taking over reviewing this for now

[CLAassistant]

All committers have signed the CLA.

[robert-milan]

Yes, we are slightly departing from our traditional layout of Metrictank files / models / separation of code, however I think we should just move forward with this. If it becomes a problem in the future we can always change it.

[robert-milan]

Can we get @djedruszczak to agree to the CLA?

[pub-djedruszczak]

Can we get @djedruszczak to agree to the CLA?

Signed!