What additional value does this feature bring to the project?
I want to define the clocktick for ldts either explicitly as variable or implicitly by defining the format for ldts columns.
Is your feature request related to a problem? Please describe.
At the moment it is possible to choose a datatype for the ldts via the variable datavault4dbt.timestamp_format but in the sat_v1 templates there is a hardcoded time difference subtracted (e.g. for fabric 100 ns).
For example when someone chooses datetime2(0) as format for their ldts in a fabric implementation then the 100 ns that are subtracted in the sat_v1 models are washed away by rounding the ledts to full seconds.
Describe the solution you'd like
When choosing the clocktick to be one second implicitly by choosing the datavault4dbt.timestamp_format the "time subtraction" in the sat_v1 models should be adjusted dynamically.
Describe alternatives you've considered
At the moment there are two possibilities from my point of view:
- Choosing a more granular data type for ldts that corresponds to the time subtracted for my chosen adapter
- Overriding the sat_v1 macro
What additional value does this feature bring to the project?
I want to define the clocktick for ldts either explicitly as variable or implicitly by defining the format for ldts columns.
Is your feature request related to a problem? Please describe.
At the moment it is possible to choose a datatype for the ldts via the variable datavault4dbt.timestamp_format but in the sat_v1 templates there is a hardcoded time difference subtracted (e.g. for fabric 100 ns).
For example when someone chooses datetime2(0) as format for their ldts in a fabric implementation then the 100 ns that are subtracted in the sat_v1 models are washed away by rounding the ledts to full seconds.
Describe the solution you'd like
When choosing the clocktick to be one second implicitly by choosing the datavault4dbt.timestamp_format the "time subtraction" in the sat_v1 models should be adjusted dynamically.
Describe alternatives you've considered
At the moment there are two possibilities from my point of view: