Feature: Add offset option for scaling measurement values

[wilbersl]

Add additive scaling option (offset) to register definition

Description

This PR introduces an additional scaling option called offset to the register definition.

The offset is applied after all existing multiplication-based scaling factors (multiplier, divisor, decimalShiftRight) have been processed.

This enables linear value transformations of the form:

            scaled_value = (
                value
                * (register_def.get("multiplier") or 1)
                * (10 ** (register_def.get("decimalshiftright") or 0))
                / (register_def.get("divisor") or 1)
            ) + (register_def.get("offset") or 0)

Motivation

Some registers cannot be represented correctly using multiplication-only scaling.

Typical use cases include:

• Temperature sensors with fixed offsets (e.g. -40°C)
• Analog values with shifted zero points
• Vendor-specific calibration offsets

Adding offset allows these cases to be modeled cleanly while keeping the existing scaling model intact.

Implementation Details

• Added optional offset field to the register definition
• Applied offset after all existing scaling operations
• Default value is 0
• No changes required for existing register definitions

Backward Compatibility

• Fully backward compatible
• Registers without offset behave exactly as before

Example

[[device.registers]]
number=1
startbit=0
nobits=16
signed=true
multiplier=1
divisor=1
decimalshiftright=0
offset=-40
input=false
name="Test_Int16"
measurementmapping.templatestring="{\"Test\":{\"Temp\":%% }}"