Refactor internals to a more generic unmarshaling approach,
and with less reflection happening inside the actual parsing loop
Implement ExcelUnmarshaler interface
Support TextUnmarshaler as fallback
Allow configuration of error handling:
Ignore or Abort when encountering missing fields
(too many columns in excel)
Ignore or Abort when encounering unknown field types
Ignore, Abort or Collect parsing errors for the actual data
While these changes should be backwards-compatibly with the previous versions, there will be small changes in handling of certain edge cases.
That makes this technically a breaking change.
While these changes should be backwards-compatibly with the previous versions, there will be small changes in handling of certain edge cases. That makes this technically a breaking change.