Many organizations regularly import data using the exact same column schema every time. It would be useful to store the mapping of columns somehow, so that Zetkin recognizes it.
This could be done as part of the regular import API, or using a separate (per-app) settings API.
Many organizations regularly import data using the exact same column schema every time. It would be useful to store the mapping of columns somehow, so that Zetkin recognizes it.
This could be done as part of the regular import API, or using a separate (per-app) settings API.