I tend to follow the logic of awk better than the one of sed (not being a vi enthusiast), but it is good to see when one might be better than the other:
I made a configurable #etl mapper in #python. I have extensive experience with ETL and a variety of middleware clients, and used that to create my own. Right now, it only reads from Google Sheets (of all things). Next I'll make it read from MySQL databases. I wonder what other connectors I should make. Anyway, it's available as #FOSS, here: https://github.com/aeveltstra/lammpster