We should allow collections to be populated from a spreadsheet or CSV.
A common usecase is that mapping rules and lookup tables are defined in spreadsheets somewhere. But then we need to extract a JSON structure to represent those mappings.
You can do this in two ways.
1: map each row to a key/value pair (ie, a new collection item)
- For each row, one column is the key (typically the first).
- The rest of the row is then parsed into a JSON object, where the column is the key (with a mapping?) and the cell is the value.
- Map the whole sheet into a single collections key, whose value is an object, and each row represents a nested object within it (with a key).
I am not sure if any one case is overwhelmingly more useful - it probably depends on your sheets structure and the density of your mapping.