Importing spreadsheets and existing data

Existing sheets can be read

An existing legacy FoodProt workbook can be imported directly, because the platform knows its column headings. The same headings are written on export, so nobody has to retype them.

The headings keep their historical spacing and punctuation on purpose, because other teams’ sheets match on the exact text. Tidying a heading would break a file the platform never sees.

You confirm the columns

A data import asks you to confirm which column means what. A guess that maps "Hardness" to the hardness field is right most of the time, and when it is wrong the dataset still looks correct. Confirming takes a minute and is recorded.

The uploaded file is kept

The uploaded file is stored unchanged and linked to whatever was created from it, so you can always download exactly what was imported. Months later, this answers the question "what exactly did we import?" without relying on anyone’s memory.

Imported values are not verified

Imported values arrive as what they are: values a person supplied, with the file as their source. To join the verified dataset they need the same review as any other value. A reference set for checking accuracy belongs in validation, not in the dataset.