I understand that a column can be “the primary column”, which is similiar to a primary key in a classic database.
Now, the content of that column can be eg “names”. And even if two individuals have exactly the same name, their names will be “two different primary keys”.
This is all fine inside spreadsheet.com, but when you sooner or later would like to take your dataset created inside spreadsheet.com and use that as “master data” for other data flows outside of spreadsheetdotcom, it would probably be good to have a generated unique ID in some contexts.
I wouldn’t be surprised if there is such a “hidden” primary key behind the scenes (?), the question is if tthat can be captured in eg “simple csv export”?
1 Like
Welcome @Johan_Myrberger!
Yes there is behind the scenes a unique identifier per row though we don’t expose this in the UI. However, one best practice we’ve been recommending is to create an Autonumber column for this purpose: https://support.spreadsheet.com/hc/en-us/articles/360039059192-Autonumber
Autonumber data types allow you to create a unique and uneditable ID for your rows, ensuring that each row has a unique identifier even if some of the data in a row is the same. Ideal for tracking product orders, bugs in a bug tracker, and as a reference to ensure data integrity.
… only one Autonumber column can be applied per worksheet.
Like Matt mentioned - yes each row has a type4 uuid associated with it. The row uuid is omnipresent in our internal API usage patterns and will be available in our public REST API as well when it is released.
The facility for a REST API consumer to look up a row using the uuid and update some field seems like a must have. I can also imagine the uuid being one of the columns in a ‘listRows’ API response.
Therefore, the ability for a UI user to view/grab a row’s uuid makes sense - May be somewhere in the Row Preview Dialog - We may need it for API/integration developers.