I'm now trying to import an array of related objects and Sanity want's a unique _key for each entry. How can I get Sanity to just pick its own key during the import?

As part of my first Sanity project I'm importing existing data. All's going well, not pretty but it works, except I'm now trying to import an array of related objects and Sanity want's a unique

_key

for each entry. How can I get Sanity to just pick its own key during the import?

You can just use something like

nanoid

to create your own.

Thanks, I guess I could generate my own UUID but I just wondering if there was some built-in feature I'd missed

An ID can be any string so just make it something predictable and you can pass it along.

Does it have to be unique across the whole dataset or just within the array of object?

Unique across the dataset I believe. If I’m doing an import I’ll just generate the with the schema hard coded so it’s unique
So if your imported document has a “title” say, make the ID

schema-name-${title}

and then you can generate the same ID again elsewhere in your imports if necessary.

Actually I'm talking about the

_key

that is required when referencing related documents rather than the

_id

of the documents.

(The

_key

only has to be unique within the array itself) 🙂

Thanks Peter, I'll work with that