🔑 No Need to Set Primary Key in Core Data/SwiftData

TL;DR: Core Data and SwiftData auto-manage primary keys using SQLite’s Z_PK (auto-increment), Z_ENT (entity type), and Z_PRIMARYKEY (key tracking), eliminating the need for manual primary key setup.

Question:

For iOS developers, one common question when using Core Data or SwiftData is: why don’t these frameworks require setting a primary key, like other databases?

Answer:

The answer lies in the underlying SQLite database where both frameworks automatically implement a robust primary key mechanism:

1. Auto-Incremented Primary Key (Z_PK): Each entity table includes a Z_PK column that starts at 1 and serves as the auto-incrementing primary key.
2. Entity Identification (Z_ENT): A Z_ENT column identifies the entity type for each record.
3. Unique Record Locator (Z_PK + Z_ENT): The combination of Z_PK and Z_ENT uniquely identifies each record across the entire database.
4. Primary Key Management (Z_PRIMARYKEY): Core Data manages auto-incrementing values for primary keys through the Z_PRIMARYKEY table, which tracks the maximum key value (Z_MAX) for each entity.

Thanks to this built-in primary key system, developers neither need nor can define their own auto-incrementing keys when setting up data models. This design ensures data uniqueness while simplifying development.

Pro Tip:

If you plan to export your data, it’s best to add an identifier property, such as a UUID, to your model. This guarantees that relationships between records remain intact after export.

References:

• How Core Data Saves Data in SQLite
• NSManagedObjectID and PersistentIdentifier: Mastering Data Identifiers in Core Data and SwiftData