A base class that implements the behavior required of a Core Data model object.
- iOS 3.0+
- macOS 10.4+
- Mac Catalyst 13.0+Beta
- tvOS 9.0+
- watchOS 2.0+
- Core Data
A managed object is associated with an entity description (
NSEntity) that provides metadata about the object, including the name of the entity that the object represents and the names of its attributes and relationships. A managed object is also associated with a managed object context that tracks changes to the object graph.
You can’t use instances of direct subclasses of
NSObject (or any other class that doesn’t inherit from
NSManaged) with a managed object context. You may create custom subclasses of
NSManaged, although this isn’t always required. If no custom logic is needed, you can create a complete Object graph with
If you instantiate a managed object directly, you must call the designated initializer
In some respects, an
NSManaged acts like a dictionary—it’s a generic container object that provides efficient storage for the properties defined by its associated
NSManaged supports a range of common types for attribute values, including string, date, and number (see
NSAttribute for full details). Therefore, typically you don’t need to define instance variables in subclasses. Sometimes, however, you want to use types that aren’t supported directly, such as colors and C structures. For example, in a graphics application you might want to define a Rectangle entity that has color and bounds attributes that are an instance of
NSColor and an
NSRect struct, respectively. For some types you can use a transformable attribute, for others this may require you to create a subclass of
Managed objects typically represent data held in a persistent store. In some situations a managed object may be a fault—an object whose property values haven’t yet been loaded from the external data store. When you access persistent property values, the fault “fires” and the data is retrieved from the store automatically. This can be a comparatively expensive process (potentially requiring a round trip to the persistent store), and you may wish to avoid unnecessarily firing a fault. See Faulting and Uniquing for more details on faults.
You can safely invoke the following methods and properties on a fault without causing it to fire:
hash don’t cause a fault to fire, managed objects can typically be placed in collections without firing a fault. Note, however, that invoking key-value coding methods on the collection object might in turn result in an invocation of
value on a managed object, which would fire the fault.
description property doesn’t cause a fault to fire, if you implement a custom
description that accesses the object’s persistent properties, this does cause a fault to fire. You are strongly discouraged from overriding
description in this way.
In combination with the entity description in the managed object model,
NSManaged provides a rich set of default behaviors including support for arbitrary properties and value validation. If you decide to subclass
NSManaged to implement custom features, make sure you don’t disrupt Core Data’s behavior.
Methods and Properties You Must Not Override
NSManaged itself customizes many features of
NSObject so that managed objects can be properly integrated into the Core Data infrastructure. Core Data relies on the
NSManaged implementation of the following methods and properties, which you therefore absolutely must not override:
Methods and Properties You Shouldn't Override
As with any class, you are strongly discouraged from overriding the key-value observing methods such as
did. Avoid overriding
description—if this method fires a fault during a debugging operation, the results may be unpredictable. Also avoid overriding
dealloc. Changing values in the
init(entity: method won't be noticed by the context, and if you aren't careful, those changes may not be saved. Perform most initialization customization in one of the
awake… methods. If you do override
init(entity:, make sure you adhere to the requirements set out in the method description. See
did is usually a better time to clear values—a managed object may not be reclaimed for some time after it has been turned into a fault. Core Data doesn’t guarantee that
dealloc will be called in all scenarios (such as when the application quits). Therefore, don’t include required side effects (like saving or changes to the file system, user preferences, and so on) in these methods.
In summary, for
dealloc, Core Data reserves exclusive control over the life cycle of the managed object (that is, raw memory management). This is so that the framework can provide features such as uniquing and by consequence, relationship maintenance, as well as much better performance than would be possible otherwise.
Additional Override Considerations
The following methods are intended to be fine grained and aren’t suitable for large-scale operations. Don’t fetch or save in these methods. In particular, they shouldn’t have side effects on the managed object context.
Turn Into Fault()
Turn Into Fault()
In addition, if you plan to override
awake, and validation methods, first invoke
super, the superclass’s implementation. Don’t modify relationships in
awake—see the method description for details.
Custom Accessor Methods
Typically, you don’t need to write custom accessor methods for properties that are defined in the entity of a managed object’s corresponding managed object model. If you need to do so, follow the implementation patterns described in Managed Object Accessor Methods in Core Data Programming Guide.
Core Data automatically generates accessor methods (and primitive accessor methods) for you. For attributes and to-one relationships, Core Data generates the standard get and set accessor methods; for to-many relationships, Core Data generates the indexed accessor methods as described in Achieving Basic Key-Value Coding Compliance in Key-Value Coding Programming Guide. You do however need to declare the accessor methods or use Objective-C properties to suppress compiler warnings. For a full discussion, see Managed Object Accessor Methods in Core Data Programming Guide.
Custom Instance Variables
NSManaged stores its properties in an internal structure as objects, and in general Core Data is more efficient working with storage under its own control rather than by using custom instance variables.
NSManaged provides support for a range of common types for attribute values, including string, date, and number (see
NSAttribute for full details). If you want to use types that aren’t supported directly, like colors and C structures, you can either use transformable attributes or create a subclass of
Sometimes it’s convenient to represent variables as scalars—in drawing applications, for example, where variables represent dimensions and x and y coordinates and are frequently used in calculations. To represent attributes as scalars, you declare instance variables as you do in any other class. You also need to implement suitable accessor methods as described in Managed Object Accessor Methods.
If you define custom instance variables for example to store derived attributes or other transient properties, clean up these variables in
did rather than
NSManaged provides consistent hooks for validating property and inter-property values. You typically shouldn’t override
validate. Instead implement methods of the form
validate<Key>:, as defined by the NSKeyValueCoding protocol. If you want to validate inter-property values, you can override
validate and/or related validation methods.
validate within custom property validation methods—if you do, you create an infinite loop when
validate is invoked at runtime. If you do implement custom validation methods, don’t call them directly. Instead, call
validate with the appropriate key. This ensures that any constraints defined in the managed object model are applied.
If you implement custom inter-property validation methods like
validate, call the superclass’s implementation first. This ensures that individual property validation methods are also invoked. If there are multiple validation failures in one operation, collect them in an array and add the array—using the key
NSDetailed—to the userInfo dictionary in the
NSError object you return. For an example, see Managed Object Validation.