CoreDataStack

CoreDataStack provides a simple all-in-one Core Data stack. It is designed to make Core Data both safe and easy to use. For applications which require high performance, via multiple writers, this project is not for you.

CoreDataStack provides a readonly NSManagedObjectContext for use in the main thread. Contexts for writing are vended as needed, and work is serialized on a background queue.

Initialization

public init(modelName: String, storeType: String) throws

Initializes a stack with the given model name and store type.

convenience public init(modelName: String) throws

Initializes a stack with the given model name and a store type of NSSQLiteStoreType.

Working with managed objects.

Modifying objects

The main workhorse is

public func perform(_ block: @escaping CoreDataManagerUpdateBlock, completion: (() -> ())? = nil)

This method will submit a block to a background context. After the operation is complete, the completion is called on the same thread as the block.

CoreDataManagerUpdateBlock receives a context and a shouldSave boolean which may be modified by the block. When set to false, all changes will be discarded.

Querying objects

public func query(_ block: @escaping CoreDataManagerQueryBlock)

This method will submit a block which receives a readonly context. The block runs on a background queue, and is intended for handling work which requires information from Core Data, but which does not need to modify it. This block runs in parallel with perform() blocks.

public func viewQuery(_ block: CoreDataManagerQueryBlock)

This method submits a block on the viewContext, and should only be called from the main thread.

Contexts

Contexts which are used for modifying core data are vendored by the perform() method. These contexts are destroyed after the block completes, and before the completion is invoked.

Contexts which are used for background queries are vendored by the query() method. These contexts are destroyed after the block is complete. If the context has changes, a fatalError() is thrown.

There is a special, persistent context, called the viewContext, used for querying Core Data on the main thread. Changes made by perform() contexts are merged into this context when they are saved.

Predicates

Extensions to the NSPredicate class provide several methods and operators to simplify the creation of predicates.

public static func predicate(for conditions: [String: Any]) -> NSPredicate

This method takes a dictionary of key-value pairs. The keys are properties of the Core Data entity being queried. The values are those which will satisfy the query. Conditions are joined together using an AND predicate.

Examples

// someKey == 5
predicate(for: ["someKey": 5])

// someKey BETWEEN 5 and 10
predicate(for: ["someKey": 5...10])

// someKey IN [1, 3, 5]
predicate(for: ["someKey", [1, 3, 5]])

Compound predicates

It is also possible to create compound statements using the following methods.

public func or(_ conditions: [String: Any]) -> NSPredicate

public func or(_ predicate: NSPredicate) -> NSPredicate

public func and(_ conditions: [String: Any]) -> NSPredicate

public func and(_ predicate: NSPredicate) -> NSPredicate

Predicates may also be combined using the following operators.

// AND
public static func && (left: NSPredicate, right: NSPredicate) -> NSPredicate

// OR
public static func || (left: NSPredicate, right: NSPredicate) -> NSPredicate

License

This software is under the MIT license. See LICENSE for details.