I am following Apple's instruction to sync SwiftData with CloudKit. While initiating the ModelContainer, right after removing the store from Core Data, the error occurs: FAULT: NSInternalInconsistencyException: This NSPersistentStoreCoordinator has no persistent stores (unknown). It cannot perform a save operation.; (user info absent) I've tried removing default.store and its related files/folders before creating the ModelContainer with FileManager but it does not resolve the issue. Isn't it supposed to create a new store when the ModelContainer is initialized? I don't understand why this error occurs. Error disappears when I comment out the #if DEBUG block. Code: import CoreData import SwiftData import SwiftUI struct InitView: View { @Binding var modelContainer: ModelContainer? @Binding var isReady: Bool @State private var loadingDots = "" @State private var timer: Timer? var body: some View { VStack(spacing: 16) { Text("Loading\(loadingDots)") .font(.title2) .foregroundColor(.gray) } .padding() .onAppear { startAnimation() registerTransformers() let config = ModelConfiguration() let newContainer: ModelContainer do { #if DEBUG // Use an autorelease pool to make sure Swift deallocates the persistent // container before setting up the SwiftData stack. try autoreleasepool { let desc = NSPersistentStoreDescription(url: config.url) let opts = NSPersistentCloudKitContainerOptions(containerIdentifier: "iCloud.my-container-identifier") desc.cloudKitContainerOptions = opts // Load the store synchronously so it completes before initializing the // CloudKit schema. desc.shouldAddStoreAsynchronously = false if let mom = NSManagedObjectModel.makeManagedObjectModel(for: [Page.self]) { let container = NSPersistentCloudKitContainer(name: "Pages", managedObjectModel: mom) container.persistentStoreDescriptions = [desc] container.loadPersistentStores { _, err in if let err { fatalError(err.localizedDescription) } } // Initialize the CloudKit schema after the store finishes loading. try container.initializeCloudKitSchema() // Remove and unload the store from the persistent container. if let store = container.persistentStoreCoordinator.persistentStores.first { try container.persistentStoreCoordinator.remove(store) } } // let fileManager = FileManager.default // let sqliteURL = config.url // let urls: [URL] = [ // sqliteURL, // sqliteURL.deletingLastPathComponent().appendingPathComponent("default.store-shm"), // sqliteURL.deletingLastPathComponent().appendingPathComponent("default.store-wal"), // sqliteURL.deletingLastPathComponent().appendingPathComponent(".default_SUPPORT"), // sqliteURL.deletingLastPathComponent().appendingPathComponent("default_ckAssets") // ] // for url in urls { // try? fileManager.removeItem(at: url) // } } #endif newContainer = try ModelContainer(for: Page.self, configurations: config) // ERROR!!! } catch { fatalError(error.localizedDescription) } modelContainer = newContainer isReady = true } .onDisappear { stopAnimation() } } private func startAnimation() { timer = Timer.scheduledTimer( withTimeInterval: 0.5, repeats: true ) { _ in updateLoadingDots() } } private func stopAnimation() { timer?.invalidate() timer = nil } private func updateLoadingDots() { if loadingDots.count > 2 { loadingDots = "" } else { loadingDots += "." } } } import CoreData import SwiftData import SwiftUI @main struct MyApp: App { @State private var modelContainer: ModelContainer? @State private var isReady: Bool = false var body: some Scene { WindowGroup { if isReady, let modelContainer = modelContainer { ContentView() .modelContainer(modelContainer) } else { InitView(modelContainer: $modelContainer, isReady: $isReady) } } } }

I have transitioned to CKSyncEngine for syncing data to iCloud, and it is working quite well. I have a question regarding best practices for modifying and saving a CKRecord which already exists in the private or shared database. In my current app, most CKRecords will never be modified after saving to the database, so I do not persist a received record locally after updating my local data model. In the rare event that the local data for that record is modified, I manually fetch the associated server record from the database, modify it, and then use CKSyncEngine to save the modified record. As an alternative method, I can create a new CKRecord locally with the corresponding recordID and the modified data, and then use CKSyncEngine to attempt to save that record to the database. Doing so generates an error in the delegate method handleSentRecordZoneChanges, where I receive the local record I tried to save back inevent.failedRecordSaves with a .serverRecordChanged error, along with the corresponding server CKRecord. I can then update that server record with the local data and re-save using CKSyncEngine. I have not yet seen any issues when doing it this way. The advantage of the latter method is that CKSyncEngine handles the entire database operation, eliminating the manual fetch step. My question is: is this an acceptable practice, or could this result in other unforeseen issues?

我正在使用 Core Data 开发一个 SwiftUI 项目。我的数据模型中有一个名为 AppleUser 的实体，具有以下属性：id （UUID）、name （String）、email （String）、password （String） 和 createdAt （Date）。所有属性都是非可选的。 我使用 Xcode 的自动生成创建了相应的 Core Data 类文件（AppleUser+CoreDataClass.swift 和 AppleUser+CoreDataProperties.swift）。我还有一个 PersistenceController，它使用模型名称 JobLinkModel 初始化 NSPersistentContainer。 当我尝试使用以下方法保存新的 AppleUser 对象时： 让用户 = AppleUser（上下文：viewContext） user.id = UUID（） user.name = “用户 1” user.email = “...” user.password = “密码 1” user.createdAt = Date（）【电子邮件格式正确，但已替换为“...”出于隐私原因】 尝试？viewContext.save（） 我在控制台中收到以下错误：核心数据保存失败：Foundation._GenericObjCError.nilError， [：] 用户快照： [“id”： ...， “name”： “User1”， “email”： “...”， “password”： “...”， “createdAt”： ...] 所有字段都有有效值，核心数据模型似乎正确。我还尝试过： • 检查 NSPersistentContainer（name：） 中的模型名称是否与 .xcdatamodeld 文件 （JobLinkModel） 匹配 • 确保正确设置 AppleUser 实体类、模块和 Codegen（类定义、当前产品模块） • 删除重复或旧的 AppleUser 类文件 • 清理 Xcode 构建文件夹并从模拟器中删除应用程序 • 对上下文使用 @Environment（.managedObjectContext） 尽管如此，在保存新的 AppleUser 对象时，我仍然会收到 _GenericObjCError.nilError。 我想了解： 为什么即使所有字段都不是零且正确分配，核心数据也无法保存？ 这可能是由于一些残留的旧类文件引起的，还是我缺少设置中的其他内容？ 我应该采取哪些步骤来确保 Core Data 正确识别 AppleUser 实体并允许保存？ 任何帮助或指导将不胜感激。

I have an app which uses key-value storage and will not sync data past a certain size -- meaning that device "A" will send the data to the cloud but device "B" will never receive the updated data. Device "B" will receive the NSUbiquitousKeyValueStoreDidChangeExternallyNotification that the KVS changed but the data is empty. The data in in the KVS is comprised of 4 keys, each containing a value of NSData generated by NSKeyedArchiver. The NSData is comprised of property-list data types (e.g. numbers, strings, dates, etc.) I've verified that the KVS meets the limits of: A total of 1 MB per app, with a per-key limit of 1 MB A per-key value size limit of 1 MB, and a maximum of 1024 keys A maximum length for key strings is 64 bytes using UTF8 encoding Also, the app has never received an NSUbiquitousKeyValueStoreQuotaViolationChange notification. Of the 4 keys, 3 of them contain no more than 30 KB of data each. However, one of the keys can contain as much as 160 KB of data which will not sync to another device. Strangely, if I constrain the data to 100 KB it will work, however, that is not ideal as it is a fraction of the necessary data. I don't see any errors in the debug log either. Any suggestions on what to try next to get this working?

I've noticed several crashes that look like they're caused by an index out of bound in internal methods of NSFetchedResultsController. This happens while changes are merged from the persistent store container into the view context. Here's an example of the last exception backtrace. Exactly which internal methods that are called in - [NSFetchedResultsController(PrivateMethods) _core_managedObjectContextDidChange:] vary between crash reports but they all end up crashing from _NSArrayRaiseBoundException. The Core Data stack consists of one persistent store, one persistent store coordinator that the view context is set up to automatically merge changes from, and data is saved to disk from background context. persistentContainer.loadPersistentStores(...) viewContext = persistentContainer.viewContext viewContext.automaticallyMergesChangesFromParent = true backgroundContext = persistentContainer.newBackgroundContext() backgroundContext.mergePolicy = NSMergeByPropertyObjectTrumpMergePolicy backgroundClientContext = persistentContainer.newBackgroundContext() backgroundClientContext.mergePolicy = NSMergeByPropertyObjectTrumpMergePolicy Does anyone have any ideas what could be causing this? Thankful for any ideas or advice on how to investigate further.

I've been trying to setup a successful migration, but it keeps failing with this error: NSCloudKitMirroringDelegate are not reusable and should have a lifecycle tied to a given instance of NSPersistentStore. I can't find any information about this online. I added breakpoints throughout the code in willMigrate, and it originally failed on this line: try? context.save() I removed that, and it still failed. After I reload the app, it doesn't run the migration again and the app loads successfully. I figured since it crashed, it would keep trying, but I guess not. Here's how my migration is setup. enum MigrationV1ToV2: SchemaMigrationPlan { static var schemas: [any VersionedSchema.Type] { [SchemaV1.self, SchemaV2.self] } static var stages: [MigrationStage] { [stage] } static let stage = MigrationStage.custom( fromVersion: SchemaV1.self, toVersion: SchemaV2.self, willMigrate: { context in // Get cycles let cycles = try? context.fetch(FetchDescriptor<SchemaV1.Cycle>()) if let cycles { for cycle in cycles { // Create new recurring objects based on what's in the cycle for income in cycle.income { let recurring = SchemaV2.Recurring(name: income.name, frequency: income.frequency, kind: .income) recurring.addAmount(.init(date: cycle.startDate, amount: income.amount)) context.insert(recurring) } for expense in cycle.expenses { let recurring = SchemaV2.Recurring(name: expense.name, frequency: expense.frequency, kind: .expense) recurring.addAmount(.init(date: cycle.startDate, amount: expense.amount)) context.insert(recurring) } for savings in cycle.savings { let recurring = SchemaV2.Recurring(name: savings.name, frequency: savings.frequency, kind: .savings) recurring.addAmount(.init(date: cycle.startDate, amount: savings.amount)) context.insert(recurring) } for investment in cycle.investments { let recurring = SchemaV2.Recurring(name: investment.name, frequency: investment.frequency, kind: .investment) recurring.addAmount(.init(date: cycle.startDate, amount: investment.amount)) context.insert(recurring) } } //try? context.save() } else { print("The cycles were not able to be fetched.") } }, didMigrate: { context in // Get new recurring objects let newRecurring = try? context.fetch(FetchDescriptor<SchemaV2.Recurring>()) if let newRecurring { for recurring in newRecurring { // Get all recurring with the same name and kind let sameName = newRecurring.filter({ $0.name == recurring.name && $0.kind == recurring.kind }) // Add amount history to recurring object, and then remove matching for match in sameName { recurring.amountHistory.append(contentsOf: match.amountHistory) context.delete(match) } } //try? context.save() } else { print("The new recurring objects could not be fetched.") } } ) } Here's is my modelContainer in the app file. There is a fatal error occurring here that's crashing the app. var sharedModelContainer: ModelContainer = { let schema = Schema(versionedSchema: SchemaV2.self) let modelConfiguration = ModelConfiguration(schema: schema, isStoredInMemoryOnly: false) do { return try ModelContainer( for: schema, migrationPlan: MigrationV1ToV2.self, configurations: [modelConfiguration] ) } catch { fatalError("Could not create ModelContainer: \(error)") } }() Does anyone have any suggestions for this? EDIT: I found this error in the console that may be relevant. BUG IN CLIENT OF CLOUDKIT: Registering a handler for a CKScheduler activity identifier that has already been registered (com.apple.coredata.cloudkit.activity.export.8F7A1261-4324-40B4-B041-886DF36FBF0A). CloudKit setup failed because it couldn't register a handler for the export activity. There is another instance of this persistent store actively syncing with CloudKit in this process. And here is the fatal error Fatal error: Could not create ModelContainer: SwiftDataError(_error: SwiftData.SwiftDataError._Error.loadIssueModelContainer, _explanation: nil)

I'm using NSPersistentCloudKitContainer with Core Data and I receive errors because my iCloud space is full. The errors printed are the following: <CKError 0x280df8e40: "Quota Exceeded" (25/2035); server message = "Quota exceeded"; op = 61846C533467A5DF; uuid = 6A144513-033F-42C2-9E27-693548EF2150; Retry after 342.0 seconds>. I want to inform the user about this issue, but I can't find a way to access the details of the error. I'm listening to NSPersistentCloudKitContainer.eventChangedNotification, I receive a error of type .partialFailure. But when I want to access the underlying errors, the partialErrorsByItemID property on the error is nil. How can I access this Quota Exceeded error? import Foundation import CloudKit import Combine import CoreData class SyncMonitor { fileprivate var subscriptions = Set<AnyCancellable>() init() { NotificationCenter.default.publisher(for: NSPersistentCloudKitContainer.eventChangedNotification) .sink { notification in if let cloudEvent = notification.userInfo?[NSPersistentCloudKitContainer.eventNotificationUserInfoKey] as? NSPersistentCloudKitContainer.Event { guard let ckerror = cloudEvent.error as? CKError else { return } print("Error: \(ckerror.localizedDescription)") if ckerror.code == .partialFailure { guard let errors = ckerror.partialErrorsByItemID else { return } for (_, error) in errors { if let currentError = error as? CKError { print(currentError.localizedDescription) } } } } } // end of sink .store(in: &subscriptions) } }

One question I often see on DevForums and in my day DTS job is if a Core Data object managed by NSPersistentCloudKitContainer can be shared with other iCloud users. The answer is yes but you need to do it using CloudKit API directly because NSPersistentCloudKitContainer doesn’t support CloudKit shared database (CKContainer.sharedCloudDatabase) today. Assuming you have a Core Data object, let’s say a document, that you’d like to collaborate with your colleagues: You are the document owner and can use NSPersistentCloudKitContainer to fully manages the document and synchronize it across your devices. You can grab a CloudKit record associated with your document from NSPersistentCloudKitContainer using record(for:) or recordID(for:), and share it to your colleagues using UICloudSharingController. See our Sharing CloudKit Data with Other iCloud Users - https://developer.apple.com/documentation/cloudkit/sharing_cloudkit_data_with_other_icloud_users sample for how to share a CloudKit record. After accepting the sharing, your colleague, as a participant, can view or edit the shared document. The document resides in the participant’s CloudKit shared database and you have to manage it with your own code. When your colleague edits and saves the shared document, the changes go to the owner’s private database, and eventually synchronize to NSPersistentCloudKitContainer on the owner side.  As you can see, you need to implement #2 and #3 with your own code because NSPersistentCloudKitContainer can’t manage the data in the participant's shared database. If you have any difficulty after going through the above sample code, you can contact Apple’s DTS for help.

The NSPersistentCloudKitContainer synchronization between core data and iCloud was working fine with phone 15.1. Connected a new iPhone iOS 15.5, it gives error: CoreData: debug: CoreData+CloudKit: -[NSCloudKitMirroringDelegate managedObjectContextSaved:](2504): <NSCloudKitMirroringDelegate: 0x28198c000>: Observed context save: <NSPersistentStoreCoordinator: 0x2809c9420> - <NSManagedObjectContext: 0x2819ad520> 2022-12-05 13:32:28.377000-0600 r2nr[340:6373] [error] error: CoreData+CloudKit: -[NSCloudKitMirroringDelegate _importFinishedWithResult:importer:](1245): <PFCloudKitImporter: 0x2837dd740>: Import failed with error: Error Domain=NSCocoaErrorDomain Code=4864 "*** -[NSKeyedUnarchiver _initForReadingFromData:error:throwLegacyExceptions:]: incomprehensible archive (0x53, 0x6f, 0x6d, 0x65, 0x20, 0x65, 0x78, 0x61)" UserInfo={NSDebugDescription=*** -[NSKeyedUnarchiver _initForReadingFromData:error:throwLegacyExceptions:]: incomprehensible archive (0x53, 0x6f, 0x6d, 0x65, 0x20, 0x65, 0x78, 0x61)} CoreData: error: CoreData+CloudKit: -[NSCloudKitMirroringDelegate _importFinishedWithResult:importer:](1245): <PFCloudKitImporter: 0x2837dd740>: Import failed with error: Error Domain=NSCocoaErrorDomain Code=4864 "*** -[NSKeyedUnarchiver _initForReadingFromData:error:throwLegacyExceptions:]: incomprehensible archive (0x53, 0x6f, 0x6d, 0x65, 0x20, 0x65, 0x78, 0x61)" UserInfo={NSDebugDescription=*** -[NSKeyedUnarchiver _initForReadingFromData:error:throwLegacyExceptions:]: incomprehensible archive (0x53, 0x6f, 0x6d, 0x65, 0x20, 0x65, 0x78, 0x61)} I go back and try with my old iPhone iOS 15.1, gives same error.

Users have been reporting that the TestFlight version of my app (compiled with Xcode 26 Beta 6 17A5305f) is sometimes crashing on startup. Upon investigating their ips files, it looks like Core Data is locking up internally during its initialization, resulting in iOS killing my app. I have not recently changed my Core Data initialization logic, and it's unclear how I should proceed. Is this a known issue? Any recommended workaround? I have attached the crash stack below. Thanks! crash_log.txt

Hi, I was testing the new iOS 18 behavior where NSPersistentCloudKitContainer wipes the local Core Data store if the user logs out of iCloud, for privacy purposes. I ran the tests both with a Core Data + CloudKit app, and a simple one using SwiftData with CloudKit enabled. Results were identical in either case. In my testing, most of the time, the feature worked as expected. When I disabled iCloud for my app, the data was wiped (consistent with say the Notes app, except if you disable iCloud it warns you that it'll remove those notes). When I re-enabled iCloud, the data appeared. (all done through the Settings app) However, in scenarios when NSPersistentCloudKitContainer cannot immediately sync -- say due to rate throttling -- and one disables iCloud in Settings, this wipes the local data store and ultimately results in data loss. This occurs even if the changes to the managed objects are saved (to the local store) -- it's simply they aren't synced in time. It can be a little hard to reproduce the issue, especially since when you exit to the home screen from the app, it generally triggers a sync. To avoid this, I swiped up to the screen where you can choose which apps to close, and immediately closed mine. Then, you can disable iCloud, and run the app again (with a debugger is helpful). I once saw a message with something along the lines of export failed (for my record that wasn't synced), and unfortunately it was deleted (and never synced). Perhaps before NSPersistentCloudKitContainer wipes the local store it ought to force sync with the cloud first?

I have been working on an app for the past few months, and one issue that I have encountered a few times is an error where quick subsequent deletions cause issues with detached tasks that are triggered from some user actions. Inside a Task.detached, I am building an isolated model context, querying for LineItems, then iterating over those items. The crash happens when accessing a Transaction property through a relationship. var byTransactionId: [UUID: [LineItem]] { return Dictionary(grouping: self) { item in item.transaction?.id ?? UUID() } } In this case, the transaction has been deleted, but the relationship existed when the fetch occurred, so the transaction value is non-nil. The crash occurs when accessing the id. This is the error. SwiftData/BackingData.swift:1035: Fatal error: This model instance was invalidated because its backing data could no longer be found the store. PersistentIdentifier(id: SwiftData.PersistentIdentifier.ID(backing: SwiftData.PersistentIdentifier.PersistentIdentifierBacking.managedObjectID(0xb43fea2c4bc3b3f5 &lt;x-coredata://A9EFB8E3-CB47-48B2-A7C4-6EEA25D27E2E/Transaction/p1756&gt;))) I see other posts about this error and am exploring some suggestions, but if anyone has any thoughts, they would be appreciated.

We are trying to solve for the following condition with SwiftData + CloudKit: Lots of data in CloudKit Perform "app-reset" to clear data & App settings and start fresh. Reset data models with try modelContext.delete(model:_) myModel.count() confirms local deletion (0 records); but iCloud Console shows expectedly slow process to delete. Old CloudKit data is returning during the On Boarding process. Questions: • Would making a new iCloud Zone for each reset work around this, as the new zone would be empty? We're having trouble finding details about how to do this with SwiftData. • Would CKSyncEngine have a benefit over the default SwiftData methods? Open to hearing if anyone has experienced a similar challenge and how you worked around it!

When a user first downloads my application they are prompted to sign into their apple account via a pop up. I have not had this pop up previously, I believe the change occurred after iOS18. I have functions that do a few things: Retrieves userRecordID Retrieves a userprofile(via userrecordid) from cloudkit.

I have a simple model @Model final class Movie: Identifiable { #Index\<Movie\>(\[.name\]) var id = UUID() var name: String var genre: String? init(name: String, genre: String?) { self.name = name self.genre = genre } } I turned on SQL debugging by including '-com.apple.CoreData.SQLDebug 3' argument on launch. When I fetch the data using the following code, it selects 3 records initially, but then also selects each record individually even though I am not referencing any other attributes. var fetchDescriptor = FetchDescriptor\<Movie\>() fetchDescriptor.propertiesToFetch = \[.id, .name\] fetchDescriptor.fetchLimit = 3 do { print("SELECT START") movies = try modelContext.fetch(fetchDescriptor) print("SELECT END") } catch { print("Failed to load Movie model.") } I see it selecting the 3 rows initially, but then it selects each one separately. Why would it do this on the initial fetch? I was hoping to select the data that I want to display and let the system select the entire record only when I access a variable that I did not initially fetch. CoreData: annotation: fetch using NSSQLiteStatement <0x600002158af0> on entity 'Movie' with sql text 'SELECT 1, t0.Z_PK, t0.ZID, t0.ZNAME FROM ZMOVIE t0 LIMIT 3' returned 3 rows with values: ( "<NSManagedObject: 0x600002158d70> (entity: Movie; id: 0xa583c7ed484691c1 <x-coredata://71E60F4C-1A40-4DB7-8CD1-CD76B4C11949/Movie/p1>; data: <fault>)", "<NSManagedObject: 0x600002158d20> (entity: Movie; id: 0xa583c7ed482691c1 <x-coredata://71E60F4C-1A40-4DB7-8CD1-CD76B4C11949/Movie/p2>; data: <fault>)", "<NSManagedObject: 0x600002158f00> (entity: Movie; id: 0xa583c7ed480691c1 <x-coredata://71E60F4C-1A40-4DB7-8CD1-CD76B4C11949/Movie/p3>; data: <fault>)" ) CoreData: annotation: fetch using NSSQLiteStatement <0x600002154d70> on entity 'Movie' with sql text 'SELECT 0, t0.Z_PK, t0.Z_OPT, t0.ZGENRE, t0.ZID, t0.ZNAME FROM ZMOVIE t0 WHERE t0.Z_PK = ? ' returned 1 rows CoreData: annotation: with values: ( "<NSSQLRow: 0x600000c89500>{Movie 1-1-1 genre=\"Horror\" id=4C5CB4EB-95D7-4DC8-B839-D4F2D2E96ED0 name=\"A000036\" and to-manys=0x0}" ) This all happens between the SELECT START and SELECT END print statements. Why is it fulfilling the faults immediately?

Here’s the situation: • You’re downloading a huge list of data from iCloud. • You’re saving it one by one (sequentially) into SwiftData. • You don’t want the SwiftUI view to refresh until all the data is imported. • After all the import is finished, SwiftUI should show the new data. The Problem If you insert into the same ModelContext that SwiftUI’s @Environment(.modelContext) is watching, each insert may cause SwiftUI to start reloading immediately. That will make the UI feel slow, and glitchy, because SwiftUI will keep trying to re-render while you’re still importing. How to achieve this in Swift Data ?

Since running on iOS 14b1, I'm getting this in my log (I have Core Data logging enabled): error: Store opened without NSPersistentHistoryTrackingKey but previously had been opened with NSPersistentHistoryTrackingKey - Forcing into Read Only mode store at 'file:///private/var/mobile/Containers/Shared/AppGroup/415B75A6-92C3-45FE-BE13-7D48D35909AF/StoreFile.sqlite' As far as I can tell, it's impossible to open my store without that key set - it's in the init() of my NSPersistentContainer subclass, before anyone calls it to load stores. Any ideas?

My use case is the following: Every user of my app can create as an owner a set of items.  These items are private until the owner invites other users to share all of them as participant. The participants can modify the shared items and/or add other items. So, sharing is not done related to individual items, but to all items of an owner. I want to use CoreData & CloudKit to have local copies of private and shared items. To my understanding, CoreData & CloudKit puts all mirrored items in a special zone „com.apple.coredata.cloudkit.zone“. So, this zone should be shared, i.e. all items in it. In the video it is said that NSPersistentCloudKitContainer uses Record Zone Sharing optionally in contrast to hierarchically record sharing using a root record. But how is this done? Maybe I can declare zone „com.apple.coredata.cloudkit.zone“ as a shared zone?

I'm writing some tests to confirm the behavior of my app. White creating a model actor to delete objects I realized that ModelContext.model(for:) does return objects that are deleted. I was able to reproduces this with this minimal test case: @Model class Activity { init() {} } struct MyLibraryTests { let modelContainer = try! ModelContainer( for: Activity.self, configurations: ModelConfiguration( isStoredInMemoryOnly: true ) ) init() throws { let context = ModelContext(modelContainer) context.insert(Activity()) try context.save() } @Test func modelForIdAfterDelete() async throws { let context = ModelContext(modelContainer) let id = try context.fetch(FetchDescriptor<Activity>()).first!.id context.delete(context.model(for: id) as! Activity) try context.save() let result = context.model(for: id) as? Activity #expect(result == nil) // Expectation failed: (result → MyLibrary.Activity) == nil } @Test func fetchDescriptorAfterDelete() async throws { let context = ModelContext(modelContainer) let id = try context.fetch(FetchDescriptor<Activity>()).first!.id context.delete(context.model(for: id) as! Activity) try context.save() let result = try context.fetch( FetchDescriptor<Activity>(predicate: #Predicate { $0.id == id }) ).first #expect(result == nil) } } Here I create a new context, insert an model and save it. The test modelForIdAfterDelete does fail, as result still contains the deleted object. I also tried to check #expect(result!.isDeleted), but it is also false. With the second test I use a FetchDescriptor to retrieve the object by ID and it correctly returns nil. Shouldn't both methods use a consistent behavior?

I am trying to add a custom JSON DataStore and DataStoreConfiguration for SwiftData. Apple kindly provided some sample code in the WWDC24 session, "Create a custom data store with SwiftData", and (once updated for API changes since WWDC) that works fine. However, when I try to add a relationship between two classes, it fails. Has anyone successfully made a JSONDataStore with a relationship? Here's my code; firstly the cleaned up code from the WWDC session: import SwiftData final class JSONStoreConfiguration: DataStoreConfiguration { typealias Store = JSONStore var name: String var schema: Schema? var fileURL: URL init(name: String, schema: Schema? = nil, fileURL: URL) { self.name = name self.schema = schema self.fileURL = fileURL } static func == (lhs: JSONStoreConfiguration, rhs: JSONStoreConfiguration) -> Bool { return lhs.name == rhs.name } func hash(into hasher: inout Hasher) { hasher.combine(name) } } final class JSONStore: DataStore { typealias Configuration = JSONStoreConfiguration typealias Snapshot = DefaultSnapshot var configuration: JSONStoreConfiguration var name: String var schema: Schema var identifier: String init(_ configuration: JSONStoreConfiguration, migrationPlan: (any SchemaMigrationPlan.Type)?) throws { self.configuration = configuration self.name = configuration.name self.schema = configuration.schema! self.identifier = configuration.fileURL.lastPathComponent } func save(_ request: DataStoreSaveChangesRequest<DefaultSnapshot>) throws -> DataStoreSaveChangesResult<DefaultSnapshot> { var remappedIdentifiers = [PersistentIdentifier: PersistentIdentifier]() var serializedData = try read() for snapshot in request.inserted { let permanentIdentifier = try PersistentIdentifier.identifier(for: identifier, entityName: snapshot.persistentIdentifier.entityName, primaryKey: UUID()) let permanentSnapshot = snapshot.copy(persistentIdentifier: permanentIdentifier) serializedData[permanentIdentifier] = permanentSnapshot remappedIdentifiers[snapshot.persistentIdentifier] = permanentIdentifier } for snapshot in request.updated { serializedData[snapshot.persistentIdentifier] = snapshot } for snapshot in request.deleted { serializedData[snapshot.persistentIdentifier] = nil } try write(serializedData) return DataStoreSaveChangesResult<DefaultSnapshot>(for: self.identifier, remappedIdentifiers: remappedIdentifiers) } func fetch<T>(_ request: DataStoreFetchRequest<T>) throws -> DataStoreFetchResult<T, DefaultSnapshot> where T : PersistentModel { if request.descriptor.predicate != nil { throw DataStoreError.preferInMemoryFilter } else if request.descriptor.sortBy.count > 0 { throw DataStoreError.preferInMemorySort } let objs = try read() let snapshots = objs.values.map({ $0 }) return DataStoreFetchResult(descriptor: request.descriptor, fetchedSnapshots: snapshots, relatedSnapshots: objs) } func read() throws -> [PersistentIdentifier : DefaultSnapshot] { if FileManager.default.fileExists(atPath: configuration.fileURL.path(percentEncoded: false)) { let decoder = JSONDecoder() decoder.dateDecodingStrategy = .iso8601 let data = try decoder.decode([DefaultSnapshot].self, from: try Data(contentsOf: configuration.fileURL)) var result = [PersistentIdentifier: DefaultSnapshot]() data.forEach { s in result[s.persistentIdentifier] = s } return result } else { return [:] } } func write(_ data: [PersistentIdentifier : DefaultSnapshot]) throws { let encoder = JSONEncoder() encoder.dateEncodingStrategy = .iso8601 encoder.outputFormatting = [.prettyPrinted, .sortedKeys] let jsonData = try encoder.encode(data.values.map({ $0 })) try jsonData.write(to: configuration.fileURL) } } The data model classes: import SwiftData @Model class Settings { private(set) var version = 1 @Relationship(deleteRule: .cascade) var hack: Hack? = Hack() init() { } } @Model class Hack { var foo = "Foo" var bar = 42 init() { } } Container: lazy var mainContainer: ModelContainer = { do { let url = // URL to file let configuration = JSONStoreConfiguration(name: "Settings", schema: Schema([Settings.self, Hack.self]), fileURL: url) return try ModelContainer(for: Settings.self, Hack.self, configurations: configuration) } catch { fatalError("Container error: \(error.localizedDescription)") } }() Load function, that saves a new Settings JSON file if there isn't an existing one: @MainActor func loadSettings() { let mainContext = mainContainer.mainContext let descriptor = FetchDescriptor<Settings>() let settingsArray = try? mainContext.fetch(descriptor) print("\(settingsArray?.count ?? 0) settings found") if let settingsArray, let settings = settingsArray.last { print("Loaded") } else { let settings = Settings() mainContext.insert(settings) do { try mainContext.save() } catch { print("Error saving settings: \(error)") } } } The save operation creates a JSON file, which while it isn't a format I would choose, is acceptable, though I notice that the "hack" property (the relationship) doesn't have the correct identifier. When I run the app again to load the data, I get an error (that there wasn't room to include in this post). Even if I change Apple's code to not assign a new identifier, so the relationship property and its pointee have the same identifier, it still doesn't load. Am I doing something obviously wrong, or are relationships not supported in custom data stores?