Lets say I have the following scenario...
ViewController1 loads a Person object from a Realm on the main thread and passes it to ViewController2. User interaction in ViewController2 causes the same Person object to change, but I only want to persist the changes once the User has pressed "Save".
Currently, when changing the passed Person object in ViewController2 a runtime error is thrown saying changes to an object need to be made in a Write block. This makes sense, but in this scenario I don't actually want to persist the changes right away.
Is there a way to detach an Object from a Realm to avoid these
checks?
If there isn't, what would be a suggested work around? (Copying the Object to a new instance? Tracking the changes to the Object separately and applying them later? Both seem pretty messy.)
Right now, you can make a 'standalone' copy of your object, via Object(value: existingObject) -- that'll probably be the simplest solution for now, until Realm adds something like nested transactions that will make undoing an arbitrary number of changes easier.
Realm has added a freeze() function for Realm Objects on RealmSwift 5.+ versions.
Imagine we have a Realm object Edible with a property name. Previously you weed need to hold a ThreadSafeReference to the object, get a realm on the other thread and unwrap the reference. Meh.
If you didn't, it would crash:
// Code running on main thread
let edible = realm.objects(Edible.self)[0]
DispatchQueue.global(qos: .background) {
let name = edible.name // Realm accessed from incorrect thread error
}
How to do detach an object in RealmSwift 5.+:
let edible = realm.objects(Edible.self)[0].freeze()
DispatchQueue.global(qos: .background) {
let name = edible.name // No longer crashes
}
Keep in mind as of Jan 11 2021, the freeze() function can create a lot of unknown weird errors affecting the users, as mentioned repeatedly in the issues for the new Realm version. For now, stay away from it. Use deep copying.
Related
I'm writing an app using Realm data persistence of certain objects.
In an attempt to clean up/remodel my code (getting realm.writes out of the Views and Controllers), I tried to put them directly in the persisted object class.
The logic is basically this:
class PersistedObject: Object {
public var data: String {
get { _saved_data }
set {
do { try realm?.write { _saved_data = newValue }
} catch { print(error) }
}
}
#objc dynamic private var _saved_data = "hello there"
}
This way, I'd be able to access and rewrite realm object properties from view controllers, without needing realm.writes directly in there. That's the idea, anyway.
This works sometimes. Other times, the app crashes with the error...
"Realm accessed from incorrect thread"
...which is what I'm currently trying to solve.
This is my first iOS app and my first time using Realm.
Does it make sense to organize the code like this (I've found little in terms of support in this approach, but also generally little at all, in terms of MVC best-practices when working with Realm)
If it does make sense, how can I solve the problem with accessing Realm from the incorrect thread, while still doing the realm.writes directly in the object class?
Thanks in advance! :)
Simon
There is no sense to organize code like this. You will be able to write only from same thread it was created
to modify objects from different thread you can use ThreadSafeReference for example
You're not going to want to do that.
There's no reason not to realm.write whenever you want to write to realm - that's what it's there for. This pattern works:
// Use them like regular Swift objects
let myDog = Dog()
myDog.name = "Rex"
myDog.age = 1
// Get the default Realm
let realm = try! Realm()
// Persist your data easily
try! realm.write {
realm.add(myDog)
}
Obviously there should be better error catching in the above code.
Another downside is if you want to write 10 objects, they are written as soon as the data property is set - what if there are three vars you want to set and heep in memory before writing it? e.g. your user is creating a list of items in your app - if the user decides not to do that and hit's Cancel, you would then have to hit the database again to delete the object(s).
Consider a case where you want to write 10 objects 'at the same time'?
realm.add([obj0, obj1, obj2...])
is a lot cleaner.
Another issue comes up if you want to guarantee objects are written within a transaction - either it all succeeds or all fails. That can't be done with your current object.
The last issue is that often you'll want to instantiate an object and add some data to it, populating the object before writing to realm. With the code in the question, you're writing it as soon as data is populated. You would have to add that same code to every property.
I want to solve next problem:
I would like to work with some NSManagedObject in context and change some properties in runtime, but without telling SQLite about any changes in it.
I just want to save NSManagedObject to database when I hit save button or similar.
As I found out from source code demo we need to use beginUnsafe for this purposes (maybe I am wrong)
func unstoredWorkout() -> WorkoutEntity {
let transaction = CoreStore.beginUnsafe()
let workout = transaction.create(Into<WorkoutEntity>())
return workout
}
let workout = unstoredWorkout()
workout.muscles = []
Now when I try to update workout.muscles = [] app crashes with error:
error: Mutating a managed object 0x600003f68b60 <x-coredata://C00A3E74-AC3F-47FD-B656-CA0ECA02832F/WorkoutEntity/tC3921DAE-BA43-45CB-8271-079CC0E4821D82> (0x600001c2da90) after it has been removed from its context.
My question how we can create object without saving it and how we can save it then when we modify some properties and avoid this crash as well.
The reason for the crash is that your transaction only lives in your unstoredWorkout() method, so it calles deinit, which resets the context (and deletes all unsaved objects).
You have to retain that unsafe transaction somewhere to keep your object alive - such as in the viewcontroller that will eventually save the changes.
But I would rather encourage you to think about that if you really want to do that. You might run into other synchronization issues with various context or other async transactions alive, like when API calls are involved.
In my program, it uses both of
DispatchQueue.global(qos: .background)
and
self.concurrentQueue.sync(flags: .barrier)
to deal with the background multithread issues.
It is swift 3 so I use the latest way to get the childContext:
lazy var context: NSManagedObjectContext = {
return (UIApplication.shared.delegate as! AppDelegate).persistentContainer.newBackgroundContext()
}()
I also enable -com.apple.CoreData.ConcurrencyDebug 1 to debug
Then the problem occurs:
1, When there's an API call and in the callback block (background thread), I need to fetch the core data, edit, then save. I tried to use self.context from the code above to call performBlockAndWait and do save inside of this block. The whole process goes fine but when I try to access my result outside of this block but inside of the callback block, the error occurs. I have also tried to get the objectId and getObjectById by both self.context and self.context.parent and the error occurs on this line. What did I do wrong and how should I do this? since I need to use the result everywhere in many different thread (not context).
2, I read a post says that I need one context per thread, then in my case, how do I determine which exact thread it is if it's a call back from API call and do I really need to do this?
3, You might ask that why do I need a privateConcurrentType, because my program has things need to be running in background thread so that it has to do it this way, (read from other post), is this right?
4, Even in my question 1, get object by passing objectId to different Context still not working in my case. Let's assume this is the proper way. How am I gonna manage passing so many objectID throughout my entire program in different thread without being super messy? To me this sounds crazy but I suppose there's a much cleaner and easier way to deal with this.
5, I have read many posts some are pretty old (before swift 3), they have to do childContext.save then parentContext.save, but since I use the code above (swift 3 only). It seems that I can do childContext.save only to make it work? Am I right?
Core data in general is not multithreading friendly. To use it on concurrent thread I can assume only bad things will happen. You may not simply manipulate managed objects outside the thread on which the context is.
As you already mentioned you need a separate context per thread which will work in most cases but by my experience you only need one background context which is read-write and a single main thread read-only context that is used for fetch result controllers or other instant fetches.
Think of a context as some in-memory module that communicates with the database (a file). Fetched entities are shared within the context but are not shared between contexts. So you can modify pretty much anything inside the context but that will not show in the database or other contexts until you save the context into the database. And if you modify the same entity on 2 contexts and then save them you will get a conflict which should be resolved by you.
All of these then make quite a mess in the code logic and so multiple contexts seem like something to avoid. What I do is create a background context and then do all of the operations on that context. Context has a method perform which will execute the code on its own thread which is not main (for background context) and this thread is serial.
So for instance when doing a smart client I will get a response from server with new entries. These are parsed on the fly and I perform a block on context to get all the corresponding objects in the database and create the ones that do not exist. Then copy the data and save the context into database.
For the UI part I do similar. Once an entry should be saved I either create or update the entity on the background context thread. Then usually do some UI stuff on completion so I have a method:
public func performBlockOnBackgroundContextAndReturnOnMain(block: #escaping (() -> Void), main: #escaping (() -> Void)) {
if let context = context {
context.perform {
block()
DispatchQueue.main.async(execute: { () -> Void in
main()
})
}
}
}
So pretty much all of the core data logic happens on a single thread which is in background. For some cases I do use a main context to get items from fetch result controller for instance; I display a list of objects with it and once user selects one of the items I refetch that item from the background context and use that one in the user interface and to modify it.
But even that may give you trouble as some properties may be loaded lazily from database so you must ensure that all the data you need will be loaded on the context and you may access them on the main thread. There is method for that but I rather use wrappers:
I have a single superclass for all the entities in the database model which include id only. So I also have a superclass wrapper which has all the logic to work with the rest of wrappers. What I am left with in the end is that for each of the subclass I need to override 2 mapping methods (from and to) managed object.
It might seem silly to create additional wrappers and to copy the data into memory from managed object but the thing is you need to do that for most of the managed objects anyway; Converting NSData to/from UIImage, NSDate to/from Date, enumerations to/from integers or strings... So in the end you are more or less just left with strings that are copied 1-to-1. Also this makes it easy to have the code that maps the response from your server in this class or any additional logic where you will have no naming conflicts with managed objects.
When I delete an NSMangedObject from the database, what happens to local variables who were assigned to it?
For example, I have a simple NSManagedObject:
class MyManagedObject: NSManagedObject {
#NSManaged var name: String
}
And then in my ViewController, I pull it out of the database, and assign it locally:
class ViewController: UIViewController {
var myManagedObject: MyManagedObject!
}
Then I delete it from the database.
If print the object name I get the following in the console
print("myManagedObject.name = \(myManagedObject.name)")
//prints: "myManagedObject.name = "
As if the object isn't there? But if I turn the variable into an optional and check it for nil, I am told it's not nil.
I'm not quite sure how to reconcile this in my mind. There seems to be something pointing to the local variable, but its properties are gone.
If I have many disparate UI objects that rely on that object for its properties, I can't assume that there is some local deep copy of it in memory?
Here is more complete code:
In viewDidLoad I create the new object, save the context, fetch the object, then assign it locally.
class ViewController: UIViewController {
var myManagedObject: MyManagedObject!
override func viewDidLoad() {
super.viewDidLoad()
//1 Create the new object
let newObject = NSEntityDescription.insertNewObject(forEntityName: "MyManagedObject", into: coreDataManager.mainContext) as! MyManagedObject
newObject.name = "My First Managed Object"
//2 Save it into the context
do {
try coreDataManager.mainContext.save()
} catch {
//handle error
}
//3 Fetch it from the database
let request = NSFetchRequest<MyManagedObject>(entityName: "MyManagedObject")
do {
let saved = try coreDataManager.mainContext.fetch(request)
//4 Store it in a local variable
self.myManagedObject = saved.first
} catch {
//handle errors
}
}
}
At this point if I print the local variable's name property, I get the correct response:
print("The object's name is: \(myManagedObject.name)")
//prints: The object's name is: My First Managed Object
So, now I delete it from the database:
if let storedObject = myManagedObject {
coreDataManager.mainContext.delete(storedObject)
do {
try coreDataManager.mainContext.save()
} catch {
//handle error
}
}
But now, when I print I get the strangest output:
print("myManagedObject.name = \(myManagedObject.name)")
//prints: "myManagedObject.name = "
This is totally not the way I'm expecting memory to work. If I create a instance of a class Foo, and then pass that instance around to different objects, it's the same instance. It only goes away once no one is pointing to it.
In this case--- what is the variable, myManagedObject? It's not nil. And what is the string, name? Is it an empty string? Or is it some other weird meta-type?
The main thing what you are probably looking here for is the core data context. The context is a connection between your memory and the actual database.
Whenever you fetch the data you fetch it through context. These are managed objects which can be modified or even deleted. Still none of these really happen on the database until you save the context.
When you delete an object it is marked for deletion but it is not deleted from the memory, it must not be since if nothing else it will still be used by the context to actually delete the object from the database itself.
What happens to the managed object once you call to delete it is pretty much irrelevant, even if documented it may change as it is a part of the framework. So it is your responsibility to check these cases and to refetch the objects once needed. So you must ensure your application has a proper architecture and uses core data responsibly.
There are very many ways on how you use your database and more or less any of them has a unique way of using it optimally. You need to be more specific on what you are doing and where do you see potential issues so we can get you on the right track.
To give you an example consider data synchronization from remote server. Here you expect that data can be synchronized at any time no matter what user is doing or what part of the application he is.
For this I suggest you have a single context which operates on a separate thread. All the managed objects should be wrapped and its properties copied once retrieved from database. On most entities you would have something like:
MyEntity.findAll { items in
...the fetch happens on context thread and returns to main, items are wrappers
}
MyEntity.find(id: idString, { item in
...the fetch happens on context thread and returns to main, items are wrappers
})()
Then since you do not have any access to the managed object directly you need some kind of method to copy the data to the managed object like:
myEntityInstance.commit() // Copies all the data to core data object. The operation is done on a context thread. A callback is usually not needed
And then to save the database
MyEntity.saveDatabse {
... save happens on the context thread and callback is called on main thread
}
Now the smart part is that saveDatabse method will report to a delegate that changes have been made. A delegate is usually the current view controller so it makes sense to have a superclass like DataBaseViewController which on view did appear assigns itself as a delegate MyEntity.delegate = self, on view did load calls some method reloadData and in the databaseDidChange delegate method calls reloadData and same in viewWillAppear.
Now each of your view controllers that are subclass of DataBaseViewController will override the reloadData and in that method you will fetch the data from the database again. Either you are fetching all items or a single one. So for those singe ones you need to save the id of the object and fetch it again by that id. If the returned object is nil then item was deleted so you catch the issue you seem to be mentioning.
All of these things are oversimplified but I hope you get a basic idea about core data and how to use it. It is not easy, it never was and it most likely never will be. It is designed for speed to be able to access the data even from a very large database in shortest time possible. The result is that it might not be very safe.
I've just starting to using Realm and feel it's very good, fast except one thing: delete an object in Realm is easily cause an exception.
Is there any way I can delete an object in Realm safety?
In my project, I usually have to create, update, delete hundred objects on the background thread. The issue is:
If the app currently display/using one object on the main thread
In the background, I delete that object.
=> On the main thread will cause an exception when using that object's properties.
I know Realm has isInvalid method to check, but I cannot add the check in every assign properties code, it's look not good.
So, as of now, what I do is: instead of actually delete, I have a property call "deleted", and in delete, I only update that value. And on the UI, I will filtered out objects which have deleted = true
I wonder is there any way better to do this?
This is intended functionality. If a background thread deletes a Realm Object, the next time you try and access that object from any thread, an exception will be thrown.
In order to handle this, Realm provides a rich notification system that you can use to automatically receive alerts of when the contents of a Realm database have been changed and to update the UI accordingly.
If you've got a view controller that is displaying the contents of a single Realm Object, you could implement a system to be notified of any changes being made to your Realm database, and then check to make sure your object is still valid:
class MyViewController : UIViewController {
var myModel: Object = nil
var notificationToken: NotificationToken? = nil
init(model: Object) {
self.myModel = model
}
override fun viewDidLoad() {
super.viewDidLoad()
notificationToken = myModel.realm.addNotificationBlock { notification, realm in
guard myModel.invalidated == false else {
// The object has been deleted, so dismiss this view controller
}
}
}
deinit() {
notificationToken?.stop()
}
}
That notification block will be triggered each time a write transaction modified something in that particular Realm file (Even on background threads), which gives you a chance to check to see if your particular Realm Object in that UI hasn't been deleted. If it has, then you can simply dismiss the UI.
Depending on your specific needs, there is also a more fine-grained notification system you can use to specifically track changes to Realm Objects that were part of the results of a query. There's sample code for that in the Collection Notifications of the Realm documentation.
Please let me know if you need additional clarification! :)