I've got a couple background tasks to process while still keeping the UI responsive. I started down the path of creating and managing threads, but soon came across the NSOperations. Sounds like a better solution . . .
However, I cannot seem to get a reference to the NSOperationQueue. The Threading Programming Guide suggests the following
#implementation MyCustomClass
- (void)launchTaskWithData:(id)data
{
NSInvocationOperation* theOp = [[NSInvocationOperation alloc] initWithTarget:self
selector:#selector(myTaskMethod:) object:data];
// Add the operation to the internal operation queue managed by the application delegate.
[[MyAppDelegate sharedOperationQueue] addOperation:theOp];
}
// This is the method that does the actual work of the task.
- (void)myTaskMethod:(id)data
{
// Perform the task.
}
. . . but I (and more importantly, the compiler) don't see the 'sharedOperationQueue' message when I implement the code in my app.
What am I missing?? Is sharedOperationQueue deprecated and no longer available? How can I get an NSOperationQueue reference?
sharedOperationQueue is not part of the official API. It is a custom method you are supposed to implement yourself, in this example as a class method in your app delegate class. The method should create and return NSOperationQueue or, if it already has created the queue, simply return the existing one.
How you implement this in your case is up to you. You don't have to follow the example. Simple create an operation queue with alloc/init (no magic here) and store a reference to it in a property so that you can later release it when you no longer need it.
Cocoa Is My Girlfriend has a good tutorial, this will help you to use NSOperationQueue.
You have to declare application delegate before adding operation:
AppDelegateClass * MyAppDelegate = [NSApp delegate];
Related
how can I stop 2nd thread to access same method that is being used by first thread?
One option is to use #synchronized in the method.
- (NSString *)someMethod {
#synchronized(self) {
// do some work
}
}
It allows the method to be called but will synchronize on itself and protect it's work (and more importantly the data it's working on).
Have a look at NSLock.
https://developer.apple.com/library/mac/#documentation/Cocoa/Reference/Foundation/Classes/NSLock_Class/Reference/Reference.html
Just found it when looking for a C# lock statement, it appears to do the same thing..
There is also another SO thread relating to synchronisation:
How does #synchronized lock/unlock in Objective-C?
I have a simple query that I'd like cleared up by someone... Is it bad-practice to retain self?
I have a server request object that I'd like to make. I'd like to be able to use it in the following fashion:
ARequest *request = [ARequest request: someParam];
request.delegate = self;
[request begin];
In order for the object not to self destruct as soon as the autorelease pool is drained, I imagine I need to call a retain in it's init method and then a release once the server response has been received, processed and delivered to it's delegate.
However, something is raising a warning bell in my head with this approach. Better ways to do it?
There is nothing wrong with retaining self, as long as you release it at some well-defined point in accordance with normal memory management protocol. If an object requires itself to exist until some condition is met, it should take responsibility for that, in the same way as it does for any other object it requires to continue existing.
Introducing otherwise extraneous manager objects or foisting the responsibility off on the object’s owner for superstitious reasons would be the real anti-pattern here.
(The equivalent approach in garbage-collected code would be for the object to exclude itself from garbage collection while results are pending, or root it through a collection of some sort if you dislike that idea.)
It's not unheard-of, but it is somewhat uncommon. The main way I've seen it used (and used it myself) is when you're dealing with some sort of semi-synchronous object (by semi-synchronous I mean that it does not block the main thread, but it also does not execute on a background thread; an NSURLConnection would fit this bill). For example, I wrote a subclass of NSWindowController that was specifically for displaying a window as a sheet and for invoking some certain delegate callbacks. Basically, you'd alloc/init a new sheet controller and invoke beginSheetForWindow:. This would run the sheet semi-synchronously, and then invoke an appropriate callback when the sheet was dismissed.
Since the invoking object doesn't necessarily "own" the sheet (think of it as a Mac version of a modal view controller on iOS), the sheet controller does [self retain] immediately before showing the sheet, and [self release] immediately after cleaning up and invoking callbacks. The purpose behind this was to ensure that the controller object would stick around until the sheet was done. (The sheet, IIRC, was retained by the runloop, but I also needed the controller to stick around)
Like I said, it's very rare to come across a situation where you would want to [self retain], but it's not impossible. However, as a general rule of thumb, if you think that you need to [self retain], you may want to think again.
Easiest way to do this would be to create an iVar for your request, retain the request when you start it and release it when the last delegate method is called.
Is ARequest a class you created? Does it create a new thread to asynchronously submit the request?
I once did the same thing as you. I wrote a Category-Method on NSString to send it it to a server, that will print it. In the Category-Method I had to call [self retain], so that the callback methods could be a NSString-Categroy-Method to.
I felt so bad about it, that I rewrote everything to use a Singleton, that is accessed by the Category-Method. So the Singleton will retain the string as long as necessary.
Here's the issue – I followed along with the Apple lazy-load image sample code to handle my graphical tables. It works great. However, my lazy-load image tables are being stacked within a navigation controller so that you can move in and out of menus. While all this works, I'm getting a persistent crash when I move into a table then move immediately back out of it using the "back" button. This appears to be a result of the network connections loading content not being closed properly, or calling back to their released delegates. Now, I've tried working through this and carefully setting all delegates to nil and calling close on all open network connections before releasing a menu. However, I'm still getting the error. Also – short posting my entire application code into this post, I can't really post a specific code snippet that illustrates this situation.
I wonder if anyone has ideas for tasks that I may be missing? Do I need to do anything to close a network connection other than closing it and setting it to nil? Also, I'm new to debugging tools – can anyone suggest a good tool to use to watch network connections and see what's causing them to fail?
Thanks!
Have you run it through the debugger (Cmd-Y)? Does it stop at the place where the crash is happening? That should show you in code where the issue is happening. I'm betting the issue has to do with over-releasing something rather than cleaning up connections. Are you getting EXC_BAD_ACCESS? Check any delegates and make sure they are nil when -viewWillDisappear gets called. That way, if anything tries to call back to a delegate, it will just be a no-op.
You may also want to try enabling zombies (NSZombieEnabled) which will tell you when an object that has been released is being accessed again. It's very helpful in finding over-released objects.
Ah ha... after a large zombie hunt (thanks, Matt Long), I discovered that the issue stems from an error in Apple's LazyTableImages sample code. That example provides the following implementation for canceling all image loads, which I turned into a general-purpose stopAllImageLoads method...
From RootViewController.m in LazyTableImages sample code:
- (void)didReceiveMemoryWarning
{
[super didReceiveMemoryWarning];
// terminate all pending download connections
NSArray *allDownloads = [self.imageDownloadsInProgress allValues];
[allDownloads performSelector:#selector(cancelDownload)];
}
There is in error in the last line of the above method where performSelector is called on an array of objects. The above implementation calls the selector on the array itself, rather that on each object in the array. Therefore, that last line should be this:
[allDownloads makeObjectsPerformSelector:#selector(cancelDownload)];
Once that line was changed, everything else fell into place. It turns out I wasn't calling my stopAllImageLoads method where I meant to – I had disabled it at one point because it was causing an error. Once that was back in place, the memory issues cleared up because image loads were successfully canceled before the table delegate was released.
Thanks all for your help.
If you're doing ANY asynchronous function (network requests, Core Location updates, etc), you run the risk that your view controller that is the delegate of that action is deallocated by the time the async function returns. i.e. you back out of the view and take the delegate target away from the background process. I've dealt with this several times.
Here's what you do. Use the ASIHTTPRequest library (which you should be doing anyway--it's brilliant). Create a synthesized property to hold your request. Then in viewWillDisappear, call -cancel on your request. To be safe, I also set its delegate to nil, but that should be unnecessary.
Here's a sketch of what you want to do. Note I typed this right here, haven't syntax-checked it or anything.
#implementation MyViewController
#synthesize req //this is an ASIHTTPRequest *req.
-(void)viewDidLoad
{
//make an NSURL object called myURL
self.req = [ASIHTTPRequest requestWithURL:myURL];
self.req.delegate = self;
[self.req startAsynchronous];
}
-(void)viewWillDisappear
{
[self.req cancel];
}
-(void)requestFinished:(ASIHTTPRequest *)request
{
NSString *string = [request responseString];
}
I have created a library which can download JSON data which is then placed into an NSDictionary. I wrap this class with a simple Twitter engine which allows me to pull my friends timeline, post an update and post an update with my GPS location. From my limited experience with Objective-C the way to connect everything is with delegation. I set a delegate property which calls back the asynchronous result to either a selector or a method signature. I can even create an optional or required interface on the delegate which will allow Xcode to assist me a little with implementing the delegate. To learn about using delegates in Objective-C I created this simple project.
http://www.smallsharptools.com/downloads/ObjC/Delegates.zip
It defines a Worker class which allows you to initialize the class with a delegate. When the work is done with the doWork method it looks for a method signature on the delegate to send a message back to it. It uses the following code.
if([[self delegate] respondsToSelector:#selector(workFinished:)]) {
NSString *msg = #"That's it? Easy!";
[[self delegate] workFinished:msg];
}
It looks for the workFinished: method to pass back a message. I declared this method signature as an optional interface with the following code in the header, Worker.h.
#protocol WorkerNotifications
#optional
- (void) workFinished: (NSString *) msg;
#end
You can see the rest of the project from the download for all of the details. But these 2 code snippets show how this delegation pattern works. But with the Twitter class I need to know the context of the method which started an asynchronous action which leads to a callback to a delegate method. If I call the sendUpdate method more than once from the calling class, how I am supposed to know the context of the callback?
Normally with a language like JavaScript, Java or C# I would create an inline closure or anonymous class which would have access to the starting context, but that is not possibly currently with Objective-C on the iPhone. I found that this question was already asked and answered on StackOverflow.
Anonymous delegate implementation in Objective-C?
So what I have done is skip the optional interface and instead passed in a selector which the Twitter class will call when the asynchronous action is completed. A call to start this action looks like...
CMTwitterEngine *engine = [[CMTwitterEngine alloc] initWithDelegate:self];
[engine setSendUpdateFinished:#selector(sendUpdateFinished:)];
[engine setSendUpdateFailed:#selector(sendUpdateFailed:)];
[engine setParsingSendUpdateFailed:#selector(parsingSendUpdateFailed:)];
[engine setUsername:TWITTER_USERNAME pass:TWITTER_PASSWORD];
[engine sendUpdate:statusUpdateText.text];
This code first initializes the engine reference with self as the delegate. To attach the callbacks I send in selectors which I originally had on the sendUpdate method signature but the method calls got pretty long. I opted to simply set properties of the selectors. This all works but I am not sure I like how this is working since it only partially solves my problem.
To complete this example, I finish the asynchronous work and eventually call a method internally which looks for the given selector and calls it if it is defined.
- (void)sendUpdateFinished:(NSDictionary *)dictionary {
if (self.sendUpdateFinished != nil) {
[self.delegate performSelector:self.sendUpdateFinished withObject:dictionary];
}
}
I can pass in the status message to send as a Twitter update but I still do not have the context of the originating call. What if I want to call sendUpdate more than once and the first asynchronous call is still running? And what if the second call finishes first? They will both have self as the delegate so I would have to either track the context somehow or pass them to a different selector to distinguish them, which also does not satisfy my needs. What happens if I have 3 or 4 or 5 asynchronous calls? I need to know which ones were sent successfully and when they are complete.
It appears the only way that I can do all this is to create a class which holds onto all of the properties needed for the context, have that class act as the delegate for the call to the asynchronous Twitter method and then report back to the parent class which is likely UIViewController. I would take this approach but I have not read about this approach or seen any sample code yet which does this.
What would you do? How would you handle multiple asynchronous calls going out which could end in a different order than going out and then process them with context upon completion?
I think your situation is a great place to use NSNotificationCenter
http://developer.apple.com/iphone/library/documentation/Cocoa/Reference/Foundation/Classes/NSNotificationCenter_Class/Reference/Reference.html
I have to second (or third) the previously posted answers in that NSNotificationCenter is probably what you're looking for here.
Essentially one typically uses notifications when there are potentially many delegates all of which need to do something in response to a single action or event that has occurred. Think of it as a one-to-many sort of delegation or an implementation of the observer pattern. The basic things to know are:
NSNotifications have a name that you define which is just an NSString. Notifications can be posted by name and objects register to receive notifications by name.
When a notification is posted a notificationSender object and/or userInfo dictionary can be provided. The notificationSender is the direct way of determining who posted a given notification when it is being handled by the receiver. The userInfo is an NSDictionary that can be used to provide additional context info along with the notification.
So, rather than forcing all of the workers to adopt to an informal protocol and messing around with reflection style calling-methods-at runtime you just register instances of the workers with NSNotificationCenter. Typically the registration with the NSNotificationCenter is done in an init method of each worker class. Instances of each type of worker are then typically set up as "freeze dried" objects in a NIB or can be programatically instantiated in the app delegate so that they get registered with the notification center early on in the app's life.
When the thing occurs you post a NSNotification to the NSNotificationCenter (which is essentially a singleton) and then everything else that has registered to receive that particular type of notification will have the method called that was specified to handle that type of notification. When done these methods can then call a common method back on the sender (obtained via NSNotification's object method) to tell the sender that they've completed their work.
Once each known worker has checked in the the common method on the sender can then go on to whatever post-worker-completion code is to be performed.
One thing to consider is using Notifications instead. Simplifies code, couples things less tightly.
I'm new to Objective C, and Mac development... It appears that I can use the Posix threads API in my app.. Is this the recommended way? Or is their some Apple API I should be using for mutexes, condition variables and threads instead?
I should add that I'm developing for the iPhone.
I'd like to add exactly what I'm trying to do. Basically, CoreLocation is asynchronous... You tell it to start updating you, and then it just calls an update method on you periodically...
The problem I have is that I need another thread to block until an update occurs... How can I make the main application thread block until at least one CoreLocation update occurs? Is their an NSConditionVariable? :)
I'd suggest an even easier way to get stuck into threads is to use the following call:
[self performSelectorInBackground:(#selector(myMethod)) withObject:nil];
This will automatically create a new background thread for you to run in. Incidentally make sure you do the following inside your background method:
-(void) myMethod {
NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
// code you want to run in background thread;
[pool drain];
}
This is necessary because there isn't a default autorelease pool set up for any threads except the main one.
Finally, talking about blocking the main thread, you could use the following from your background thread to do this:
[self performSelectorOnMainThread:(#selector(myOtherMethod)) withObject:nil waitUntilDone:YES];
The optional third parameter will hold up the main thread for you if you want it to do so.
Hope that helps!
It depends on what you are trying to do, but I would start with NSOperation and NSOperationQueue. It makes it pretty simple to hand off background tasks. Take a look at Dave Dribin's blog post on NSOperation concurrency as well: http://www.dribin.org/dave/blog/archives/2009/05/05/concurrent_operations/
Instead of blocking the user interface by making it hang, I would suggest showing some kind of loading screen until you've received your first update. The could would look something like this:
- (void)viewDidLoad {
...
[myCLLocationManager beginUpdates];
[self showLoadingIndicator];
....
}
- (void)locationManager:(CLLocationManager *)manager didReceiveUpdates {
[self hideLoadingIndicator];
// Additionally load the rest of your UI here, if you haven't already
}
Don't quote me on those method calls verbatim, but that's how I suggest solving your problem, in essence.
Yes there is an NSCondition object, and it will probably do exactly what you want for the CoreLocation scenario you mentioned.