I'm implementing support for Lion's "Resume" feature in my OS X app.
I have a custom subclass of NSViewController in which I implemented the method
encodeRestorableStateWithCoder: as:
#implementation MyClass (Restoration)
-(void)encodeRestorableStateWithCoder:(NSCoder*)coder {
[coder encodeObject:_dataMember forKey:#"object_key"]; // I get the warning below when this line is executed for the second time
}
- (void)restoreStateWithCoder:(NSCoder *)coder {
_dataMember = [coder decodeObjectForKey:#"object_key"];
}
#end
However, since I have multiple instances of MyClass, different values are saved into the same key ("object_key") and I get the following warning from Cocoa:
NSKeyedArchiver warning: replacing existing value for key
'object_key'; probable duplication of encoding keys in class hierarchy
What is the best practice to overcome this problem?
Edit: I found here that each instance automatically has its own namespace to avoid collisions, so the problem might be in the way I'm manually calling encodeRestorableStateWithCoder to different instances with the same NSCoder object without telling it that these are different instances. However, I still can't figure out how to do that properly.
Thanks in advance!
To overcome this problem, it is possible to create a new NSMutableData where each of which is written by a separate (new) NSKeyArchiver, and store them all in an array which is stored in the original NSCoder object.
Here is an example for encoding the restorable state of subitems. The decoding part can be straight-forward given this code.
- (void)encodeRestorableStateWithCoder:(NSCoder *)coder
{
[super encodeRestorableStateWithCoder:coder];
// Encode subitems states:
NSArray* subitems = self.items;
NSMutableArray* states = [NSMutableArray arrayWithCapacity: subitems.count];
for (SubItemClass* item in subitems)
{
NSMutableData* state = [NSMutableData data];
NSKeyedArchiver *archiver = [[NSKeyedArchiver alloc] initForWritingWithMutableData:state];
[item encodeRestorableStateWithCoder:archiver];
[archiver finishEncoding];
[states addObject:state];
}
[coder encodeObject:states forKey:#"subitems"];
}
Related
I am able to support the Make New command of AppleScript for my app, however the returned 'specified object' (an NSUniqueIDSpecifier) for the core data managed object is useless. The following AppleScript returns the error message:
error "SpellAnalysis got an error: Invalid key form." number -10002 from level id "x-coredata:///Levels/tC5A49E01-1CE1-4ED6-8F6B-BC0AE90E279A2"
tell application "SpellAnalysis"
set thisLevel to make new «class Slev» with properties {«class Saln»:3}
get properties of thisLevel
end tell
So the newly created Levels object can not be acted upon in AppleScript. I've combed the Web for a solution to this and the closest thing I have found is Bill Cheeseman's example app, "WareroomDemo" which specifically deals with Cocoa Scriptability for Core Data apps (the Sketch example does not use Core Data). Unfortunately, it is a dated example, running only on pre-64-bit XCode and I can't actually run it--I can only look at the code. His app's Make Command may have the same limitations for all I know.
The returned 'objectSpecifier' is unable to refer to the created object either as a safe-guard against corrupting Core Data's organizing scheme, or perhaps because the returned object is an un-cashed 'fault'. I think the latter possibility is unlikely because I can force the fault to cash (by getting a property value on the managed object) , yet I get the same error message with the AppleScript.
Here is the method that creates my class:
- (id)newScriptingObjectOfClass:(Class)class forValueForKey:(NSString *)key withContentsValue:(id)contentsValue properties:(NSDictionary *)properties { // Creates a new Lesson object in response to the AppleScript 'make' command.
// Documentation for 'newScriptingObject…' states that to create a new class object when using Core Data, you intercede using the following method (or you can subclass the NSCreateCommand's 'performDefaultImplementation' method and put your NSManagedObject init code there):
if (class == [Levels class]) {
//NSLog(#"class: %#",class);
NSEntityDescription *levelsEntity = [NSEntityDescription
entityForName:#"Levels"
inManagedObjectContext:levelsDBase];
NSManagedObject *levelObject = [[NSManagedObject alloc] initWithEntity:levelsEntity insertIntoManagedObjectContext:levelsDBase];
SLOG(#"lessonObject: %#", lessonObject);
NSString *levelNumberString = [[properties objectForKey:#"levelNumber"] stringValue];
SLOG(#"levelNumberString: %#", levelNumberString);
[levelObject setValue:levelNumberString forKey:#"levelNumber"];
return levelObject; // When using Core Data, it seems that you return the newly created object directly
}
return [super newScriptingObjectOfClass:(Class)class forValueForKey:(NSString *)key withContentsValue:(id)contentsValue properties:(NSDictionary *)properties];
}
Here is my object specifier method:
- (NSScriptObjectSpecifier *)objectSpecifier {
// This NSScriptObjectSpecifiers informal protocol returns a unique ID specifier specifying the absolute string of the URI representation of this managed object. // AppleScript return value: 'level id <id>'.
// The primary container is the application.
NSScriptObjectSpecifier *containerRef = nil; // I understand that if the application is the container, this is value you use for the container reference
NSString *uniqueID = [[[self objectID] URIRepresentation] absoluteString];
return [[[NSUniqueIDSpecifier alloc] initWithContainerClassDescription:[NSScriptClassDescription classDescriptionForClass:[NSApp class]] containerSpecifier:containerRef key:#"levelsArray" uniqueID:uniqueID] autorelease];
}
The problem lies with the specifier method. The Sketch example actually uses the technique that I needed. I overlooked it many times because I didn't see how it would apply to Core Data managed objects. Instead of returning the objects uniqueID, you make it return the managedObject index using the 'indexOfObjectIdenticalTo:' method as follows:
- (NSScriptObjectSpecifier *)objectSpecifier {
NSArray *levelsArray = [[NSApp delegate] levelsArray]; // Access your exposed to-many relationship--a mutable array
unsigned index = [levelsArray indexOfObjectIdenticalTo:self]; // Determin the current objects index
if (index != (unsigned)NSNotFound) {
// The primary container is the document containing this object's managed object context.
NSScriptObjectSpecifier *containerRef = nil; // the appliation
return [[[NSIndexSpecifier allocWithZone:[self zone]] initWithContainerClassDescription:[NSScriptClassDescription classDescriptionForClass:[NSApp class]] containerSpecifier:containerRef key:#"levelsArray" index:index] autorelease];
} else {
return nil;
}
}
Note that this method resides within a subclass of your Core Data managedObject--in this case, the 'Levels' class. The 'self' within the 'indexOfObjectIndenticalToSelf:' method refers to the current managedObject ('Levels') being handled. Also, be sure to provide the specifier (accessor) type to your 'sdef' file, like this:
<element type="level">
<cocoa key="levelsArray"/>
<accessor style="index"/>
</element>
I'm curious to know what the best way is to create a new NSManagedObject in RestKit 0.20? Currently my code looks something like this:
#pragma mark - navigation buttons
- (void)createButtonDidTouch
{
// create new album object
NSManagedObjectContext *context = [[NSManagedObjectContext alloc] initWithConcurrencyType:NSMainQueueConcurrencyType];
NSManagedObjectContext *parentContext = RKObjectManager.sharedManager.managedObjectStore.mainQueueManagedObjectContext;
context.parentContext = parentContext;
NSEntityDescription *entityDescription = [NSEntityDescription entityForName:#"Album" inManagedObjectContext:parentContext];
Album *newAlbum = [[Album alloc] initWithEntity:entityDescription insertIntoManagedObjectContext:context];
// pass object to create view to manipulate
AlbumCreateViewController *createViewController = [[AlbumCreateViewController alloc] initWithData:newAlbum];
createViewController.delegate = self;
createViewController.managedObjectContext = context;
UINavigationController *navController = [[UINavigationController alloc] initWithRootViewController:createViewController];
navController.modalTransitionStyle = UIModalTransitionStyleCoverVertical;
[self presentViewController:navController animated:YES completion:nil];
}
#pragma mark - create view controller delegate
- (void)createViewControllerDidSave:(NSManagedObject *)data
{
// dismiss the create view controller and POST
// FIXME: add restkit code to save the object
NSLog(#"save the object...");
NSDictionary *userInfo = [KeychainUtility load:#"userInfo"];
NSString *path = [NSString stringWithFormat:#"/albums/add/%#/%#", userInfo[#"userID"], userInfo[#"apiKey"]];
[RKObjectManager.sharedManager postObject:data path:path parameters:nil success:^(RKObjectRequestOperation *operation, RKMappingResult *mappingResult) {
operation.targetObject = data;
} failure:^(RKObjectRequestOperation *operation, NSError *error) {
NSLog(#"create album error: %#", error);
}];
[self dismissViewControllerAnimated:YES completion:nil];
}
- (void)createViewControllerDidCancel:(NSManagedObject *)data
{
// dismiss the create view controller
NSLog(#"delete the object...");
// FIXME: add restkit code to delete the object
[self dismissViewControllerAnimated:YES completion:nil];
}
I'm also curious to know what my responsibilities are for saving / deleting this object. If I POST to the server via RestKit is the managed object context saved?
What if I decide to cancel this creation process — what's the preferred way to delete this object?
Basically how much is RestKit doing for me, and what should I make sure I'm doing. I haven't found much documentation on this and would like to be clear on it.
When you initialize an RKManagedObjectRequestOperation for a given object, RestKit will obtain a permanent object ID for that object and then create a child managed object context whose parent context is the context the object is inserted into. The operation then executes the HTTP request to completion and obtains a response.
If the response is successful and the mapping of the response is successful (note that the mapping occurs within this private child context), then the private child context is saved. The type of save invoked is determined by the value of the savesToPersistentStore property (see http://restkit.org/api/0.20.0/Classes/RKManagedObjectRequestOperation.html#//api/name/savesToPersistentStore).
When YES, the context is saved recursively all the way back to the persistent store via the NSManagedObjectContext category method saveToPersistentStore (see http://restkit.org/api/0.20.0/Categories/NSManagedObjectContext+RKAdditions.html).
When NO, the context is saved via a vanilla [NSManagedObjectContext save:] message, which 'pushes' the changes back to the parent context. They will remain local to that context until you save them back. Keep in mind that managed object context parent/child hierarchies can be as long as you create within the application.
If the HTTP request failed or there was an error during the mapping process, the private context is not saved and the operation is considered failed. This means that no changes are saved back to the original MOC, leaving your object graph just as it was before the operation was started (except the object being sent, if temporary, now has a permanent object ID but is still unsaved).
The way you do it should works (calling each time the MOC in each of your VC), but is not "recommended".
What Apple suggests, just like any Core Data app, is the "pass the baton" style.
Nested contexts make it more important than ever that you adopt the
“pass the baton” approach of accessing a context (by passing a context
from one view controller to the next) rather than retrieving it
directly from the application delegate.
See here:
http://developer.apple.com/library/ios/#releasenotes/DataManagement/RN-CoreData/_index.html
As for your second question, RestKit should manage saving/updating your Core Data stack upon success of your api calls if everything is well mapped/setup.
From blake the RK creator:
if you POST or PUT a Core Data object, RK obtains a permanent object
ID for it and then creates a secondary managed object context, fires
the request, and maps the response (if possible). if the response and
the mapping are successful, it will either save it back to the parent
context or all the way back to the persistent store (i.e. into SQLite)
based on the value of the savesToPersistentStore.
I have this application with 3 classes: AppController, Profile, ProfileBuilder. I also need 3 windows: one for each class. I tried keeping all 3 as subclasses of NSObject and applying initWithNibName to an NSWindowController class WindowController variable but when I tried outputting some values on each window it wouldn't work, and also the window resulted as null using NSLog. I was wondering what was the best way to manage multiple windows, perhaps all from a same class like an AppWindowsController involving as least as possible specific code in the other classes, and keeping, if possible, the other classes as subclasses of NSObject and not NSWindowController. So if there is, maybe a way to control the behavior of the windows remotely, adding as least as possible code inside the specific classes, just to keep them as clear as possible and uniquely focused on their content. Thanks, hope I made myself clear, I'm actually pretty new to the Cocoa framework.
You should be able to load the nib files with your windows in an init method for your different classes. For example, in Profile, you could do something like this:
-(id)init {
if (self = [super init]) {
NSArray *array;
BOOL success = [[NSBundle mainBundle] loadNibNamed:#"ProfileWindow" owner: self topLevelObjects:&array];
if (success) {
for (id obj in array) {
if ([obj isKindOfClass:[NSWindow class]]) {
self.profileWindow = obj;
}
}
[self.profileWindow makeKeyAndOrderFront:self];
}
}
return self;
}
profileWindow is a property (typed as strong). In the xib file, I set the File's Owner to Profile.
I just like to improve the solution of rdelmar.
You don't need to iterate over the array to find the NSWindow class.
If you define profileWindow as an outlet and connect it in the IB, the call
[[NSBundle mainBundle] loadNibNamed:#"ProfileWindow" owner:self topLevelObjects:&array];
will assign the window object to your outlet, the array stuff is not required.
The key here is the owner object which act as interface. In the IB you can define the class type of the owner and if so, see its outlets.
I have these methods in a class at the moment which seem to work fine. I would like to create a subclass which inherits these methods. The problem I have is that in the third method (shiftViewUpForKeyboard) I want the if statement to be a different UITextField (mirror being the current example).
I've read that to override a method in the subclass, you have to basically copy it exactly with the new coding, but if I want to just change that small section what is the best way to do it?
Thank you in advance.
- (void) viewWillAppear:(BOOL)animated {
[[NSNotificationCenter defaultCenter] addObserver: self
selector: #selector(shiftViewUpForKeyboard:)
name: UIKeyboardWillShowNotification
object: nil];
[[NSNotificationCenter defaultCenter] addObserver: self
selector: #selector(shiftViewDownAfterKeyboard)
name: UIKeyboardWillHideNotification
object: nil];
}
- (void) viewDidDisappear:(BOOL)animated {
[[NSNotificationCenter defaultCenter] removeObserver: self
name: UIKeyboardWillShowNotification
object: nil];
[[NSNotificationCenter defaultCenter] removeObserver: self
name: UIKeyboardWillHideNotification
object: nil];
}
- (void) shiftViewUpForKeyboard: (NSNotification*) theNotification;
{
if(mirror.isEditing == YES)
{
CGRect keyboardFrame;
NSDictionary* userInfo = theNotification.userInfo;
keyboardSlideDuration = [[userInfo objectForKey: UIKeyboardAnimationDurationUserInfoKey] floatValue];
keyboardFrame = [[userInfo objectForKey: UIKeyboardFrameBeginUserInfoKey] CGRectValue];
UIInterfaceOrientation theStatusBarOrientation = [[UIApplication sharedApplication] statusBarOrientation];
if UIInterfaceOrientationIsLandscape(theStatusBarOrientation)
keyboardShiftAmount = keyboardFrame.size.width;
else
keyboardShiftAmount = keyboardFrame.size.height;
[UIView beginAnimations: #"ShiftUp" context: nil];
[UIView setAnimationDuration: keyboardSlideDuration];
self.view.center = CGPointMake( self.view.center.x, self.view.center.y - keyboardShiftAmount);
[UIView commitAnimations];
viewShiftedForKeyboard = TRUE;
}
}
In any case if you don't want to copy the full method in the subclass, and adding your little customization, the only other possible approach I see is to change the original class. To do it I can suggest two possibilities:
1)
You could create in the original class a method called:
-(UITextField *)keyboardShiftTextField
and then in this class replace the mirror.isEditing code with:
[[self keyboardShiftTextField] isEditing]
In such case the only difference between the two classes will be in the implementation of the new method, that for the original class will be:-(UITextField *)keyboardShiftTextField {
return mirror;
}
while in the subclass this return the right text field.
2)
A second approach is more elegant as it requires the definition of the delegate pattern. This requires some overhead in term of code but we'll provide you more flexibility. Besides if the only reason to make the subclass is just to override this third method, then using the delegate pattern you can avoid creating the subclass at all, as the "custom" work will be done by the delegate. If the number of methods to override is more than one, you can still use this mechanism by moving into the protocol section all the parts that need customization. This is a quite common technique for Obj-C and Cocoa, which limits the need for some classes in many cases. Typically you use a subclass when you want to provide a different functionality, but in your case you're not providing a different functionality, but just a customization for the same functionality (the view shift up).
The usual approach would be the http://en.wikipedia.org/wiki/Template_method_pattern: pull out just the bit of the method that varies between classes, make that a separate method, and override that in subclasses.
[EDITED to add ...] An alternative approach -- I can't tell whether it would work well here without seeing more of your code -- would be to make the thing that varies a parameter that's passed into the method (or select it on the basis of a parameter passed into the method, or something else of the kind). (You'd typically then use other mechanisms rather than inheritance+polymorphism to get the effect you want, of multiple things with similar behaviour: they'd be instances of the same class but fed with different data.)
I would like to use an NSArrayController to provide data to an NSTableView. The problem I am facing is that I do not want to pre-load all my data into an array and then use the array controllers setContent: method. My data model is a large existing code base that manages millions of records. It contains methods to efficiently return a set of data rows.
Following an example I found on limiting the number of objects in an NSArrayController, I tried subclassing NSArrayController and overriding the arrangedObjects: method to return an array proxy class I wrote. The array proxy class provided count: and objectAtIndex: methods. The object returned by objectAtIndex: is an NSDictionary. When I tried returning my array proxy from the arrangedObjects: method both count: and objectAtIndex: get called, but I also get an unrecognized selector error on my array proxy class for _valueForKeyPath:ofObjectAtIndex:. This looked like a private method, so I did not continue down this path.
I also thought of returning a smaller array of data from arrangedObjects:, but could not figure out how I would determine which rows the NSTableView was trying to display.
Is a datasource the "correct" way to interface with my existing data model or is there some way to make an NSArrayController work?
NSArrayController already works, with proxies and indexes and lazy-loading and the whole shabang. Have you tried just using it as-is? If afterwards you feel the need to micro-manage the data-loading, use NSFetchRequest. Subclass NSArrayController and add an initializer along these lines:
+ (id)arrayControllerWithEntityName: (NSString *)entityName error:(NSError **)error
{
id newInstance = [[[self alloc] initWithContent:nil] autorelease];
[newInstance setManagedObjectContext:[[NSApp delegate] managedObjectContext]];
[newInstance setEntityName:entityName];
[newInstance setAutomaticallyPreparesContent:YES];
[newInstance setSelectsInsertedObjects:YES];
[newInstance setAvoidsEmptySelection:YES];
[newInstance setAlwaysUsesMultipleValuesMarker:YES];
NSFetchRequest *dontGobbleRequest = [[[NSFetchRequest alloc] init] autorelease];
//configure the request with fetchLimit and fetchOffset an' all that
NSError *err;
if ([newInstance fetchWithRequest:dontGobbleRequest merge:YES error:&err] == NO) {
//better check docs whether merge:YES is what you want
if(*error && err) {
*error = err;
}
return nil;
}
return newInstance;
}
You'll have to do some research into the various possibilities and configurations, but you get the picture.