I can create an NSPredicate easily using an NSPredicateEditor (a subclass of NSRuleEditor). What I'd like to know is this:
How can I take an existing NSPredicate (one created by the editor) and reload it into the editor so that I can alter it?
EDIT: I tried #John's suggestion of using setObjectValue:, but that didn't quite work. Let me explain my set up a bit more:
I've got a Document-based cocoa app, and the Document window just has an NSPredicateEditor on it. In the dataOfType:error: method, I have:
NSPredicate * pred = [predicateEditor objectValue];
NSData * predicateData = [NSKeyedArchiver archivedDataWithRootObject:pred];
return predicateData;
In the readFromData:ofType:error: method, I have:
NSPredicate * pred = [NSKeyedUnarchiver unarchiveObjectWithData:data];
[predicateEditor setObjectValue:pred];
return (pred != nil);
I've verified that the predicate is getting correctly archived and unarchived, but after opening a saved predicate, the predicate is not loaded into the predicateEditor. (Yes, predicateEditor is hooked up as an IBOutlet)
Set the objectValue property of the NSPredicateEditor to the predicate in question.
The documentation has this description of the loading process; does any of this seem like it might cause a problem with your setup?
First, an instance of
NSPredicateEditor is created, and some
row templates are set on it—either
through a nib file or
programmatically. The first thing
predicate editor does is ask each of
the templates for their views, using
templateViews.
After setting up the predicate editor,
you typically send it a
setObjectValue: message to restore a
saved predicate. NSPredicateEditor
needs to determine which of its
templates should display each
predicate in the predicate tree. It
does this by sending each of its row
templates a matchForPredicate: message
and choosing the one that returns the
highest value.
After finding the best match for a
predicate, NSPredicateEditor copies
that template to get fresh views,
inserts them into the proper row, and
then sets the predicate on the
template using setPredicate:. Within
that method, the
NSPredicateEditorRowTemplate object
must set its views' values to
represent that predicate.
NSPredicateEditorRowTemplate next asks
the template for the “displayable
sub-predicates” of the predicate by
sending a
displayableSubpredicatesOfPredicate:
message. If a template represents a
predicate in its entirety, or if the
predicate has no subpredicates, it can
return nil for this. Otherwise, it
should return a list of predicates to
be made into sub-rows of that
template's row. The whole process
repeats for each sub-predicate.
Related
I'm implementing a search screen in my app using UITableViewDiffableDataSource. Each cell represents a search hit and highlights the search match in the cell title, kind of like Xcode's Open Quickly window highlights portions of its result items. As text is typed into the search field, I update the results list. Results move up and down in the list as their relevance changes.
The trick is that I need to force every cell to re-render every time the search text changes, because a new search string means an update to the highlighted portions of the cell title. But I don't want to animate a deletion and insert, because it's still the same item. How can I tell the data source using the snapshot that it needs to reload cells?
I declare the data source like this:
#property (retain) UITableViewDiffableDataSource<NSString *, SearchHit *> *dataSource;
SearchHit represents one search result; it has properties for a display title and an array of ranges to highlight in the title. And it overrides hash and isEqual: so that every result row is uniquely identified.
My code looks something like this:
-(void)searchBar:(UISearchBar *)searchBar textDidChange:(NSString *)searchText
{
NSArray<SearchHit *> *hits = [self fetchHits:searchText];
NSDiffableDataSourceSnapshot<NSString *, SearchHit *> *snap = [[[NSDiffableDataSourceSnapshot alloc] init] autorelease];
[snap appendSectionsWithIdentifiers:#[#""]];
[snap appendItemsWithIdentifiers:hits];
[snap reloadItemsWithIdentifiers:hits];
[self.dataSource applySnapshot:snap animatingDifferences:YES];
}
At first I didn't have the reloadItemsWithIdentifiers call there, and then no cell would change at all once it was in the result list. Adding the reload call helped, but now most of the cells are constantly one update behind. This smells like a logic error somewhere in my code, but I've verified that the hits passed to the snapshot are correct and the hits passed to the data source's cell creation callback are not.
This article by Donny Wals and this related Twitter thread involving Steve Breen suggests that the way to fix this is to make the item identifier type only represent the properties needed to display the cell. So I updated SearchHit's hash and equality comparison to include the highlighted portions of the title, which they didn't before. Then I got delete and insert animations for all the cells on every update, which I don't want.
This seems like what reloadItemsWithIdentifiers should do...right?
Sample project here on GitHub.
The diffable datasource API may not be the right tool to effect animations on cells themselves. It’s geared towards the animation of the appearance, disappearance and ordering of cells. If your data source has a change that is expressed via Hashable conformance the api will see it as a change and delete/insert etc.
My advice would be to remove the search text from the item identifier and have each cell observe the search text and effect an animation or redraw independently from the datasource.
The proper solution to this is actually in the names of the APIs - the objects you give to the data source should be identifiers, like rowid values from a database. In my case, when the item identifiers don't represent rows in a database that I can look up, I just need to keep the state of the objects in some sort of lookup structure, so that when I call reloadItemsWithIdentifiers, I get the state for each cell from that structure, not from the object that the data source hands to me.
I'm a nwebie in Core Data, i have designed a navigation based application and some of the data i use are created on run time(come from a URL via JSON). I took a few tutorials an searched for almost a day but haven't still realized how to save the incoming JSON data to the Entity (or event?) in my Core Data model. I fetch the data in the DetailViewController class and i need to save this data to Core Data(I have prepared an Entity with 7 properties). Can anyone please help?(If you know a good tutorial or sample code i will be pleased)
EDIT This may be a little specific but i really have trouble with and need just a little help.
My data comes to the app from a kind of restful server(i wrote it in PHP), firstly user enters his/her login informations(which i have saved to the database on server before) and when the response data comes i will use different elements of it in differen views(for example the user_id will be used on a view and the buttonData etc on other views). My question is, how will i save JSON data into my core data model(has tree Entities for the moment). Thanks in advance
Note: I lokked arround a lot but couldn't find any answer&tutorial about an app like mine
The best way to do that would be to create entities corresponding to JSON structure. Easiest was is when each JSON object becomes an entity, and arrays become arrays of entities. Be reasonable, however, and don't introduce too much overkill for JSON subobjects that are essentially part of its superobject.
When you have created entities, you can start off with the parsing and translation. Use some JSON framework (starting from iOS5 there's one from Apple) and parse JSON string into object tree, where root item is either an NSArray or NSDictionary, and subelements will be NSArray, NSDictionary, NSNumber, NSString or NSNull.
Go over them one by one in iterational loops and assign according values to your core data entity attributes. You can make use of NSKeyValueCoding here and avoid too much manual mapping of the attribute names. If your JSON attributes are of the same name as entity attributes, you'll be able to just go over all dictionary elements and parse them into attributes of the same name.
Example
My parsing code in the similar situation was as follows:
NSDictionary *parsedFeed = /* your way to get a dictionary */;
for (NSString *key in parsedFeed) {
id value = [parsedFeed objectForKey:key];
// Don't assign NSNull, it will break assignments to NSString, etc.
if (value && [value isKindOfClass:[NSNull class]])
value = nil;
#try {
[yourCreatedEntity setValue:value forKey:property];
} #catch (NSException *exception) {
// Exception means such attribute is not defined in the class or some other error.
}
}
This code will work in trivial situation, however, it may need to be expanded, depending on your needs:
With some kinds of custom mappings in case you want your JSON value be placed in differently named attribute.
If your JSON has sub-objects or arrays of sub-objects, you will need to detect those cases, for example in setters, and initiate new parsing one level deeper. Otherwise with my example you will face the situation that assigns NSDictionary object to an NSManagedObject.
I don't think it is reasonable to dive into these, more advanced matters in scope of this answer, as it will expand it too much.
I suggest you to use this library : https://github.com/TouchCode/TouchJSON
And then if you want to make a factory to parse json and feed your code data, you can use selectors to call methods to fill all your attributes.
Chances are your JSON data gets converted to an NSDictionary or NSArray (or some combination of the two). Simply extract the key/values from the JSON structure and add them to your entity class.
This lib helps me lot
Features
Attribute and relationship mapping to JSON key paths.
Value transformation using named NSValueTransformer objects.
Object graph preservation.
Support for entity inheritance
Works vice-versa
Newbie question here. I'd like to be able to specify through data (i.e. an XML file), the appropriate Objective-C message to send. Any advice on if this is possible or how I can do this?
The next best thing, if I can't do this, would be some way to create a map object that would correlate a key (an int) with a function (I guess also a selector). Is that possible if the above isn't?
If someone could point me to some tutorial or example code as reference, that'd be great. Right now I'm doing things with a big switch statement, and I don't like it. (I'm switching on the id and in each case, explicitly calling the method relevant to the particular id.)
I love that you asked this question; too often, I see Satan's Swollen Switch Statement. It's nice to see someone wanting to using a function-table instead.
If you're OK with using a property list file (which is usually encoded in XML), this is really easy.
Just make a property list where the root element is a dictionary, which maps from some keys to some selectors.
Key Type Value
----------------------------------------------
Root Dictionary
firstKey String someSelector
secondKey String anotherSelector
Load the contents of your property list into an NSDictionary:
id path = [[NSBundle mainBundle] pathForResource:#"filename" ofType:#"plist"];
id dict = [NSDictionary dictionaryWithContentsOfFile:path];
SEL selector = NSSelectorFromString([dict objectForKey:#"firstKey"]);
if ([someObject respondsToSelector:selector]) {
[someObject performSelector:selector];
}
Of course, you'll want to refactor this logic into an appropriate method, and probably cache the property list as an instance variable.
Note: I personally think it's better to just put this function table inline; property lists are cool, but I'm not sure that it is very helpful in this case. Also, if you are cool with using Objective-C++, std::map will allow you to get away with not wrapping and unwrapping the selectors in NSString objects, etc.
I am playing with an app that uses Core Data and NSManagedObjects to populate a UITableView. There is only one class in my application, called Event. I have created the following custom instance method on Event:
- (BOOL)isExpired {
return ([[self.endOn dateAtEndOfDay] timeIntervalSinceNow] < 0);
}
I would like to limit the UITableView that displays Event objects to only the Events that are expired - that is, where isExpired returns YES. I have tried to do this by adding an NSPredicate to the NSFetchRequest:
NSPredicate *predicate = [NSPredicate predicateWithBlock:^BOOL(id evaluatedObject, NSDictionary * bindings) {return([evaluatedObject isExpired]);}];
[fetchRequest setPredicate:predicate];
but I get the error: *** Terminating app due to uncaught exception 'NSInvalidArgumentException', reason: 'Problem with subpredicate BLOCKPREDICATE(0x272ac)'
***. Does this mean that you can't use a block predicate with an NSFetchRequest? Or have I just constructed it improperly?
Thank you!
So, it appears that we've established in the comments to the original post that this is likely caused by SQLite stores being incompatible with block predicates, since Core Data cannot translate these to SQL to run them in the store (thanks, JoostK).
There might be a couple of ways to overcome this:
Provided that the end date of your entities is a regular attribute, you might be able to express the expiry constraint as a predicate format string instead of a block predicate, which Core Data should be able to translate into a SQL clause.
If the above is possible, you will probably prefer to use a fetch request template to retrieve the expired items. You would need to pass in a substitution variable like $NOW to give access to the current date, though. This has the advantage of making the predicate template show up in the model editor.
Both approaches, however, have the disadvantage of duplicating existing functionality (i.e., your isExpired method). So another way would be fetch all qualifiying entities regardless of their expiry state first, and then run a dedicated filtering step on the resulting set of entities to weed out the non-expired ones. Since by that point, they have been fully resurrected from the store, you should be able to use a block predicate for this.
You can do a normal fetch request without specifying the predicate, and afterwards filter the resulting array:
NSArray *allEvents = [context executeFetchRequest:fetchRequest];
if (!allEvents) { // do error handling here
}
NSArray *expiredEvents = [allEvents filteredArrayUsingPredicate:predicate];
I want to load a client object and then pull their related purchase orders based on whether they have been placed or not, purchase orders have an IsPlaced BOOL property.
So I have my client object and I can get all purchase orders like this, which is working great:
purchaseordersList =[[myclient.purchaseorders allObjects] mutableCopy];
But ideally I would actually like 2 array's - one for each order type: IsPlaced=YES and IsPlaced=NO
How do I do that here? Or do I need to do another fetch?
First, there is no reason to be turning the set into an array unless you are sorting it and there is no reason to be turning that array into a mutable array. Did you get that from some example code?
Second, you can filter an array or a set by using a predicate so you can create two sets (or arrays) easily via:
NSSet *placed = [[myclient purchaseorders] filteredSetUsingPredicate:[NSPredicate predicateWithFormat:#"isPlaced == YES"]];
NSSet *notPlaced = [[myclient purchaseorders] filteredSetUsingPredicate:[NSPredicate predicateWithFormat:#"isPlaced == YES"]];
If you are wanting to use this for a UITableView then look into a NSFetchedResultsController instead. It will save you a LOT of boiler-plate code.
Do you remember what example code you got that from? Been seeing that -mutableCopy a lot lately and would love to quash it. :)