Trying to create a object for ABPersonRef
example:
ABpersonRef ref;
have included Addressbook and AddressBookUI framework
even then when i compile it states 'ABPersonRef' Undeclared Identifier
Have you added imports to your source file?
#import <AddressBook/AddressBook.h>
#import <AddressBookUI/AddressBookUI.h>
Edit: I've read your question more carefully. There's no ABPersonRef type (at least public) in iPhone SDK. You should work with ABRecordRef type which is generic for both person and group records - as you can see all AB*** methods work with ABRecordRef type.
You will also need to link against AddressBook framework and
#import <AddressBook/AddressBook.h>
Related
I am creating an iPhone custom framework which should be able to be integrated into any iPhone app.
I have created some properties in one of the public header files of my framework. I want to give #package level access to those properties, so that those properties can be accessed only with in the classes inside the framework.
I do not want the user to use those properties.
Please tell me whether doing this is possible. If yes, please give me some idea on how to achieve this!.
First you should know that there is no way to completely forbid a user of your library to call a method. Even if you don't declare it in your header, a user could declare it on its own and use it. He would still have to find out the name of your property though, for instance by running classdump on your library.
Therefore in Objective-C, we make properties private by not declaring them in the header (which is the public part of your class), but by declaring them in the implementation (which is the "private" part of your class).
If you create another header which contains a category on your class, you can add some properties to it that will not be in the main header - so not on the public declaration of your class - but that can still be imported by other classes of your library that know about this header.
For instance:
MyClass+SecretProperties.h:
#interface MyClass ()
#property (strong) NSString *secretString;
#end
MyClass.m:
#import "MyClass.h"
#import "MyClass+SecretProperties.h"
#implementation MyClass
#synthesize secretString; // Depending on your runtime, you may not even need to this - properties are auto-synthesized in the latest SDKs.
…
#end
OtherClass.m:
#import "MyClass.h"
#import "MyClass+SecretProperties.h"
// Now you can use secretString on instances of MyClass
Then since you only export MyClass.h with your library, users have no idea that there is a secretString property :) This is the closest you can get to a #package scope AFAIK.
If you want to make those property as private then use below things.
In your .m file use extension characteristics of objective c.
#interface InitialViewController ()
//declare your property here
#end
I am reading about iOS programming and I bought the Programming iOS 4 book. There is a introductory part where among several things "Files" is mentioned.
I don't understand how the source files is put together. You have a header file with function declarations, then you have a corresponding file with the function definitions.
Let say you have a Car.h and Car.m & Person.h and Person.m.
Now, if you want to use the Car in the Person class you would import only the Car.h file. How is this sufficient? I don't understand the sequence it put together and builds a program. (Not thinking about the technical stuff, just h/m files.)
The .h or "header file" contains the interface.
The .m or "implementation file" contains the implementation.
Each implementation file is also called a "compilation unit" because the compiler compiles each one separately. Within each compilation unit, the compiler needs to know about types and methods. All it needs to know about a class to create the right code is information about the methods it implements.
So let's imagine you have these files:
Car.h
#import <Foundation/Foundation.h>
#interface Car : NSObject
- (void)drive;
#end
Car.m
#import "Car.h"
#implementation Car
- (void)drive {
NSLog(#"I'm driving!");
}
#end
Person.h
#import <Foundation/Foundation.h>
#class Car;
#interface Person : NSObject
#property (nonatomic, strong) Car *car;
- (void)start;
#end
Person.m
#import "Person.h"
#import "Car.h"
#implementation Person
#synthesize car;
- (void)start {
[car drive];
}
#end
Now when the compiler does its business, it compiles both Car.m and Person.m into Car.o and Person.o respectively. [These then get linked into the final binary, but that's beyond the scope of this question for now].
When it compiles Person.m, the compiler doesn't need to know how - (void)drive of Car is implemented, but it does need to know that it exists, that it is a method that takes no arguments and returns nothing. It doesn't care about the implementation, just that it exists. So you just need to #import the header file of Car to tell the compiler about the methods that exist on Car. The compiler knows that the implementation exists, because you've told it so, and then later on the linker will do it's business to correctly wire up the method call to the correct implementation. How the linker actually does that is a huge topic and I encourage you to go and read about it separately if you don't already understand it.
Note that it's the same for all of the standard NS classes that you use such as NSObject, NSString, etc. You just need to #import Foundation.h from the Foundation framework which tells the compiler about what these classes are and what methods are defined on them.
Creating an executable from a set of source code files is a two stage process.
Firstly, all of the .m files are individually compiled with the Objective-C compiler. This turns each one into a .o file which is an object code file. However, if the code in a .m file refers to things that are defined in other .m files, the compiler does not know about these so it just leaves unresolved references in the .o file.
The second stage is called linking. This takes all the .o files and combines them into an executable. When the linker finds unresolved references in one .o file, it checks all the others to resolve the reference.
Header files allow the compiler to have some information from outside the particular .m file it is currently compiling. So if you have two classes Foo and Bar they are conventionally defined in files Foo.m and Bar.m In order for the compiler to know what class Bar looks like when it is compiling Foo.m we put class Bars interface declaration in a header file (conventionally Bar.h) and import it into your .m file. If you see the line
#import "Bar.h"
it is literally as if the compiler has copy-pasted the entire header file into the source code file before compiling it.
What language have you been using until now? Many languages do it this way including c and c++. The m files are compiled into an actual program, and the h files provide a list of ways to interact with it. While you can still call the methods if you interact with the objective c runtime, the compiler will not guarantee their existence unless they are in the h file.
Now, I say guarantee, but if you dont provide an implementation in the m file, the sibling to the compiler, the linker will have a fit. It will try to make a jump into another m file based on its h file only to tragically discover that it is not there.
The benefits of splitting like this is that you can compile your source into a library and distribute it along with the h files and another application can use it without having the implementation source code.
In summary the m files compile into a lost island of bits and the h files are the map to get around it. If something is on the map that doesnt exist then you will get lost. If something exists but is not on the map then you will have a lot of trouble finding it.
Header files specify what messages ("methods" in other languages) can be passed to a class. That's all the compiler needs to know to compile your code; the linker will eventually wire everything up, so to speak, using the *.m files.
The compiler will handle that for you.
As you stated, a header file contains just the declarations.
Its like a interface to the actual code and that is the only thing Compiler needs to know to fetch the rest.
I am working on iOS5.I am new for SQLite. I have created database named DemoDB.sqlite
Now for making connection with database I have added libsql3.0.dylib -library for SQLite.
But I received an error. "Error in readDataFromDatabase method"
Which library should I add?
-(void) readDataFromDatabase
{
sqlite3 *database;
}
The error I received is "use of undeclared identifier database"
The compiler cannot see the declaration of the sqlite 3 struct. Try importing or including the main SQLite3 API header file:
#include <sqlite3.h>
or
#import <sqlite3.h>
Yes, The Compiler cannot see the declaration of the sqlite3 library ,So you have to import it by
#import "/usr/include/sqlite3.h"
or
#import <sqlite3.h>
it will work after that .
In my iOS5 app, I have NSObject States class, and trying to init it:
states = [states init];
here is init method in States:
- (id) init
{
if ((self = [super init]))
{
pickedGlasses = 0;
}
return self;
}
But there is error in the line states = [states init];
receiver type "States" for instance message is a forward declaration
What does it mean? What am I doing wrong?
That basically means that you need to import the .h file containing the declaration of States.
However, there is a lot of other stuff wrong with your code.
You're -init'ing an object without +alloc'ing it. That won't work
You're declaring an object as a non-pointer type, that won't work either
You're not calling [super init] in -init.
You've declared the class using #class in the header, but never imported the class.
FWIW, I got this error when I was implementing core data in to an existing project. It turned out I forgot to link CoreData.h to my project. I had already added the CoreData framework to my project but solved the issue by linking to the framework in my pre-compiled header just like Apple's templates do:
#import <Availability.h>
#ifndef __IPHONE_5_0
#warning "This project uses features only available in iOS SDK 5.0 and later."
#endif
#ifdef __OBJC__
#import <UIKit/UIKit.h>
#import <Foundation/Foundation.h>
#import <CoreData/CoreData.h>
#endif
I got this sort of message when I had two files that depended on each other. The tricky thing here is that you'll get a circular reference if you just try to import each other (class A imports class B, class B imports class A) from their header files. So what you would do is instead place a forward (#class A) declaration in one of the classes' (class B's) header file. However, when attempting to use an ivar of class A within the implementation of class B, this very error comes up, merely adding an #import "A.h" in the .m file of class B fixed the problem for me.
I was trying to use #class "Myclass.h".
When I changed it to #import "Myclass.h", it worked fine.
If you are getting this error while trying to use Swift class or method in Objective C: you forgot one of 2 steps Apple defined on this diagram:
Example:
Error shows up in your Test.m file:
Receiver 'MyClass' for class message is a forward declaration
In Obj-C files:
Step 1: check that Test.h has
#class MyClass;
Step 2: find *-Swift.h file name in Build Settings (look for Objective-C Generated Interface Header Name). Name will be something like MyModule-Swift.h
Step 3: check that Test.m imports the above header
#import <MyModule/MyModule-Swift.h>
In Swift file:
Ensure MyClass (or it's base class) inherits NSObject class.
Ensure #objc is before each method you want call from Obj-C.
Also, check Target Membership section (in File Inspector).
You are using
States states;
where as you should use
States *states;
Your init method should be like this
-(id)init {
if( (self = [super init]) ) {
pickedGlasses = 0;
}
return self;
}
Now finally when you are going to create an object for States class you should do it like this.
State *states = [[States alloc] init];
I am not saying this is the best way of doing this. But it may help you understand the very basic use of initializing objects.
Check if you imported the header files of classes that are throwing this error.
Make sure the prototype for your unit method is in the .h file.
Because you're calling the method higher in the file than you're defining it, you get this message. Alternatively, you could rearrange your methods, so that callers are lower in the file than the methods they call.
There are two related error messages that may tell you something is wrong with declarations and/or imports.
The first is the one you are referring to, which can be generated by NOT putting an #import in your .m (or .pch file) while declaring an #class in your .h.
The second you might see, if you had a method in your States class like:
- (void)logout:(NSTimer *)timer
after adding the #import is this:
No visible #interface for "States" declares the selector 'logout:'
If you see this, you need to check and see if you declared your "logout" method (in this instance) in the .h file of the class you're importing or forwarding.
So in your case, you would need a:
- (void)logout:(NSTimer *)timer;
in your States class's .h to make one or both of these related errors disappear.
I'm getting a warning in XCode which says NSValue may not respond to +valueWithCMTime:' but I'm not sure what I'm doing wrong. In my header file, I have this:
#import <Foundation/Foundation.h>
#import <CoreMedia/CoreMedia.h>
...
-(void)displayTime:(CMTime)time;
and in my implementation file, I have this:
-(void)displayTime:(CMTime)time
{
//This line has the warning
[mutableArray addObject:[NSValue valueWithCMTime:time]];
}
I've used the above line elsewhere in another class and there's no warning, so what could be missing from this one? Is it something to do with the way the method is set up?
You've probably forgotten to import the AVFoundation headers.
#import <AVFoundation/AVFoundation.h>
The iPhone class NSValue has no valueWithCMTime.
Check out the online documentation
It is added by AVFoundation framework, as documented here.
Added the corresponding header inclusion :
#import <AVFoundation/AVFoundation.h>
Make an NSMutableArray.... because you need an NSValue