I'm writing a couple classes that make use of the Objective-C runtime library. This includes retrieving Protocol definitions at runtime based on their name. However, it appears that Protocols that are not explicitly adopted by a class or referenced in code using #protocol(ProtocolName) are excluded from compilation and not available at runtime.
Example:
#protocol MyProtocol <NSObject>
-(void)doSomething;
#end
//Somewhere else in code
Protocol *protocol = NSProtocolFromString(#"MyProtocol");
// ^ value of "protocol" will be nil when I run the application!
//However, if I use do the following:
Protocol *whyDoIHaveToDoThis = #protocol(MyProtocol);
Protocol *protocol = NSProtocolFromString(#"MyProtocol");
// ^ value of "protocol" will now be a pointer as expected when I run the application!
Does anyone know why this is, and even better, how to force the compiler to include Protocol definitions that are not in use at compile time, but which I will later want available at runtime?
You can force the compiler to include the protocol by making a dummy method that's not called that uses it. I've done this before:
void preserveProtocolFromBeingTrimmed()
{
(void)#protocol(BrightnessProtocol);
}
I see that Apple uses this in their FxBrightness plug-in from the FxPlug SDK.
Related
I've read along here and here about conforming to Swift protocols in Objective-C headers, and I'm not quite getting the behaviour I want - I'm also looking for a better understanding of how this all works. Here's my scenario.
I have a protocol in Swift:
PersonalDetailsProtocol.swift
#objc
protocol PersonalDetailsProtocol {
func doSomeWork()
}
I then have an Objective-C class with a header and implementation file
RegistrationPersonalDetails.h
#protocol PersonalDetailsProtocol; // Forward declare my Swift protocol
#interface RegistrationPersonalDetails : NSObject <PersonalDetailsProtocol>
#end
RegistrationPersonalDetails.m
#import "RegistrationPersonalDetails.h"
#implementation RegistrationPersonalDetails
- (void)doSomeWork {
NSLog(#"Working...");
}
#end
At this point everything compiles, although there is a warning in the RegistrationPersonalDetails.h file stating Cannot find protocol definition for 'PersonalDetailsProtocol'. Other than that warning, the issue I'm facing is I can't publicly call the doSomeWork method on an instance of RegistrationPersonalDetails.
The call site in Swift would look something like:
let personalDetails = RegistrationPersonalDetails()
personalDetails.doSomeWork()
but I get an error stating:
Value of type 'RegistrationPersonalDetails' has no member 'doSomeWork'
I get that the method isn't public, because it's not declared in the header file. But I didn't think it should have to be as long as the protocol conformance is public i.e. declared in the header file.
Can anyone point me on the right path here and offer an explanation? Or is this even possible? I can obviously rewrite the protocol in ObjC, but I just always try to add new code as Swift.
In pure objective-c, you cannot make a class conform to a protocol without importing it in the header. To make the use of a protocol private, the conformance shouldn't be declared in the header. It remains possible to call such a method using a cast however, but it should be done with caution because it's a little bit unsafe :
if ([anInstance respondToSelector:#selector(myProtocolMethod)])
[(id<MyProtocol>)anInstance myProtocolMethod];
I'm not familiar with Swift, but I think you can do the same this way (or something close to it) :
if let conformingInstance = anInstance as? MyProtocol
conformingInstance.myProtocolMethod
EDIT : To complete my first assertion, forward declarations can still be used in the header when you need to declare a method receiving or returning an instance conforming to that protocol :
#SomeProtocol;
// This is not possible
#interface MyClass : NSObject <SomeProtocol>
// But this is possible
#property (nonatomic) id<SomeProtocol> someProperty;
-(void) someMethod:(id<SomeProtocol>)object;
#end
In this document Apple clearly said that :
Forward declarations of Swift classes and protocols can be used only
as types for method and property declarations.
So it seems that the rule is the same whatever the protocol is an Objective-c protocol or a Swift protocol.
I've been looking for documents of reasons of why a singleton method definition in Objective-C class is not available on the Swift interface on Xcode.
The Objective-C class is defined like this
/**
* A general accessor across the sample App to remove the dependency of QPSLibraryManager from resusable components.
*/
#interface QPSSharedAccessor : NSObject
/**
* Required by QPSLibraryManager and some UI components.
*/
#property (nonatomic, strong) QPSApplicationConfiguration *qpsConfiguration;
/**
* Provides Commands to app
*/
#property (nonatomic, strong) id<QPSAppController> appController;;
/**
* Shared singleton.
*/
+ (instancetype)sharedAccessor;
/**
* Returns access to a configuration manager
*/
- (QPSConfigurationManager *)configurationManager;
#end
On Swift Interface, its defined like this
/**
* A general accessor across the sample App to remove the dependency of QPSLibraryManager from resusable components.
*/
open class QPSSharedAccessor : NSObject {
/**
* Required by QPSLibraryManager and some UI components.
*/
open var qpsConfiguration: QPSApplicationConfiguration!
/**
* Provides Commands to app
*/
open var appController: QPSAppController!
/**
* Returns access to a configuration manager
*/
open func configurationManager() -> QPSConfigurationManager!
}
I'm expected to see the sharedAccessor() singleton method on Swift but it is missing as you can see. Calling the said method on a separate swift file results in a compiler error, saying that the sharedAccessor() method doesn't exist. Converting everything to Swift is not viable btw. Advice on fixing this problem?
After some experimentation, I see that sharedAccessor() appears to have some special meaning. Using a different name for it, e.g. sharedAccessor1() or sharedInstance(), worked for me just fine. One related observation is that trying to call sharedAccessor() when it is NOT part of the interface results in this error: Type QPSSharedAccessor has no member sharedAccessor, which makes sense. However, adding sharedAccessor() to the Objective-C code results in:
'sharedAccessor()' is unavailable: use object construction 'QPSSharedAccessor()'
Opening the error details sheds more light on this:
'sharedAccessor()' has been explicitly marked unavailable here (__ObjC.QPSSharedAccessor)
Now, renaming the QPSSharedAccessor type to something else makes sharedAccessor() acceptable, but if the new type name is, for example, QPSMyClass, then naming the method myClass() becomes a problem!
To work around this strange problem, evidently having to do with compiler internals, obscure method naming conventions, or a bug, you can simply rename the sharedAccessor() method to something else. Alternatively, you can write a wrapper method in C or Objective-C and make it available to Swift via the bridging header. For example:
QPSSharedAccessor * getGlobalQPSSharedAccessor()
{
return [QPSSharedAccessor sharedAccessor];
}
You could also add a category to QPSSharedAccessor with a method having a different name and delegating to sharedAccessor().
Also, please see if these references are useful:
Can't build in Xcode with error "[method] has been explicitly marked unavailable"
https://developer.apple.com/library/content/documentation/Swift/Conceptual/BuildingCocoaApps/AdoptingCocoaDesignPatterns.html (see the Singleton section).
Working on a mixed framework. imported inside the Obj-C file but the internal classes are not visible, only the public ones.
The documentation clearly states the internal clasees should be available between Swift and Obj-C:
Importing Swift into Objective-C To import a set of Swift files in the same framework target as your Objective-C code, you don’t
need to import anything into the umbrella header for the framework.
Instead, import the Xcode-generated header file for your Swift code
into any Objective-C .m file you want to use your Swift code from.
Because the generated header for a framework target is part of the
framework’s public interface, only declarations marked with the public
modifier appear in the generated header for a framework target. You
can still use Swift methods and properties that are marked with the
internal modifier from within the Objective-C part of your framework,
as long they are declared within a class that inherits from an
Objective-C class. For more information on access-level modifiers, see
Access Control in The Swift Programming Language (Swift 2).
Code Sample (Create a new project with a framework)
// SwiftObject.swift
public class SwiftObject: NSObject {
public class func doSomething() {}
}
internal class YetAnotherSwiftObject: NSObject {
internal class func doSomething() {}
}
// SomeObject.m file
#implementation SomeObject
- (void)someMethod {
[SwiftObject doSomething];
}
- (void)someOtherMethod {
[YetAnotherSwiftObject doSomething]; // Use of undeclared identifier
}
#end
As indicated in the docs, declarations marked with internal modifier don't appear in the generated header, so the compiler does not know about them and thus complaints. Of course, you could send messages using performSelector approach, but that's not convenient and bug-prone. We just need to help the compiler know that those declarations are there.
First, we need to use #objc attribute variant that allows you to specify name for your symbol in Objective-C:
// SwiftObject.swift
#objc(SWIFTYetAnotherSwiftObject)
internal class YetAnotherSwiftObject: NSObject {
internal class func doSomething() {}
}
And then you just need to create #interface declaration with the methods you want to use in your code - so the compiler will be happy, and also apply SWIFT_CLASS macro with the symbol name you've specified earlier - so the linker would pick the actual implementation:
// SomeObject.m file
SWIFT_CLASS("SWIFTYetAnotherSwiftObject")
#interface YetAnotherSwiftObject : NSObject
+ (void)doSomething;
#end
#implementation SomeObject
- (void)someOtherMethod {
[YetAnotherSwiftObject doSomething]; // Should work now !!!
}
#end
I've used the interface declaration in .m file just for clarity, the better option would be to combine such declarations in .h file, and include it.
By declaring methods in that interface we're making a promise to compiler, and it won't complain if you'll put there a method that does not exist (or with wrong signature, etc.) Obviously, you'll crash in runtime in that case - so be cautious.
For me it just worked by checking: "Allow app extension API only". You find it by going to the project setting, select your target and then it is in the General tab under Deployment Info.
Can someone explain to me, why this does solve the problem?
While the above solution works (https://stackoverflow.com/a/33159964/5945317), it seems overly complicated and unintuitive:
Complicated, because it seems to add more things than necessary – I will provide a smoother solution below.
Unintuitive, because the objc macro SWIFT_CLASS resolves to SWIFT_RUNTIME_NAME, and the provided value is not actually the runtime name – nor is the objc class name in the header matching the Swift attribute param in #objc. Still, surprisingly, the solution works – but to me it is not clear why.
Here is what we have tested in our own project, and believe to be the better solution (using the example above):
// YetAnotherSwiftObject.swift
#objc(OBJCPREFIXYetAnotherSwiftObject)
internal class YetAnotherSwiftObject: NSObject {
#objc internal class func doSomething() {}
}
// OBJCPREFIXYetAnotherSwiftObject.h
#interface OBJCPREFIXYetAnotherSwiftObject : NSObject
+ (void)doSomething;
#end
That's it. The interface looks like a regular objc interface. This gives the added benefit that you can include it in other header files (which you cannot do if you use the SWIFT_CLASS macro, as it comes from the autogenerated Swift header file, which in turn you cannot include in an objc header, due to circular dependency).
On the Swift side, the only thing relevant is that you provide the class with the proper objc name. Mind that I only used the name prefix for language consistency – you can even just use YetAnotherSwiftObject everywhere (i.e., in the objc header and in the #objc attribute in Swift – but you need to keep this attribute with explicit naming in any case, and need to keep it consistent with the class name in the header).
This also makes your life easier if you're in the process of converting your objc framework step by step to Swift. You just keep the objc header as before, and now provide the implementation in Swift.
Methods and properties that are marked with the internal modifier and declared within a class that inherits from an Objective-C class are accessible to the Objective-C runtime.
so let's make use of that:
class MyInternalClass: NSObject {
#objc var internalProperty = 42
}
#interface MyPublicClass()
#end
#implementation MyPublicClass
+ (void) printValue {
Class myInternalClass = NSClassFromString(#"MyPackageNameIfAny.MyInternalClass");
id myInternalClassInstance = [myInternalClass new];
int value = [myInternalClassInstance performSelector:#selector(internalProperty)];
NSLog(#"Value is %d ", value); // "value is 42"
}
#end
Using the SWIFT_CLASS macro and #objc class attribute could easily lead to errors when archiving. This approach is safer.
In my class, I have a reference on an UIViewController and want to implement a delegate on this ViewController during runtime. The delegate has only one method (with two parameters) and when the delegate-method on the ViewController is invoked, my class should handle the call.
I am quite sure this is possible with some kind of method swizzling, etc. but I don't know how to accomplish this.
What you want is possible, but it's not method swizzling, since you don't want to switch to methods but add a new one. It can be done, thanks to Objective-C's dynamic nature, but it's still a dirty hack so also file a feature request with the library vendor.
What you want is class_addMethod() and a C function with the actual implementation for that. One more thing, Objective-C methods are C methods, but with two implicit parameters, self and _cmd, which have to keep in mind (both when creating your C method and when telling class_addMethod your methods signature. And here is an SSCE of how to pull something like that off:
#import <Foundation/Foundation.h>
#import <objc/runtime.h> // Required for class_addMethod()
#interface MyClass : NSObject
#end
#implementation MyClass
#end
#protocol MyProtocol <NSObject>
- (void)printString:(NSString *)string;
#end
// Note the method signature containing the
// two implicit parameters self and _cmd!
void MyClassPrinStringIMP(id self, SEL _cmd, NSString *string)
{
NSLog(#"Hi I'm %#:%s and this is the string: %#", self, sel_getName(_cmd), string);
}
void PimpMyClass()
{
// The last argument is the signature. First character is the return type, in our case void
// Then comes self and _cmd, followed by the NSString. You can use #encode() to find out how your
// type is encoded. Best is to build this string at runtime, since the encoding can change with architectures
class_addMethod([MyClass class], #selector(printString:), (IMP)MyClassPrinStringIMP, "v#:#");
}
int main(int argc, const char * argv[])
{
#autoreleasepool
{
PimpMyClass();
id foo = [[MyClass alloc] init]; // id, to silence the compiler!
[foo printString:#"Hello World"];
}
return 0;
}
Example output:
Hi I'm <MyClass: 0x100101810>:printString: and this is the string: Hello World
Edit: Something that you may find is that the passed object is checked at runtime wether it conforms to a protocol or not using conformsToProtocol:. Since this code just adds the method implementation, it would still fail, but you can tell the runtime that you totally do implement that protocol with this one function call:
class_addProtocol([MyClass class], #protocol(MyProtocol));
Alternative: proxies
Objective-Cs dynamism and message forwarding is already praised by #JasperBlues, however, there is one particular class in Objective-C that is designed to do just that: NSProxy. It is designed to intercept sent messages and dispatching them dynamically to the relevant target, and does use the high-level NSInvocation approach. If you can pass a proxied object in some way as the delegate (depending on what your code allows for and what not), creating a NSProxy subclass might be the cleanest way to go.
However, note though that you then end up with a shim object that wraps over your other object, which comes with its own bag of pain and will break when you try to directly access variables via -> syntax. It's not a perfectly invisible proxy, but good enough for most cases.
Firstly, some comments indicate that what you're asking is instantly "a bad thing to do" or a "dirty hack". I disagree here. Most modern Object Oriented languages support these features, and they are used to good effect by numerous system-level frameworks. Of course it is human-nature to perhaps use these dynamic features where they're not really required (for fun or practice), even when a simpler approach would work fine. Beware of this.
Objective-C is admirable in that its somewhat of a legacy language and close to the "bare metal", and yet features a surprising level of dynamism, making it relatively easy to support these requirements without any external libraries or frameworks.
Besides using the class_addMethod guide that another answer correctly indicates, some other approaches are:
Message Forwarding: (recommended)
All NSObject sub-classes have the ability to forward a method that they're not able to respond to, to another target object. This is similar to the lower-level concept of trampolines. Apple publishes a guide on using this approach.
The advantages of using forward invocation is that it uses the NSInvocation level of abstraction, instead of directly calling the C ObjC runtime API. This abstracts the following details away:
Structs and primitives will be box/unboxed automatically
Dispatching to methods with a dynamic/unknown number of arguments becomes easy. Until arm64, this could be done using va_args, however on arm64 va_args can be copied directly to registers, and not popped off the stack.
Resolve Instance Method:
Instance methods are created by by registering a C function as the implementation to respond to a given message. This can be done neatly with blocks, using IMP_ImplementationWithBlock:
+ (BOOL)resolveInstanceMethod:(SEL)sel
{
IMP imp = imp_implementationWithBlock((__bridge id) objc_unretainedPointer(
^(id me, BOOL firstParam, NSString* secondParam)
{
//Implementation goes in here
return something; //something of type 'id'
}));
class_addMethod(self, sel, imp, "##:");
return YES;
}
return NO;
}
Using libffi:
Libffi can also do this kind of thing, though it should not be necessary if you're using pure Objective-C, as the runtime already has these features baked in.
I've made use of Class Extensions in the .m as a way to have "private" methods and variables. I've read that since Xcode 4.4, the compiler no longer needed the private methods declared.
For example this would compile even though helperMethodC is not declared:
in .h
#interface MyClass : NSObject
-(void)publicMethodA;
#end
in .m
#interface MyClass ()
- (void) pseudoPrivateMethodB;
#end
#implementation MyClass
- (void)publicMethodA
{
//Do Something
}
- (void)pseudoPrivateMethodB
{
[self helperMethodC];
}
- (void) helperMethodC
{
// Do something
}
While private methods no longer have to be declared to compile (helperMethodC), is there a style guide, historical reason, or rule that all private methods (i.e. helperMethodC) should still be declared? Or a "rule" for when to declare and not declare private methods?
Declare them if they help you. From a documentation point of view they are very useful. The compiler will also tell you if you have specified that a method will exist and then not implemented it. There is no rule, but its a good idea to add them. Consider how you'll feel if you have to come back in 6 months and edit the class - will having the methods listed there help you?
While private methods no longer have to be declared to compile (helperMethodC), is there a style guide, historical reason, or rule that all private methods (i.e. helperMethodC) should still be declared? Or a "rule" for when to declare and not declare private methods?
There are multiple conventions, but no standard.
You really should have them when/if you need to support older toolchains -- GCC or older versions of Clang.
Once that restriction is removed, I think it's best if you just phase the (redundant) declarations out where they are not needed. High warning levels and ARC semantics can guide you here.
If you introduce types:
Something * s = [array objectAtIndex:i];
s.string = #"string";
// rather than: [array objectAtIndex:i].string = #"string";
And name your selectors uniquely for the parameter/return types:
// BAD: May be ambiguous.
// Solution: Choose a more descriptive selector name.
// class A
- (void)setX:(int)x;
// class B
- (void)setX:(double)x;
then the compiler can inform you of ambiguities.