13.iOS一些面试题分析

一、什么是Runtime

• runtime是由C和C++、汇编实现的一套API，为OC语言加入了面向对象、运行时的功能
• 运行时（runtime）将数据类型的确定由编译时推迟到了运行时
• 平时编写的OC代码，在程序运行过程中，最终会转换成runtime的C语言代码——runtime是Objective-C 的幕后⼯作者

如类结构中的ro和rw属性

• ro（read-only）在编译时已经确定
• rw（read-write）在运行时才确定，因此可以使用runtime进行修改

二、方法的本质是什么

方法的本质是发送消息objc_msgSend，即寻找IMP的过程

发送消息会有以下⼏个流程：

• 1.快速查找流程——通过汇编objc_msgSend查找缓存cache_t是否有imp实现
• 2.慢速查找流程——通过C++中lookUpImpOrForward递归查找当前类和父类的rw中methodlist的方法
• 3.动态方法解析——通过调用resolveInstanceMethod和resolveClassMethod来动态方法决议——实现消息动态处理
• 4.快速转发流程——通过CoreFoundation来触发消息转发流程，forwardingTargetForSelector实现快速转发，由其他对象来实现处理方法
• 5.慢速转发流程——先调用methodSignatureForSelector获取到方法的签名，生成对应的invocation；再通过forwardInvocation来进行处理
• 6.以上流程均无法挽救就崩溃并报错

三、SEL和IMP的关系

SEL是方法编号，也是方法名，在dyld加载镜像到内存时，通过_read_image方法加载到内存的表中了

IMP是函数实现指针，找IMP就是找函数实现的过程

SEL和IMP的关系就可以解释为：

• SEL就相当于书本的⽬录标题
• IMP就是书本的⻚码
• 函数就是具体页码对应的内容

四、能否向运⾏时创建的类中添加实例变量

具体情况具体分析：

• 编译好的类不能添加实例变量
• 运行时创建的类可以添加实例变量，但若已注册到内存中就不行了

原因：

• 编译好的实例变量存储的位置在ro，而ro是在编译时就已经确定了的
• ⼀旦编译完成，内存结构就完全确定就⽆法修改
• 只能修改rw中的方法或者可以通过关联对象的方式来添加属性

五、isKindOfClass 和 isMemberOfClass

#import "DHPerson.h"
#import <objc/runtime.h>

int main(int argc, const char * argv[]) {
    @autoreleasepool {

        BOOL re1 = [(id)[NSObject class] isKindOfClass:[NSObject class]];//1
        BOOL re2 = [(id)[NSObject class] isMemberOfClass:[NSObject class]];// 0
        BOOL re3 = [(id)[DHPerson class] isKindOfClass:[DHPerson class]];//0
        BOOL re4 = [(id)[DHPerson class] isMemberOfClass:[DHPerson class]];// 0
        NSLog(@" re1 :%hhd\n re2 :%hhd\n re3 :%hhd\n re4 :%hhd\n",re1,re2,re3,re4);

        BOOL re5 = [(id)[NSObject alloc] isKindOfClass:[NSObject class]];//1
        BOOL re6 = [(id)[NSObject alloc] isMemberOfClass:[NSObject class]];// 1
        BOOL re7 = [(id)[DHPerson alloc] isKindOfClass:[DHPerson class]];//1
        BOOL re8 = [(id)[DHPerson alloc] isMemberOfClass:[DHPerson class]];// 1
        NSLog(@" re5 :%hhd\n re6 :%hhd\n re7 :%hhd\n re8 :%hhd\n",re5,re6,re7,re8);
    }
    return 0;
}

我们先看一下源码的实现

5.1 +isKindOfClass

+ (Class)class {
    return self;
}
// 获取元类
Class object_getClass(id obj) {
    if (obj) return obj->getIsa();
    else return Nil;
} 

+ (BOOL)isKindOfClass:(Class)cls { 
    for (Class tcls = object_getClass((id)self); tcls; tcls = tcls->superclass) {
        if (tcls == cls) return YES;
    }
    return NO;
} 

+isKindOfClass是元类及其父类 vs 类

5.2 +isMemberOfClass

+ (BOOL)isMemberOfClass:(Class)cls {
    return object_getClass((id)self) == cls;
}

相较于+isKindOfClass少了父类的比较，因此+isMemberOfClass为YES时, +isKindOfClass一定为YES

+isMemberOfClass是元类 vs 类

结合isa走位图 可以得出前面四个打印结果：

• NSObject元类与NSObject类不相等，NSObject元类的父类（指向NSObject类）与NSObject类相等——YES
• NSObject元类与NSObject类不相等——NO
• DHPerson元类与DHPerson类不相等，DHPerson元类的父类与DHPerson类不相等——NO
• DHPerson元类与DHPerson类不相等——NO

5.3 -isKindOfClass

- (BOOL)isKindOfClass:(Class)cls {
    for (Class tcls = [self class]; tcls; tcls = tcls->superclass) {
        if (tcls == cls) return YES;
    }
    return NO;
}   

-isKindOfClass是类本身及其父类 vs 类

5.4 -isMemberOfClass

- (BOOL)isMemberOfClass:(Class)cls {
    return [self class] == cls;
}

-isMemberOfClass是当前类和cls作比较，-isKindOfClass多了一步for循环类对象的父类

-isMemberOfClass是类本身 vs 类

后面四个结果分析如下：

• NSObject类与NSObject类相等——YES
• NSObject类与NSObject类相等——YES
• DHPerson类与DHPerson类相等——YES
• DHPerson类与DHPerson类相等——YES

六、 [self class]和[super class]

DHSon 继承于 DHFather，主程序初始化 DHSon，求问打印内容以及思路

#import "DHSon.h"
@implementation DHSon

- (instancetype)init {
    self = [super init];
    if (self) {
        NSLog(@"[self class] = %@", NSStringFromClass([self class]));//DHSon
        NSLog(@"[super class] = %@", NSStringFromClass([super class]));//DHSon
        NSLog(@"[self superclass] = %@", NSStringFromClass([self superclass]));//DHFather
        NSLog(@"[super superclass] = %@", NSStringFromClass([super superclass]));//DHFather
    }
    return self;
}

@end

结果有点出乎意料，我们查看一下源码

+ (Class)class {
    return self;
}

- (Class)class {
    return object_getClass(self);
}

+ (Class)superclass {
    return self->superclass;
}

- (Class)superclass {
    return [self class]->superclass;
}
 

从这段代码能解释[self class]和[self superclass]，但是另外两个又怎么解释呢？

终端clang编译代码得到super.cpp，就能看到初始化的底层代码了(也可以断点查看汇编)

clang -rewrite-objc DHSon.m -o super.cpp

可见[super class]和[super superclass]底层都是调用的objc_msgSendSuper

objc_msgSendSuper(void /* struct objc_super *super, SEL op, ... */ )
    OBJC_AVAILABLE(10.0, 2.0, 9.0, 1.0, 2.0);
    
/// Specifies the superclass of an instance. 
struct objc_msgSendSuper {
    /// Specifies an instance of a class.
    __unsafe_unretained _Nonnull id receiver;

    /// Specifies the particular superclass of the instance to message. 
#if !defined(__cplusplus)  &&  !__OBJC2__
    /* For compatibility with old objc-runtime.h header */
    __unsafe_unretained _Nonnull Class class;
#else
    __unsafe_unretained _Nonnull Class super_class;
#endif
    /* super_class is the first class to search */
}; 

使用objc_msgSendSuper向objc_super发送消息，而objc_super在objc2.0下有两个元素： id类型的receiver和Class类型的super_class

我们仿写一下

struct objc_super lg_super = {
    self,
    class_getSuperclass([self class]),
};

NSLog(@"%@",NSStringFromClass(objc_msgSendSuper(&lg_super,@selector(class))));

就能得到和以下一样的结果了

NSLog(@"[super class] = %@", NSStringFromClass([super class]));

记得导入<objc/message.h>，报错Too many arguments就去修改编译期配置

那苹果为什么要这么设计呢？把消息查找和isa走位图联系起来就明白了！

son实例对象的实例方法存在DHSon类中

• 调用[self class]就是son照着DHSon->DHFather->NSObject顺序问老爸要-class方法

• 调用[super class]就是son跳过DHSon，直接通过DHFather->NSObject查找

• 还有比[super class]更快找到class方法的写法, lg_super结构体中[self class]改为[super class]，直接找到NSObjct

• 当lg_super结构体中[self class]改为[DHFather class]时，因为类方法存在元类中，会按DHFather元类->NSObject元类->NSObject根元类找-class方法，最后也是会输出DHSon

结论：

• [self class]就是发送消息objc_msgSend，消息接收者是self，方法编号是class
• [super class]就是发送消息objc_msgSendSuper，消息接收者是self，方法编号是class，只不过objc_msgSendSuper会跳过self的查找

七、 weak置空原理

int main(int argc, const char * argv[]) {
    @autoreleasepool {
        DHPerson *person = [[DHPerson alloc] init];
        id __weak person = object;
    }
    return 0;
}

id __weak一行打断点，Debug->Debug Workflow->Always show Disassembly查看汇编——汇编代码会来到libobjc库的objc_initWeak

7.1objc_initWeak

id objc_initWeak(id *location, id newObj)
{
    if (!newObj) {
        *location = nil;
        return nil;
    }

    return storeWeak<DontHaveOld, DoHaveNew, DoCrashIfDeallocating>
        (location, (objc_object*)newObj);
} 

location：表示__weak指针的地址（我们研究的就是__weak指针指向的内容怎么置为nil）

newObj：所引用的对象，即例子中的 person

7.2storeWeak

查看storeWeak源码，根据注释，可以知道如下几点

• HaveOld：weak指针之前是否已经指向了一个弱引用
• HaveNew：weak指针是否需要指向一个新引用
• CrashIfDeallocating：如果被弱引用的对象正在析构，此时再弱引用该对象，是否应该crash。

// Update a weak variable.
// If HaveOld is true, the variable has an existing value 
//   that needs to be cleaned up. This value might be nil.
// If HaveNew is true, there is a new value that needs to be 
//   assigned into the variable. This value might be nil.
// If CrashIfDeallocating is true, the process is halted if newObj is 
//   deallocating or newObj's class does not support weak references. 
//   If CrashIfDeallocating is false, nil is stored instead.
enum CrashIfDeallocating {
    DontCrashIfDeallocating = false, DoCrashIfDeallocating = true
};
template <HaveOld haveOld, HaveNew haveNew,
          CrashIfDeallocating crashIfDeallocating>
static id 
storeWeak(id *location, objc_object *newObj)
{
    assert(haveOld  ||  haveNew);
    if (!haveNew) assert(newObj == nil);

    Class previouslyInitializedClass = nil;
    id oldObj;
    SideTable *oldTable;
    SideTable *newTable;

    // Acquire locks for old and new values.
    // Order by lock address to prevent lock ordering problems. 
    // Retry if the old value changes underneath us.
 retry:
    ///🌹 如果weak指针之前弱引用过一个obj，则将这个obj所对应的SideTable取出，赋值给oldTable
    if (haveOld) {
        oldObj = *location;
        oldTable = &SideTables()[oldObj];
    } else {
        // 没有弱引用过，则oldTable = nil
        oldTable = nil;
    }
  ///🌹 如果weak指针要弱引用一个新的obj，则将该obj对应的SideTable取出，赋值给newTable
    if (haveNew) {
        newTable = &SideTables()[newObj];
    } else {
        newTable = nil;
    }
    //加锁操作，防止多线程中竞争冲突
    SideTable::lockTwo<haveOld, haveNew>(oldTable, newTable);
    // location 应该与 oldObj 保持一致，如果不同，说明当前的 location 已经处理过 oldObj 可是又被其他线程所修改
    if (haveOld  &&  *location != oldObj) {
        SideTable::unlockTwo<haveOld, haveNew>(oldTable, newTable);
        goto retry;
    }

    // Prevent a deadlock between the weak reference machinery
    // and the +initialize machinery by ensuring that no 
    // weakly-referenced object has an un-+initialized isa.
    if (haveNew  &&  newObj) {
        Class cls = newObj->getIsa();
        // 如果cls还没有初始化，先初始化，再尝试设置弱引用
        if (cls != previouslyInitializedClass  &&  
            !((objc_class *)cls)->isInitialized()) 
        {
            SideTable::unlockTwo<haveOld, haveNew>(oldTable, newTable);
            _class_initialize(_class_getNonMetaClass(cls, (id)newObj));

            // If this class is finished with +initialize then we're good.
            // If this class is still running +initialize on this thread 
            // (i.e. +initialize called storeWeak on an instance of itself)
            // then we may proceed but it will appear initializing and 
            // not yet initialized to the check above.
            // Instead set previouslyInitializedClass to recognize it on retry.
            // 完成初始化后进行标记
            previouslyInitializedClass = cls;
            //newObj 初始化后，重新获取一遍newObj
            goto retry;
        }
    }

    // Clean up old value, if any.
  ///🌹 如果weak指针之前弱引用过别的对象oldObj，则调用weak_unregister_no_lock，在oldObj的weak_entry_t中移除该weak指针地址
    if (haveOld) {
        weak_unregister_no_lock(&oldTable->weak_table, oldObj, location);
    }

    // Assign new value, if any.
   ///🌹 如果weak指针需要弱引用新的对象newObj
    if (haveNew) {
      ///🌹 调用weak_register_no_lock方法，将weak指针的地址记录到newObj对应的weak_entry_t中
        newObj = (objc_object *)
            weak_register_no_lock(&newTable->weak_table, (id)newObj, location, 
                                  crashIfDeallocating);
        // weak_register_no_lock returns nil if weak store should be rejected

        // Set is-weakly-referenced bit in refcount table.
         //更新newObj的isa指针的weakly_referenced bit标志位
        if (newObj  &&  !newObj->isTaggedPointer()) {
            newObj->setWeaklyReferenced_nolock();
        }

        // Do not set *location anywhere else. That would introduce a race.
       //*location 赋值，也就是将weak指针直接指向了newObj，而且没有将newObj的引用计数+1
        *location = (id)newObj;
    }
    else {
        // No new value. The storage is not changed.
    }
    
    SideTable::unlockTwo<haveOld, haveNew>(oldTable, newTable);

    return (id)newObj;
} 

因为我们这里是第一次调用，所以是一个新的对象，也就是haveNew的情况，获取到的是新的散列表SideTable，主要执行了weak_register_no_lock方法来进行插入。

7.3weak_register_no_lock

id weak_register_no_lock(weak_table_t *weak_table, id referent_id, 
                      id *referrer_id, bool crashIfDeallocating)
{
  //先获取需要弱引用对象
    objc_object *referent = (objc_object *)referent_id;
    objc_object **referrer = (objc_object **)referrer_id;
   //果被弱引用对象referent为nil 或者被弱引用对象采用了TaggedPointer计数方式，则直接返回
    if (!referent  ||  referent->isTaggedPointer()) return referent_id;

    // ensure that the referenced object is viable
    //确保被引用的对象可用（没有在析构，同时应该支持weak弱引用）
    bool deallocating;
    if (!referent->ISA()->hasCustomRR()) {
        deallocating = referent->rootIsDeallocating();
    }
    else {
        BOOL (*allowsWeakReference)(objc_object *, SEL) = 
            (BOOL(*)(objc_object *, SEL))
            object_getMethodImplementation((id)referent, 
                                           SEL_allowsWeakReference);
        if ((IMP)allowsWeakReference == _objc_msgForward) {
            return nil;
        }
        deallocating =
            ! (*allowsWeakReference)(referent, SEL_allowsWeakReference);
    }
   //如果是正在析构的对象，那么不能够被弱引用
    if (deallocating) {
        if (crashIfDeallocating) {
            _objc_fatal("Cannot form weak reference to instance (%p) of "
                        "class %s. It is possible that this object was "
                        "over-released, or is in the process of deallocation.",
                        (void*)referent, object_getClassName((id)referent));
        } else {
            return nil;
        }
    }

    // now remember it and where it is being stored
   ///🌹在 weak_table 中找到被弱引用对象 referent 对应的 weak_entry,并将 referrer 加入到 weak_entry 中
    weak_entry_t *entry;
    if ((entry = weak_entry_for_referent(weak_table, referent))) {
        //如果能找到 weak_entry,则将referrer 插入到 weak_entry 中
        append_referrer(entry, referrer);
    } 
    else {
       //如果找不到 weak_entry，就新建一个
        weak_entry_t new_entry(referent, referrer);
        weak_grow_maybe(weak_table);
        weak_entry_insert(weak_table, &new_entry);
    }

    // Do not set *referrer. objc_storeWeak() requires that the 
    // value not change.

    return referent_id;
}

• 如果可以被弱引用，则将被弱引用对象所在的weak_table中的weak_entries哈希数组中取出对应的weak_entry_t，然后将指向被弱引用对象地址的指针referrer通过函数append_referrer插入到对应的weak_entry_t引用数组。
• 如果weak_entry_t不存在，则会新建一个。weak_grow_maybe判断要不要扩容，然后weak_entry_insert插入到weak表
• 至此就完成了弱引用。

7.4 append_referrer

主要是找到弱引用对象的对应的weak_entry哈希数组中，基本就是个遍历插入的过程

static void append_referrer(weak_entry_t *entry, objc_object **new_referrer)
{
    //果weak_entry 使用静态数组 inline_referrers
    if (! entry->out_of_line()) {
        // Try to insert inline.
        //试将 referrer 插入数组
        for (size_t i = 0; i < WEAK_INLINE_COUNT; i++) {
            if (entry->inline_referrers[i] == nil) {
                entry->inline_referrers[i] = new_referrer;
                return;
            }
        }

        // Couldn't insert inline. Allocate out of line.
        //如果inline_referrers的位置已经存满了，则要转型为 referrers，动态数组
        weak_referrer_t *new_referrers = (weak_referrer_t *)
            calloc(WEAK_INLINE_COUNT, sizeof(weak_referrer_t));
        // This constructed table is invalid, but grow_refs_and_insert
        // will fix it and rehash it.
        for (size_t i = 0; i < WEAK_INLINE_COUNT; i++) {
            new_referrers[i] = entry->inline_referrers[i];
        }
        entry->referrers = new_referrers;
        entry->num_refs = WEAK_INLINE_COUNT;
        entry->out_of_line_ness = REFERRERS_OUT_OF_LINE;
        entry->mask = WEAK_INLINE_COUNT-1;
        entry->max_hash_displacement = 0;
    }

    assert(entry->out_of_line());
   //如果动态数组中元素个数大于或等于数组总空间的3/4，则扩展数组空间为当前长度的一倍，然后将 referrer 插入数组
    if (entry->num_refs >= TABLE_SIZE(entry) * 3/4) {
        return grow_refs_and_insert(entry, new_referrer);
    }
   //如果不需要扩容，直接插入到weak_entry中
   //& (entry->mask) 保证 begin 的位置只能大于或等于数组的长度
    size_t begin = w_hash_pointer(new_referrer) & (entry->mask);
    size_t index = begin;
    size_t hash_displacement = 0;
    while (entry->referrers[index] != nil) {
        hash_displacement++;
        index = (index+1) & entry->mask;
        if (index == begin) bad_weak_table(entry);
    }
    if (hash_displacement > entry->max_hash_displacement) {
        entry->max_hash_displacement = hash_displacement;
    }
    weak_referrer_t &ref = entry->referrers[index];
    ref = new_referrer;
    entry->num_refs++;
} 

7.5 weak_unregister_no_lock

如果weak指针在指向obj之前，已经弱引用了其他的对象，则需要先将weak指针从其他对象的weak_entry_t的hash数组中移除。在storeWeak方法中会调用weak_unregister_no_lock函数来做移除操作

weak_unregister_no_lock函数首先会在weak_table中找出以前被弱引用的对象referent对应的weak_entry_t，在weak_entry_t中移除被弱引用的对象referrer。移除元素后，判断此时weak_entry_t中是否还有元素。如果此时weak_entry_t已经没有元素了，则需要将weak_entry_t从weak_table中移除。

void weak_unregister_no_lock(weak_table_t *weak_table, id referent_id, 
                        id *referrer_id)
{
    //拿到以前弱引用的对象和对象的地址
    objc_object *referent = (objc_object *)referent_id;
    objc_object **referrer = (objc_object **)referrer_id;

    weak_entry_t *entry;

    if (!referent) return;
    //查找到以前弱引用的对象 referent 所对应的 weak_entry_t
    if ((entry = weak_entry_for_referent(weak_table, referent))) {
        //在以前弱引用的对象 referent 所对应的 weak_entry_t 的 hash 数组中，移除弱引用 referrer
        remove_referrer(entry, referrer);
        //移除元素之后， 要检查一下 weak_entry_t 的 hash 数组是否已经空了
        bool empty = true;
        if (entry->out_of_line()  &&  entry->num_refs != 0) {
            empty = false;
        }
        else {
            for (size_t i = 0; i < WEAK_INLINE_COUNT; i++) {
                if (entry->inline_referrers[i]) {
                    empty = false; 
                    break;
                }
            }
        }
        //如果 weak_entry_t 的hash数组已经空了，则需要将 weak_entry_t 从 weak_table 中移除
        if (empty) {
            weak_entry_remove(weak_table, entry);
        }
    }

    // Do not set *referrer = nil. objc_storeWeak() requires that the 
    // value not change.
} 

7.6 以上是weak的创建，现在看看weak的销毁

- (void)dealloc {
    _objc_rootDealloc(self);
}
**********************************
void _objc_rootDealloc(id obj)
{
    assert(obj);

    obj->rootDealloc();
}
***********************************
inline void objc_object::rootDealloc()
{
    //如果是Tagged Pointer，就直接返回
    if (isTaggedPointer()) return;  // fixme necessary?

    /*
    如果同时满足 
    1. 是优化过的isa、
    2. 没有被weak指针引用过、
    3. 没有关联对象、
    4. 没有C++析构函数、
    5. 没有sideTable，
    就可以直接释放内存free()
    */
    if (fastpath(isa.nonpointer  &&  
                 !isa.weakly_referenced  &&  
                 !isa.has_assoc  &&  
                 !isa.has_cxx_dtor  &&  
                 !isa.has_sidetable_rc))
    {
        assert(!sidetable_present());
        free(this);
    } 
    else {//否则的话就需要通过下面的函数处理
        object_dispose((id)this);
    }
} 

id 
object_dispose(id obj)
{
    if (!obj) return nil;

    objc_destructInstance(obj);    
    free(obj);

    return nil;
}

void *objc_destructInstance(id obj) 
{
    if (obj) {
        // Read all of the flags at once for performance
        bool cxx = obj->hasCxxDtor();
        bool assoc = obj->hasAssociatedObjects();

        // This order is important.
        // 如果有C++析构函数，则从运行相关函数
        if (cxx) object_cxxDestruct(obj); 
        // 如果有关联对象，则移除所有的关联对象，并将其自身从Association Manager的map中移除
        if (assoc) _object_remove_assocations(obj); 
        // 继续清理其它相关的引用
        obj->clearDeallocating(); 
    }
    return obj;
} 

inline void 
objc_object::clearDeallocating()
{
    if (slowpath(!isa.nonpointer)) {
        // Slow path for raw pointer isa.
        // 如果要释放的对象没有采用了优化过的isa引用计数
        sidetable_clearDeallocating();
    }
    else if (slowpath(isa.weakly_referenced  ||  isa.has_sidetable_rc)) {
        // Slow path for non-pointer isa with weak refs and/or side table data.
        // 如果要释放的对象采用了优化过的isa引用计数，并且有弱引用或者使用了sideTable的辅助引用计数
        clearDeallocating_slow();
    }

    assert(!sidetable_present());
} 

NEVER_INLINE void
objc_object::clearDeallocating_slow()
{
    assert(isa.nonpointer  &&  (isa.weakly_referenced || isa.has_sidetable_rc));
    // 在全局的SideTables中，以this指针(要释放的对象)为key，找到对应的SideTable
    SideTable& table = SideTables()[this];
    table.lock();
    if (isa.weakly_referenced) {
        //要释放的对象被弱引用了，通过weak_clear_no_lock函数将指向该对象的弱引用指针置为nil
        weak_clear_no_lock(&table.weak_table, (id)this);
    }
    //使用了sideTable的辅助引用计数,直接在SideTable中擦除该对象的引用计数
    if (isa.has_sidetable_rc) {
        table.refcnts.erase(this);
    }
    table.unlock();
} 

void 
weak_clear_no_lock(weak_table_t *weak_table, id referent_id) 
{
    //获取被弱引用对象的地址
    objc_object *referent = (objc_object *)referent_id;
    // 根据对象地址找到被弱引用对象referent在weak_table中对应的weak_entry_t
    weak_entry_t *entry = weak_entry_for_referent(weak_table, referent); 
    if (entry == nil) {
        /// XXX shouldn't happen, but does with mismatched CF/objc
        //printf("XXX no entry for clear deallocating %p\n", referent);
        return;
    }

    // zero out references
    weak_referrer_t *referrers;
    size_t count;
    
    // 找出弱引用该对象的所有weak指针地址数组
    if (entry->out_of_line()) {
        referrers = entry->referrers;
        count = TABLE_SIZE(entry);
    } 
    else {
        referrers = entry->inline_referrers;
        count = WEAK_INLINE_COUNT;
    }
    // 遍历取出每个weak指针的地址
    for (size_t i = 0; i < count; ++i) {
        objc_object **referrer = referrers[i]; 
        if (referrer) {
            // 如果weak指针确实弱引用了对象 referent，则将weak指针设置为nil
            if (*referrer == referent) { 
                *referrer = nil;
            }
            // 如果所存储的weak指针没有弱引用对象 referent，这可能是由于runtime代码的逻辑错误引起的，报错
            else if (*referrer) { 
                _objc_inform("__weak variable at %p holds %p instead of %p. "
                             "This is probably incorrect use of "
                             "objc_storeWeak() and objc_loadWeak(). "
                             "Break on objc_weak_error to debug.\n", 
                             referrer, (void*)*referrer, (void*)referent);
                objc_weak_error();
            }
        }
    }
    weak_entry_remove(weak_table, entry);
} 

主要流程是 dealloc->_objc_rootDealloc->rootDealloc->object_dispose->objc_destructInstance->clearDeallocating->clearDeallocating_slow->weak_clear_no_lock

7.7 小结

• 当一个对象obj被weak指针指向时，这个weak指针会以obj作为key，被存储到sideTable中的weak_table这个散列表上对应的一个weak指针数组weak_entries里面。
• 当一个对象obj的dealloc方法被调用时，Runtime会以obj为key，从sideTable的weak_table散列表中，找出对应的weak指针列表，然后将里面的weak指针逐个置为nil。

创建流程图

销毁流程图引用

八、 Method Swizzing坑点

坑点一，类簇

比如处理数组越界问题

NSArray *array = @[@"b", @"o", @"o", @"k", @"s"];
NSLog(@"%@", array[5]);
NSLog(@"%@", [array objectAtIndex:5]);

• 新建NSArray分类
• 导入runtime头文件——<objc/runtime.h>
• 写下新的方法
• 在+load利用黑魔法交换方法

#import "NSArray+EX.h"
#import <objc/runtime.h>

@implementation NSArray (EX)

+ (void)load {
    // 交换objectAtIndex方法
    Method oriMethod1 = class_getInstanceMethod(self, @selector(objectAtIndex:));
    Method swiMethod1 = class_getInstanceMethod(self, @selector(ex_objectAtIndex:));
    method_exchangeImplementations(oriMethod1, swiMethod1);
    
    // 交换取下标方法
    Method oriMethod2 = class_getInstanceMethod(self, @selector(objectAtIndexedSubscript:));
    Method swiMethod2 = class_getInstanceMethod(self, @selector(ex_objectAtIndexedSubscript:));
    method_exchangeImplementations(oriMethod2, swiMethod2);
}

- (void)ex_objectAtIndex:(NSInteger)index {
    if (index > self.count - 1) {
        NSLog(@"objectAtIndex————————数组越界");
        return;
    }
    return [self ex_objectAtIndex:index];
}

- (void)ex_objectAtIndexedSubscript:(NSInteger)index {
    if (index > self.count - 1) {
        NSLog(@"取下标————————数组越界");
        return;
    }
    return [self ex_objectAtIndexedSubscript:index];
}

@end 

然而还是越界崩溃了

其实在iOS中NSNumber、NSArray、NSDictionary等这些类都是类簇(Class Clusters)，一个NSArray的实现可能由多个类组成。所以如果想对NSArray进行方法交换，必须获取到其真身进行方法交换，直接对NSArray进行操作是无效的

类名	真身
NSArray	__NSArrayI
NSMutableArray	__NSArrayM
NSDictionary	__NSDictionaryI
NSMutableDictionary	__NSDictionaryM

这样就好办了，可以使用runtime取出本类

#import "NSArray+EX.h"
#import <objc/runtime.h>

@implementation NSArray (EX)

+ (void)load {

    Class cls = objc_getClass(@"__NSArrayI")
    // 交换objectAtIndex方法
    Method oriMethod1 = class_getInstanceMethod(cls, @selector(objectAtIndex:));
    Method swiMethod1 = class_getInstanceMethod(cls, @selector(ex_objectAtIndex:));
    method_exchangeImplementations(oriMethod1, swiMethod1);
    
    // 交换取下标方法
    Method oriMethod2 = class_getInstanceMethod(cls, @selector(objectAtIndexedSubscript:));
    Method swiMethod2 = class_getInstanceMethod(cls, @selector(ex_objectAtIndexedSubscript:));
    method_exchangeImplementations(oriMethod2, swiMethod2);
}

- (void)ex_objectAtIndex:(NSInteger)index {
    if (index > self.count - 1) {
        NSLog(@"objectAtIndex————————数组越界");
        return;
    }
    return [self ex_objectAtIndex:index];
}

- (void)ex_objectAtIndexedSubscript:(NSInteger)index {
    if (index > self.count - 1) {
        NSLog(@"取下标————————数组越界");
        return;
    }
    return [self ex_objectAtIndexedSubscript:index];
}

@end 

坑点二，最好写成单例，避免多次交换

比如说添加了[NSArray load]代码，方法实现又交换回去了导致了崩溃

NSArray *array = @[@"b", @"o", @"o", @"k", @"s"];
[NSArray load];
NSLog(@"%@", array[5]); 

将+load方法改写成单例

+ (void)load {
    static dispatch_once_t onceToken;
    dispatch_once(&onceToken, ^{
        // 交换objectAtIndex方法
        Method oriMethod1 = class_getInstanceMethod(objc_getClass("__NSArrayI"), @selector(objectAtIndex:));
        Method swiMethod1 = class_getInstanceMethod(objc_getClass("__NSArrayI"), @selector(fx_objectAtIndex:));
        method_exchangeImplementations(oriMethod1, swiMethod1);
        
        // 交换取下标方法
        Method oriMethod2 = class_getInstanceMethod(objc_getClass("__NSArrayI"), @selector(objectAtIndexedSubscript:));
        Method swiMethod2 = class_getInstanceMethod(objc_getClass("__NSArrayI"), @selector(fx_objectAtIndexedSubscript:));
        method_exchangeImplementations(oriMethod2, swiMethod2);
    });
} 

坑点3， 子类没有实现 - 父类实现

Student 继承 Person, Person 继承 NSObject,子类没有实现父类方法personInstanceMethod

Student *s = [[Student alloc] init];
[s personInstanceMethod];

Person *p = [[Person alloc] init];
[p personInstanceMethod];

子类打印出结果，而父类调用却崩溃了，为什么会这样呢？

因为Student交换方法personInstanceMethod会先在本类查找方法，再在父类里查找，在父类Person中找到了方法实现就把它跟新方法进行交换了。可是新方法是在Student分类中的，Person找不到imp就unrecognized selector sent to instance 0x600002334250

所以这种情况下应该给子类添加方法实现，只交换子类的方法，不动父类的方法

+ (void)load {
    static dispatch_once_t onceToken;
    dispatch_once(&onceToken, ^{
    
        // oriSEL       personInstanceMethod
        // swizzledSEL  ex_personInstanceMethod
        Method oriMethod = class_getInstanceMethod(self, @selector(personInstanceMethod));
        Method swiMethod = class_getInstanceMethod(self, @selector(ex_personInstanceMethod));
        //给原方法添加方法实现，如果之前没实现就能添加成功，如果之前实现了，就会添加失败
        BOOL didAddMethod = class_addMethod(self, @selector(personInstanceMethod), method_getImplementation(swiMethod), method_getTypeEncoding(swiMethod));
        if (didAddMethod) {
        ///如果添加成功 则替换ex_的方法实现,这样也达到了方法交换的效果
        /**
        添加后：personInstanceMethod(sel) - ex_personInstanceMethod(imp)
        替换后：ex_personInstanceMethod (swizzledSEL) - personInstanceMethod(imp，这里因为父类有实现，所以调用的是父类的实现)
        */
            class_replaceMethod(self, @selector(ex_personInstanceMethod), method_getImplementation(oriMethod), method_getTypeEncoding(oriMethod));
        } else {
            method_exchangeImplementations(oriMethod, swiMethod);
        }
    }); 
} 

- (void)ex_personInstanceMethod{
    [self ex_personInstanceMethod];
    NSLog(@"Student分类添加的ex对象方法:%s",__func__);
}

这样就没有动父类的方法，相当于子类重写了父类的方法personInstanceMethod

坑点4， 子类没有实现 - 父类也没有实现 会造成死循环

因为都没有实现，导致class_replaceMethod时，获取原方法的实现personInstanceMethod(imp)无效，ex_personInstanceMethod方法中想调用旧的方法实现，可是找不到personInstanceMethod实现,就会死循环

所以我们判断一下，如果原方法没实现，则添加空实现

+ (void)load {
    static dispatch_once_t onceToken;
    dispatch_once(&onceToken, ^{
        Method oriMethod = class_getInstanceMethod(self, @selector(doInstance));
        Method swiMethod = class_getInstanceMethod(self, @selector(fx_doInstance));
        /// 原方法没实现，添加一个空的实现
        if (!oriMethod) {
            class_addMethod(self, @selector(doInstance), method_getImplementation(swiMethod), method_getTypeEncoding(swiMethod));
            method_setImplementation(swiMethod, imp_implementationWithBlock(^(id self, SEL _cmd) {
                NSLog(@"方法未实现");
            }));
        }
        
        BOOL didAddMethod = class_addMethod(self, @selector(doInstance), method_getImplementation(swiMethod), method_getTypeEncoding(swiMethod));
        if (didAddMethod) {
            class_replaceMethod(self, @selector(fx_doInstance), method_getImplementation(oriMethod), method_getTypeEncoding(oriMethod));
        } else {
            method_exchangeImplementations(oriMethod, swiMethod);
        }
    });
} 

• 尽可能在+load方法中交换方法 使用单例保证只交换一次
• 自定义方法名不能产生冲突
• 对于系统方法要调用原始实现，避免对系统产生影响
• 做好注释（因为方法交换比较绕）
• 迫不得已情况下才去使用方法交换

+ (void)DHMethodSwizzlingWithClass:(Class)cls oriSEL:(SEL)oriSEL swizzledSEL:(SEL)swizzledSEL{
    
    if (!cls) NSLog(@"传入的交换类不能为空");
    
    Method oriMethod = class_getInstanceMethod(cls, oriSEL);
    Method swiMethod = class_getInstanceMethod(cls, swizzledSEL);
    
    if (!oriMethod) {
        // 在oriMethod为nil时，替换后将swizzledSEL复制一个不做任何事的空实现,代码如下:
        class_addMethod(cls, oriSEL, method_getImplementation(swiMethod), method_getTypeEncoding(swiMethod));
        method_setImplementation(swiMethod, imp_implementationWithBlock(^(id self, SEL _cmd){ }));
    } 
    
    BOOL didAddMethod = class_addMethod(cls, oriSEL, method_getImplementation(swiMethod), method_getTypeEncoding(swiMethod));
    if (didAddMethod) {
        class_replaceMethod(cls, swizzledSEL, method_getImplementation(oriMethod), method_getTypeEncoding(oriMethod));
    }else{
        method_exchangeImplementations(oriMethod, swiMethod);
    }
    
}

九、 内存偏移问题

程序能否运行吗？是否正常输出？

#import "ViewController.h"

@interface Person : NSObject
@property (nonatomic, copy) NSString *name;
@end

@implementation Person

- (void)printMyProperty {
    NSLog(@"当前打印内容为%s", __func__);
}

@end


@implementation ViewController

- (void)viewDidLoad {
    [super viewDidLoad];
    
    id cls = [Person class];
    void *obj= &cls;
    [(__bridge id)obj printMyProperty];
    
    Person *p = [Person alloc];
    [p printMyProperty];
}

@end 

运行结果与普通初始化对象一模一样

正常操作：指针p->实例对象的首地址isa->类对象

指针操作：指针obj->指针cls->类对象

isa不是可以强转位cls吗，所以都能正常打印

拓展一

修改打印方法printMyProperty——不但打印方法，同时打印属性name

- (void)printMyProperty {
    NSLog(@"当前打印内容为%s——————%@", __func__, self.name);
}

当前打印内容为 -[Person printMyProperty]——————<ViewController: 0x10472bd50>

为什么属性name还没有赋值，却打印出了ViewController的内存地址？

• 由于栈先入后出，viewDidLoad入栈先拉伸栈空间，然后依次放入self、_cmd局部变量
• 调用[super viewDidLoad]，继续放入super_class、self
• 正常情况下获取name，本质是p的内存地址往下偏移8字节

拓展二

修改viewDidLoad——在obj前面加个临时字符串变量

- (void)viewDidLoad {
    [super viewDidLoad];
    
    NSString *temp = @"dh";
    
    id cls = [Person class];
    void *obj= &cls;
    [(__bridge id)obj printMyProperty];
    
    Person *p = [Person alloc];
    [p printMyProperty];
}

打印如下：

当前打印内容为 -[Person printMyProperty]——————dh

同理，在obj入栈前已经有了temp变量，此时访问self.name就会访问到temp

拓展三

去掉临时变量，Person类新增字符串属性hobby，打印方法改为打印hobby，运行

#import "ViewController.h"

@interface Person : NSObject
@property (nonatomic, copy) NSString *name;
@property (nonatomic, copy) NSString *hobby;
@end

@implementation Person

- (void)printMyProperty {
    NSLog(@"当前打印内容为%s——————%@", __func__, self.hobby);
}

@end


@implementation ViewController

- (void)viewDidLoad {
    [super viewDidLoad];
    
    id cls = [Person class];
    void *obj= &cls;
    [(__bridge id)obj printMyProperty];
    
    Person *p = [Person alloc];
    [p printMyProperty];
}

@end 

当前打印内容为 -[Person printMyProperty]——————ViewController ViewController就是obj偏移16字节拿到的super_class,见上图

拓展四

去掉[super viewDidLoad]

#import "ViewController.h"

@interface Person : NSObject
@property (nonatomic, copy) NSString *name;
@property (nonatomic, copy) NSString *hobby;
@end

@implementation Person

- (void)printMyProperty {
    NSLog(@"当前打印内容为%s——————%@", __func__, self.name);
}

@end


@implementation ViewController

- (void)viewDidLoad {
    //[super viewDidLoad];
    
    id cls = [Person class];
    void *obj= &cls;
    [(__bridge id)obj printMyProperty];
    
    Person *p = [Person alloc];
    [p printMyProperty];
}

@end 

崩溃保存Thread 1: EXC_BAD_ACCESS (code=1, address=0x477656990)

野指针——指针偏移的offset不正确，获取不到对应变量的首地址

打印改一下

- (void)printMyProperty {
    NSLog(@"当前打印内容为%s——————%@", __func__, self.hobby);
}

当前打印内容为-[LGPerson saySomething]——————<ViewController: 0x11fd365c0>

变量的存放顺序，是根据定义的先后顺序，从函数栈底（高地址）开始，一个一个排列