【深入了解cocos2d-x 3.x】定时器(scheduler)的使用和原理探究(2)

上篇说到定时器的用法。这篇主要分析它的实现原理。

1.哈希链表

cocos2dx封装了一个结构体,叫做UT_hash_handle。仅仅要在自己定义的结构体中声明这个结构体变量。就实现了哈希链表,而且能使用一系列的哈希链表专用的宏。这个结构体的详细实现例如以下:
typedef struct UT_hash_handle {
   struct UT_hash_table *tbl;
   void *prev;                       /* prev element in app order      */
   void *next;                       /* next element in app order      */
   struct UT_hash_handle *hh_prev;   /* previous hh in bucket order    */
   struct UT_hash_handle *hh_next;   /* next hh in bucket order        */
   void *key;                        /* ptr to enclosing struct's key  */
   unsigned keylen;                  /* enclosing struct's key len     */
   unsigned hashv;                   /* result of hash-fcn(key)        */
} UT_hash_handle;

这个结构体主要实现的是一个双向链表,详细实现哈希验证的还要看UT_hash_table 结构体
typedef struct UT_hash_table {
   UT_hash_bucket *buckets;
   unsigned num_buckets, log2_num_buckets;
   unsigned num_items;
   struct UT_hash_handle *tail; /* tail hh in app order, for fast append    */
   ptrdiff_t hho; /* hash handle offset (byte pos of hash handle in element */

   /* in an ideal situation (all buckets used equally), no bucket would have
    * more than ceil(#items/#buckets) items. that's the ideal chain length. */
   unsigned ideal_chain_maxlen;

   /* nonideal_items is the number of items in the hash whose chain position
    * exceeds the ideal chain maxlen. these items pay the penalty for an uneven
    * hash distribution; reaching them in a chain traversal takes >ideal steps */
   unsigned nonideal_items;

   /* ineffective expands occur when a bucket doubling was performed, but 
    * afterward, more than half the items in the hash had nonideal chain
    * positions. If this happens on two consecutive expansions we inhibit any
    * further expansion, as it's not helping; this happens when the hash
    * function isn't a good fit for the key domain. When expansion is inhibited
    * the hash will still work, albeit no longer in constant time. */
   unsigned ineff_expands, noexpand;

   uint32_t signature; /* used only to find hash tables in external analysis */
#ifdef HASH_BLOOM
   uint32_t bloom_sig; /* used only to test bloom exists in external analysis */
   uint8_t *bloom_bv;
   char bloom_nbits;
#endif

} UT_hash_table;

然后看看与哈希链表相关的宏定义,使用这些宏能非常方便的插入链表。删除链表,查找链表。
/**
 * 查找元素
 * head:哈希链表的头指针
 * findptr:要查找的元素指针
 * out:查找结果
 */
HASH_FIND_PTR(head,findptr,out) 
/**
 * 加入元素
 * head:哈希链表的头指针
 * ptrfield:要加入的元素指针
 * add:要加入的哈希链表元素
 */
HASH_ADD_PTR(head,ptrfield,add) 
/**
 * 替换元素
 * head:哈希链表的头指针
 * ptrfield:要替换的元素指针
 * add:要替换的哈希链表元素
 */
HASH_REPLACE_PTR(head,ptrfield,add)
/**
 * 删除
 * head:哈希链表的头指针
 * delptr:要删除的元素指针
 */
HASH_DEL(head,delptr) 

以上是引擎中实现的哈希链表的相关知识,接下来再看看与定时器相关的哈希链表。定时器的实现中。将一个定时器存储在哈希链表中,那么在scheduler是怎样实现以后哈希链表的结构体的呢?例如以下:
// 不同优先级的update定时器的双向链表
typedef struct _listEntry
{
    struct _listEntry   *prev, *next;
    ccSchedulerFunc     callback;
    void                *target;
    int                 priority;
    bool                paused;
    bool                markedForDeletion; // selector will no longer be called and entry will be removed at end of the next tick
} tListEntry;
//内置的update定时器
typedef struct _hashUpdateEntry
{
    tListEntry          **list;        // Which list does it belong to ?
    tListEntry          *entry;        // entry in the list
    void                *target;
    ccSchedulerFunc     callback;
    UT_hash_handle      hh;
} tHashUpdateEntry;

// 自己定义定时器
typedef struct _hashSelectorEntry
{
    ccArray             *timers;
    void                *target;
    int                 timerIndex;
    Timer               *currentTimer;
    bool                currentTimerSalvaged;
    bool                paused;
    UT_hash_handle      hh;
} tHashTimerEntry;

以上就是相关的哈希链表的知识。接下来从定义定时器的函数Node::schedule中一步一步的分析定时器是怎样增加到哈希链表中的。

2.怎样定义自己定义定时器

首先。上一篇文章中说到了非常多个自己定义定时器的函数。可是终于会调用的函数仅仅有两个,各自是
    /**
     * 定义一个自己定义的定时器
	 * selector:回调函数
	 * interval:反复间隔时间。反复运行间隔的时间。假设传入0,则表示每帧调用
	 * repeat:反复运行次数,假设传入CC_REPEAT_FOREVER则表示无限循环
	 * delay:延时秒数。延迟delay秒開始运行第一次回调
     */
    void schedule(SEL_SCHEDULE selector, float interval, unsigned int repeat, float delay);
	
    /**
     * 使用lambda函数定义一个自己定义定时器
     * callback:lambda函数
	 * interval:反复间隔时间。反复运行间隔的时间,假设传入0,则表示每帧调用
	 * repeat:反复运行次数。假设传入CC_REPEAT_FOREVER则表示无限循环
	 * delay:延时秒数。延迟delay秒開始运行第一次回调
     * key:lambda函数的Key,用于取消定时器
     * @lua NA
     */
    void schedule(const std::function<void(float)>& callback, float interval, unsigned int repeat, float delay, const std::string &key);

本文从传统的定义定时器的方法入手,也就是第一个方法。

接下来看看这种方法的实现:

void Node::schedule(SEL_SCHEDULE selector, float interval, unsigned int repeat, float delay)
{
    CCASSERT( selector, "Argument must be non-nil");
    CCASSERT( interval >=0, "Argument must be positive");

    _scheduler->schedule(selector, this, interval , repeat, delay, !_running);
}

看到事实上还是调用_scheduler的schedule方法。那么_scheduler又是个什么鬼?
Scheduler *_scheduler;          ///< scheduler used to schedule timers and updates
查看定义能够知道是一个Scheduler 的指针。可是这个指针从哪里来?在构造函数中有真相
Node::Node(void)
{
    // set default scheduler and actionManager
    _director = Director::getInstance();
    _scheduler = _director->getScheduler();
    _scheduler->retain();
}

是从导演类中引用的。这一块临时我们无论。接下来深入到_scheduler->schedule函数中分析。例如以下是函数的详细实现
void Scheduler::schedule(SEL_SCHEDULE selector, Ref *target, float interval, unsigned int repeat, float delay, bool paused)
{
    CCASSERT(target, "Argument target must be non-nullptr");
    
	//定义而且查找链表元素
    tHashTimerEntry *element = nullptr;
    HASH_FIND_PTR(_hashForTimers, &target, element);
    
	//没找到
    if (! element)
    {
		//创建一个链表元素
        element = (tHashTimerEntry *)calloc(sizeof(*element), 1);
        element->target = target;
        
		//加入到哈希链表中
        HASH_ADD_PTR(_hashForTimers, target, element);
        
        // Is this the 1st element ?

Then set the pause level to all the selectors of this target element->paused = paused; } else { CCASSERT(element->paused == paused, ""); } //检查这个元素的定时器数组,假设数组为空 则new 10个数组出来备用 if (element->timers == nullptr) { element->timers = ccArrayNew(10); } else { //循环查找定时器数组,看看是不是以前定义过同样的定时器。假设定义过,则仅仅须要改动定时器的间隔时间 for (int i = 0; i < element->timers->num; ++i) { TimerTargetSelector *timer = dynamic_cast<TimerTargetSelector*>(element->timers->arr[i]); if (timer && selector == timer->getSelector()) { CCLOG("CCScheduler#scheduleSelector. Selector already scheduled. Updating interval from: %.4f to %.4f", timer->getInterval(), interval); timer->setInterval(interval); return; } } //扩展1个定时器数组 ccArrayEnsureExtraCapacity(element->timers, 1); } //创建一个定时器,而且将定时器加入到当前链表指针的定时器数组中 TimerTargetSelector *timer = new (std::nothrow) TimerTargetSelector(); timer->initWithSelector(this, selector, target, interval, repeat, delay); ccArrayAppendObject(element->timers, timer); timer->release(); }



这一段代码详细分析了怎样将自己定义定时器增加到链表中。而且在链表中的存储结构是怎么样的。接下来看看内置的Update定时器。

3.怎样定义Update定时器

Update定时器的开启方法有两个。各自是:
    /**
     * 开启自带的update方法,这种方法会每帧运行一次,默认优先级为0,而且在全部自己定义方法运行之前运行
     */
    void scheduleUpdate(void);

    /**
     * 开启自带的update方法。这种方法会每帧运行一次。设定的优先级越小,越优先运行
     */
    void scheduleUpdateWithPriority(int priority);
第一个方法实际上是直接调用第二个方法。而且把优先级设置为0,我们直接看第二个方法就能够了。

void Node::scheduleUpdateWithPriority(int priority)
{
    _scheduler->scheduleUpdate(this, priority, !_running);
}
详细调用还是要进入到_scheduler->scheduleUpdate。

/** Schedules the 'update' selector for a given target with a given priority.
     The 'update' selector will be called every frame.
     The lower the priority, the earlier it is called.
     @since v3.0
     @lua NA
     */
    template <class T>
    void scheduleUpdate(T *target, int priority, bool paused)
    {
        this->schedulePerFrame([target](float dt){
            target->update(dt);
        }, target, priority, paused);
    }

能够看到这里主要还是调用了一个schedulePerFrame函数,而且传入了一个lambda函数。这个函数实际上调用的是target->update,接下来走进schedulePerFrame看看它的实现:
void Scheduler::schedulePerFrame(const ccSchedulerFunc& callback, void *target, int priority, bool paused)
{
	//定义而且查找链表元素
    tHashUpdateEntry *hashElement = nullptr;
    HASH_FIND_PTR(_hashForUpdates, &target, hashElement);
	
	//假设找到,就直接改优先级
    if (hashElement)
    {
        // 检查优先级是否改变
        if ((*hashElement->list)->priority != priority)
        {
			//检查是否被锁定
            if (_updateHashLocked)
            {
                CCLOG("warning: you CANNOT change update priority in scheduled function");
                hashElement->entry->markedForDeletion = false;
                hashElement->entry->paused = paused;
                return;
            }
            else
            {
            	// 在这里先停止到update。后面会加回来 
                unscheduleUpdate(target);
            }
        }
        else
        {
            hashElement->entry->markedForDeletion = false;
            hashElement->entry->paused = paused;
            return;
        }
    }

    // 优先级为0,增加到_updates0List链表中,而且增加到_hashForUpdates表中
    if (priority == 0)
    {
        appendIn(&_updates0List, callback, target, paused);
    }
	// 优先级小于0,增加到_updatesNegList链表中。而且增加到_hashForUpdates表中
    else if (priority < 0)
    {
        priorityIn(&_updatesNegList, callback, target, priority, paused);
    }
	// 优先级大于0,增加到_updatesPosList链表中,而且增加到_hashForUpdates表中
    else
    {
        // priority > 0
        priorityIn(&_updatesPosList, callback, target, priority, paused);
    }
}
在这里看上去逻辑还是非常清晰的,有两个函数要重点分析一下。各自是
void Scheduler::appendIn(_listEntry **list, const ccSchedulerFunc& callback, void *target, bool paused)
void Scheduler::priorityIn(tListEntry **list, const ccSchedulerFunc& callback, void *target, int priority, bool paused)

第一个用于加入默认优先级。第二个函数用于加入指定优先级的。首先看加入默认优先级的。
void Scheduler::appendIn(_listEntry **list, const ccSchedulerFunc& callback, void *target, bool paused)
{
	//创建一个链表元素
    tListEntry *listElement = new tListEntry();

    listElement->callback = callback;
    listElement->target = target;
    listElement->paused = paused;
    listElement->priority = 0;
    listElement->markedForDeletion = false;
	
	//加入到双向链表中
    DL_APPEND(*list, listElement);

    //创建一个哈希链表元素
    tHashUpdateEntry *hashElement = (tHashUpdateEntry *)calloc(sizeof(*hashElement), 1);
    hashElement->target = target;
    hashElement->list = list;
    hashElement->entry = listElement;
	//加入到哈希链表中
    HASH_ADD_PTR(_hashForUpdates, target, hashElement);
}

接下来看还有一个函数
void Scheduler::priorityIn(tListEntry **list, const ccSchedulerFunc& callback, void *target, int priority, bool paused)
{
	//同上一个函数
    tListEntry *listElement = new tListEntry();

    listElement->callback = callback;
    listElement->target = target;
    listElement->priority = priority;
    listElement->paused = paused;
    listElement->next = listElement->prev = nullptr;
    listElement->markedForDeletion = false;

    //假设链表为空
    if (! *list)
    {
        DL_APPEND(*list, listElement);
    }
    else
    {
        bool added = false;
		//保证链表有序
        for (tListEntry *element = *list; element; element = element->next)
        {
			// 假设优先级小于当前元素的优先级,就在这个元素前面插入
            if (priority < element->priority)
            {
                if (element == *list)
                {
                    DL_PREPEND(*list, listElement);
                }
                else
                {
                    listElement->next = element;
                    listElement->prev = element->prev;

                    element->prev->next = listElement;
                    element->prev = listElement;
                }

                added = true;
                break;
            }
        }

        //假设新增加的优先级最低,则增加到链表的最后
        if (! added)
        {
            DL_APPEND(*list, listElement);
        }
    }

    //同上一个函数
    tHashUpdateEntry *hashElement = (tHashUpdateEntry *)calloc(sizeof(*hashElement), 1);
    hashElement->target = target;
    hashElement->list = list;
    hashElement->entry = listElement;
    HASH_ADD_PTR(_hashForUpdates, target, hashElement);
}

本文简单的分析了哈希链表以及定时器的存储和加入,下一篇文章将分析定时器是怎样运转起来的。


posted @ 2017-07-18 15:56  cxchanpin  阅读(298)  评论(0编辑  收藏  举报