1. 程式人生 > >leveldb原始碼分析4:SkipList

leveldb原始碼分析4:SkipList

skiplist思想可以具體參考這:

  // Create a new SkipList object that will use "cmp" for comparing keys,
  // and will allocate memory using "*arena".  Objects allocated in the arena
  // must remain allocated for the lifetime of the skiplist object.
  explicit SkipList(Comparator cmp, Arena* arena);

  // Insert key into the list.
  // REQUIRES: nothing that compares equal to key is currently in the list.
  void Insert(const Key& key);

  // Returns true iff an entry that compares equal to key is in the list.
  bool Contains(const Key& key) const;

private成員變數:
// 最大的level
  enum { kMaxHeight = 12 };

  // Immutable after construction
  Comparator const compare_;
  // 記憶體分配器
  Arena* const arena_;    // Arena used for allocations of nodes

    // 指向第一個節點,建構函式中初始化
  Node* const head_;

  // Modified only by Insert().  Read racily by readers, but stale
  // values are ok.
  port::AtomicPointer max_height_;   // Height of the entire list
我們下面來首先分析初始化操作,如下:
// 初始化:
// 1. 初始化compare_
// 2. 初始化arena_
// 3. 初始化head_,指向指標陣列
// 4. 初始化max_height_
// 5. 初始化rnd_隨機數的seed
// 6. 初始化head_指向的陣列
template<typename Key, class Comparator>
SkipList<Key,Comparator>::SkipList(Comparator cmp, Arena* arena)
    : compare_(cmp),
      arena_(arena),
      head_(NewNode(0 /* any key will do */, kMaxHeight)),
      max_height_(reinterpret_cast<void*>(1)),
      rnd_(0xdeadbeef) {
          // 初始化head_指向的陣列
  for (int i = 0; i < kMaxHeight; i++) {
    head_->SetNext(i, NULL);
  }
}

下面是一個插入操作的示意圖:


leveldb中實現的插入程式碼就是按照上面的思路實現,首先查詢到合適的位置,並記錄查詢過程中經過的路徑,之後新生成一個節點,修改指標。

// 插入操作
// 這裡的key其實已經是經過處理的key,包含了使用者指定的key和value
template<typename Key, class Comparator>
void SkipList<Key,Comparator>::Insert(const Key& key) {
  // TODO(opt): We can use a barrier-free variant of FindGreaterOrEqual()
  // here since Insert() is externally synchronized.
  // prev記錄的是查詢路徑,下面需要使用prev來修改新生成
  // 節點的指標
  Node* prev[kMaxHeight];
  Node* x = FindGreaterOrEqual(key, prev);

  // Our data structure does not allow duplicate insertion
  // 不允許插入重複的值
  assert(x == NULL || !Equal(key, x->key));

    // 隨即生成節點高度
  int height = RandomHeight();
  // 對prev陣列中未賦值的元素進行賦值
  if (height > GetMaxHeight()) {
    for (int i = GetMaxHeight(); i < height; i++) {
      prev[i] = head_;
    }


    // It is ok to mutate max_height_ without any synchronization
    // with concurrent readers.  A concurrent reader that observes
    // the new value of max_height_ will see either the old value of
    // new level pointers from head_ (NULL), or a new value set in
    // the loop below.  In the former case the reader will
    // immediately drop to the next level since NULL sorts after all
    // keys.  In the latter case the reader will use the new node.
    // 設定max_height變數
    max_height_.NoBarrier_Store(reinterpret_cast<void*>(height));
  }

    // 新生成一個節點,之後插入資料
  x = NewNode(key, height);
  for (int i = 0; i < height; i++) {
    // NoBarrier_SetNext() suffices since we will add a barrier when
    // we publish a pointer to "x" in prev[i].
    // 修改兩部分的指標,一部分是需要執行新插入節點的指標
    // 另外的一部分是x節點的指標
    x->NoBarrier_SetNext(i, prev[i]->NoBarrier_Next(i));
    prev[i]->SetNext(i, x);
  }
}

函式FindGreaterOrEqual中完成查詢操作,就是向下(level控制)和向右(x控制)移動過程,並不斷經經過路徑儲存到引數prev中。

template<typename Key, class Comparator>
typename SkipList<Key,Comparator>::Node* 
SkipList<Key,Comparator>::FindGreaterOrEqual(const Key& key,
                                                                                      Node** prev)
    const {
        // 從最高層開始查詢
  Node* x = head_;
  int level = GetMaxHeight() - 1;
  while (true) {
    Node* next = x->Next(level);
    if (KeyIsAfterNode(key, next)) {    // 向右移動
      // Keep searching in this list
      x = next;
    }
    else    // 向下移動
    {
        // 記錄查詢路徑
      if (prev != NULL)
        prev[level] = x;
      if (level == 0) {
        return next;
      } else {
        // Switch to next list下一層尋找
        level--;
      }
    }
  }
}

查詢操作基本上就是呼叫函式上面的函式FindGreaterOrEqual實現:

// 查詢操作
template<typename Key, class Comparator>
bool SkipList<Key,Comparator>::Contains(const Key& key) const {
  Node* x = FindGreaterOrEqual(key, NULL);
  if (x != NULL && Equal(key, x->key)) {
    return true;
  } else {
    return false;
  }
}

上面基本上就是skiplist在leveldb中實現,leveldb中沒有使用複雜的紅黑樹等機制去保證資料的有序性,而是使用了輕快的skiplist實現。最後需要注意skiplist中每個節點儲存key是使用者傳遞keyvalue經過變幻(變幻方法參考http://blog.csdn.net/xuqianghit/article/details/6948164)得到的。