Re: [問題] 利用指標完成quicksort

[dendrobium (石斛蘭)]

演算法本身就是錯的吧...

考慮 input[] = {3,1,2,3,4,5};


  front                last
    |                   |
  +---+---+---+---+---+---+
  | 3 | 1 | 2 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3












因為 *last >= meddle 所以 last--

  front            last
    |               |
  +---+---+---+---+---+---+
  | 3 | 1 | 2 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3













again, *last >= meddle 所以 last--

  front        last
    |           |
  +---+---+---+---+---+---+
  | 3 | 1 | 2 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3













有等號的關係...again

  front    last
    |       |
  +---+---+---+---+---+---+
  | 3 | 1 | 2 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3













跳出while(), swap(*front,*last);

  front    last
    |       |
  +---+---+---+---+---+---+
  | 2 | 1 | 3 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3













接著 *front <= meddle 所以 front++

      front last
        |   |
  +---+---+---+---+---+---+
  | 2 | 1 | 3 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3













直到

           last   front
            |       |
  +---+---+---+---+---+---+
  | 2 | 1 | 3 | 3 | 4 | 5 |
  +---+---+---+---+---+---+

   meddle:3













然後 swap(*last,*front)

           last   front
            |       |
  +---+---+---+---+---+---+
  | 2 | 1 | 4 | 3 | 3 | 5 |
  +---+---+---+---+---+---+

   meddle:3

你會發現, 這個演算法根本就不正確...￣▽￣||



--
人家可不是為了你才這樣做的哦!

--
※ 發信站: 批踢踢實業坊(ptt.cc) 
◆ From: 60.198.35.85
推 nowar100:喔喔  有圖有推 XD
→ dendrobium:都是複製貼上啦XD
推 VictorTom:太熱心了, 推:)
推 mmmaaatttttt:推圖是一定要的啦XD
推 courant:有圖有推
推 loveme00835:有圖有推