After treatment in a number of initialization, MD5 to 512 division to deal with...

代码说明：

在一些初始化处理后，MD5以512位分组来处理输入文本，每一分组又划分为16个32位子分组。算法的输出由四个32位分组组成，将它们级联形成一个128位散列值。 首先填充消息使其长度恰好为一个比512位的倍数仅小64位的数。填充方法是附一个1在消息后面，后接所要求的多个0，然后在其后附上64位的消息长度（填充前）。这两步的作用是使消息长度恰好是512位的整数倍（算法的其余部分要求如此），同时确保不同的消息在填充后不相同。 四个32位变量初始化为：-After treatment in a number of initialization, MD5 to 512 division to deal with input text, each sub-divided into 16 sub-32 seat. Algorithm output consists of four sub-groups, 32, cascade them to form a 128-bit hash value. First of all, it filled the news for a length of just over 512 in multiples of 64 only a few small. Filling method is attached to the back of a 1 in the message, followed by the required number of 0, and then subsequently attached to the message length of 64 (before filling). The role of these two steps so that messages are exactly the length of integer multiple of 512 (the rest of the algorithm requested the case), while ensuring that the different sources in the same after filling. Four 32-bit variable is initialized as follows: