从纯算法上来讲, 这是一个寻找最长子序列的变体...
标准答案就是用动态规划算法.
但是从实际上来讲, 试卷答案不只是检查下相似程度, 而是有"关键分数点", 只要"关键分数点"正确, 其他其实并不重要.
你需要设定一些关键分数点, 然后再考虑相似程度.
private unsafe static int GetSimilarPercent(string str1, string str2)
{
//check if they are equal
if (String.Compare(str1, str2, true) == 0)
{
return 100;
}
else if (String.IsNullOrEmpty(str1) || String.IsNullOrEmpty(str2))
{
//one string is null, return 0
return 0;
}
else
{
//both two strings are not empty, then we can start to check the similar percent.
str1 = str1.Trim();
str2 = str2.Trim();
//if the two strings is a sentence(contains blank) not a single word, then split the sentence to words, check each word.
if (str1.IndexOf(" ") > 0 && str2.IndexOf(" ") > 0)
{
string[] str1Arr = str1.Split(new char[] { ' ' }, StringSplitOptions.RemoveEmptyEntries);
string[] str2Arr = str2.Split(new char[] { ' ' }, StringSplitOptions.RemoveEmptyEntries);
//if the two strings have the same number of words, check each word.
if (str1Arr.Length > 1 && str1Arr.Length == str2Arr.Length)
{
int totalSimPercent = 0;
float weighted = 0;
for (int i = 0; i < str1Arr.Length; i++)
{
weighted = (float)(str1Arr[i].Length + str2Arr[i].Length) / (float)(str1.Length + str2.Length);
totalSimPercent += Convert.ToInt32(GetSimilarPercent(str1Arr[i], str2Arr[i]) * weighted);
}
return totalSimPercent;
}
}
int len1 = str1.Length;
int len2 = str2.Length;
//dynamic programming array.
//to improve performance, we use 1 dim stack arrary and unsafe code.
int* dpArr = stackalloc int[(len1 + 1) * (len2 + 1)];
//init array
int curIndex = 0;
for (int i = 0; i < len1; i++)
{
for (int j = 0; j < len2; j++)
{
curIndex = (j + 1) * (len1 + 1) + i + 1;
if (str1[i] == str2[j])
{
if (i == 0 || j == 0)
{
dpArr[curIndex] = 1;
}
else
{
dpArr[curIndex] = dpArr[j * (len1 + 1) + i] + 1;
}
}
else
{
dpArr[curIndex] = 0;
}
}
}
int currentSameCharCount = 0;
int totalSameCharCount = 0;
int totalLen = len1 + len2;
int str1Index = len1;
int str2Index = len2;
//the max number's position in the array of sepcific line and row.
int maxStr1Index = str1Index;
int maxStr2Index = str2Index;
while (str1Index > 0 && str2Index > 0)
{
currentSameCharCount = 0;
for (int i = str1Index; i > 0; i--)
{
curIndex = str2Index * (len1 + 1) + i;
if (dpArr[curIndex] > currentSameCharCount)
{
currentSameCharCount = dpArr[curIndex];
maxStr1Index = i;
maxStr2Index = str2Index;
}
if (currentSameCharCount >= i)
{
break;
}
}
for (int j = str2Index; j > 0; j--)
{
curIndex = j * (len1 + 1) + str1Index;
if (dpArr[curIndex] > currentSameCharCount)
{
currentSameCharCount = dpArr[curIndex];
maxStr1Index = str1Index;
maxStr2Index = j;
}
if (currentSameCharCount >= j)
{
break;
}
}
totalSameCharCount += currentSameCharCount;
if (currentSameCharCount > 0)
{
str1Index = maxStr1Index - currentSameCharCount;
str2Index = maxStr2Index - currentSameCharCount;
}
else
{
str1Index--;
str2Index--;
}
}
float percent = (float)(totalSameCharCount * 2) / (float)totalLen;
return Convert.ToInt32(percent * 100);
}
}
