private int GetCommonPrefix(String str1, String str2)
{
if (str1 == null || str2 == null)
{
return(this.prefixLength);
}
int commonPrefix = 0;
int length = MathExtension.Min(str1.Length,
str2.Length,
this.prefixLength);
// check for prefix similarity of length n
for (int i = 0; i < length; i++)
{
// check the prefix is the same so far
if (str1[i] == str2[i])
{
// count up
commonPrefix++;
}
}
return(commonPrefix);
}
public int BuildDamerauLevenshteinDistance(string str1, string str2)
{
if (str1 == null || str2 == null)
{
return(0);
}
int length1 = str1.Length;
int length2 = str2.Length;
if (length1 == 0)
{
return(length2);
}
if (length2 == 0)
{
return(length1);
}
int[,] matrix = new int[str1.Length + 1, str2.Length + 1];
// initialize matrix
for (int i = 0; i <= length1; i++)
{
matrix[i, 0] = i;
}
for (int j = 0; j <= length2; j++)
{
matrix[0, j] = j;
}
// analyze
for (int i = 1; i <= length1; i++)
{
for (int j = 1; j <= length2; j++)
{
int deletionCost = 1;
int insertionCost = 1;
// substitution is based on a cost function for different chars. e.g. m->n has lower cost than a->X
int substitionCost = (int)this.CostFunction.GetCost(str1, i - 1, str2, j - 1);
int del = matrix[i - 1, j]; // above is deletion
int ins = matrix[i, j - 1]; // left is insertion
int subst = matrix[i - 1, j - 1]; // diagonal is substitution
// new cell is the minium of all three precessors (left, diagonal, above).
// So the minumium moves to this cell plus the cost of the operation
int cell = MathExtension.Min(del + deletionCost,
ins + insertionCost,
subst + substitionCost);
// transposition e.g. "ab" <-> "ba" (Damerau-extension to Levensthein)
if (i > 1 && j > 1)
{
int transposition = matrix[i - 2, j - 2] + 1;
if (str1[i - 2] != str2[j - 1])
{
transposition++;
}
if (str1[i - 1] != str2[j - 2])
{
transposition++;
}
//trans += (int)this.CostFunction.GetCost(str1, i - 2, str2, j - 1);
//trans += (int)this.CostFunction.GetCost(str1, i - 1, str2, j - 2);
cell = MathExtension.Min(cell,
transposition);
}
// put the new cellvalue back in the matrix
matrix[i, j] = cell;
}
}
// return the last cell (right, bottom) of the matrix
return(matrix[length1, length2]);
}