unsigned int levenshtein_distance(const T &s1, const T & s2) {
const size_t len1 = s1.length(), len2 = s2.length();
std::vector<unsigned int> col(len2+1), prevCol(len2+1);
for (unsigned int i = 0; i < prevCol.size(); i++)
prevCol[i] = i;
for (unsigned int i = 0; i < len1; i++) {
col[0] = i+1;
for (unsigned int j = 0; j < len2; j++)
col[j+1] = std::min( std::min(prevCol[1 + j] + 1, col[j] + 1),
prevCol[j] + (s1[i]==s2[j] ? 0 : 1) );
col.swap(prevCol);
}
return prevCol[len2];
}
// https://en.wikibooks.org/wiki/Algorithm_Implementation/Strings/Levenshtein_distance#C.2B.2B
unsigned int computeEditDistance(const string& s1, const string& s2)
{
const size_t len1 = s1.size(), len2 = s2.size();
vector<unsigned int> col(len2 + 1), prevCol(len2 + 1);
for (unsigned int i = 0; i < prevCol.size(); i++)
prevCol[i] = i;
for (unsigned int i = 0; i < len1; i++)
{
col[0] = i + 1;
for (unsigned int j = 0; j < len2; j++)
col[j + 1] = min(min(prevCol[1 + j] + 1, col[j] + 1), prevCol[j] + (s1[i] == s2[j] ? 0 : 1)) ;
col.swap(prevCol);
}
return prevCol[len2];
}
unsigned int
GestureRecognizer::levenshteinDistance(const std::vector<int>& a, const std::vector<int>& b)
{
const size_t len1 = a.size(), len2 = b.size();
std::vector<unsigned int> col(len2 + 1), prevCol(len2 + 1);
for (unsigned int i = 0; i < prevCol.size(); i++)
prevCol[i] = i;
for (unsigned int i = 0; i < len1; i++)
{
col[0] = i + 1;
for (unsigned int j = 0; j < len2; j++)
col[j + 1] = std::min(std::min(1 + col[j], 1 + prevCol[1 + j]), prevCol[j] + (a[i] == b[j] ? 0 : 1));
col.swap(prevCol);
}
return prevCol[len2];
}
unsigned int levenshteinDistance(const T &s1, const T & s2)
{
const size_t len1 = s1.size(), len2 = s2.size();
QVector<unsigned int> col(len2+1), prevCol(len2+1);
for (int i = 0; i < prevCol.size(); i++)
prevCol[i] = i;
for (unsigned int i = 0; i < len1; i++)
{
col[0] = i+1;
for (unsigned int j = 0; j < len2; j++)
col[j+1] = qMin( qMin( 1 + col[j], 1 + prevCol[1 + j]),
prevCol[j] + (s1[i]==s2[j] ? 0 : 1) );
col.swap(prevCol);
}
return prevCol[len2];
}
int KNUtil::similarity(const QString &string1, const QString &string2)
{
// Replace this into Sift4 algorithm.
//Wagner–Fischer algorithm Levenshtein distance.
//Based on
/* http://en.wikibooks.org/wiki/Algorithm_implementation/Strings/Levenshtein
_distance#C.2B.2B*/
//And
/*http://www.codeproject.com/Articles/13525/Fast-memory-efficient-Levenshtei
n-algorithm*/
//The original Levenshtein Algorithm will used a matrix whose size is
//length(string1) * length(string2), this is a common fast, memory efficient
//version. It can reduce the memory usage to length(string1) +
//length(string2). This won't reduce the time complexity.
//Thanks for QString and QChar, a chinese character will be description as
//a char.
//Initial the size.
const unsigned len1=string1.size(), len2=string2.size();
std::vector<size_t> col(len2+1), prevCol(len2+1);
//Fills the vector with ascending numbers, starting by 0
//Because of the F**K clang, we can NOT use itoa, or else these thing
//following can be done in only one sentence:
for(unsigned i=0; i<prevCol.size(); i++)
{
prevCol[i]=i;
}
//Use double std::min instead of std::min({,,}).
for(unsigned i=0; i<len1; i++)
{
col[0]=i+1;
for(unsigned j=0; j<len2; j++)
{
col[j+1]=std::min(std::min(1+col[j],
1+prevCol[1+j]),
prevCol[j]+(string1[i]!=string2[j]));
}
std::swap(col, prevCol);
}
return prevCol[len2];
}
unsigned int levenshteinDistance(const T &s1, const T & s2)
{
const size_t len1 = s1.size(), len2 = s2.size();
QVector<unsigned int> col(len2+1), prevCol(len2+1);
QVector<unsigned int> &c = col;
QVector<unsigned int> &p = prevCol;
for (int i = 0; i < p.size(); i++)
p[i] = i;
for (unsigned int i = 0; i < len1; i++)
{
c[0] = i+1;
for (unsigned int j = 0; j < len2; j++)
c[j+1] = qMin( qMin( 1 + c[j], 1 + p[1 + j]),
p[j] + (s1[i]==s2[j] ? 0 : 1) );
p = col;
c = prevCol;
}
return p[len2];
}
