Created
May 25, 2024 04:19
-
-
Save jrade/8fd24da00d665934f07134fe83aa4337 to your computer and use it in GitHub Desktop.
Fast and memory efficient C++ implementations of the following string distances: LCS (Longest common subsequence), Levenshtein, Damerau-Levenshtein
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
// Copyright 2024 Johan Rade ([email protected]) | |
// Distributed under the MIT license (https://opensource.org/licenses/MIT) | |
inline size_t lcsDistance(const std::string& s, const std::string& t) | |
{ | |
size_t m = s.size() + 1; | |
size_t n = t.size() + 1; | |
const char* p = s.data(); | |
const char* q = t.data(); | |
if (m > n) { | |
std::swap(m, n); | |
std::swap(p, q); | |
} | |
static thread_local std::vector<size_t> buf; | |
buf.resize(std::max(2 * m, buf.size())); | |
size_t* prevDist = buf.data(); | |
size_t* dist = buf.data() + m; | |
for (size_t i = 0; i != m; ++i) | |
dist[i] = i; | |
for (size_t j = 1; j != n; ++j) { | |
std::swap(prevDist, dist); | |
dist[0] = j; | |
for (size_t i = 1; i != m; ++i) { | |
const size_t cost = (p[i - 1] == q[j - 1]) ? 0 : 2; | |
dist[i] = std::min(prevDist[i - 1] + cost, std::min(prevDist[i], dist[i - 1]) + 1); | |
} | |
} | |
return dist[m - 1]; | |
} | |
inline size_t levenshteinDistance(const std::string& s, const std::string& t) | |
{ | |
size_t m = s.size() + 1; | |
size_t n = t.size() + 1; | |
const char* p = s.data(); | |
const char* q = t.data(); | |
if (m > n) { | |
std::swap(m, n); | |
std::swap(p, q); | |
} | |
static thread_local std::vector<size_t> buf; | |
buf.resize(std::max(2 * m, buf.size())); | |
size_t* prevDist = buf.data(); | |
size_t* dist = buf.data() + m; | |
for (size_t i = 0; i != m; ++i) | |
dist[i] = i; | |
for (size_t j = 1; j != n; ++j) { | |
std::swap(prevDist, dist); | |
dist[0] = j; | |
for (size_t i = 1; i != m; ++i) { | |
const size_t cost = (p[i - 1] == q[j - 1]) ? 0 : 1; | |
dist[i] = std::min(prevDist[i - 1] + cost, std::min(prevDist[i], dist[i - 1]) + 1); | |
} | |
} | |
return dist[m - 1]; | |
} | |
inline size_t damerauLevenshteinDistance(const std::string& s, const std::string& t) | |
{ | |
size_t m = s.size() + 1; | |
size_t n = t.size() + 1; | |
const char* p = s.data(); | |
const char* q = t.data(); | |
if (m > n) { | |
std::swap(m, n); | |
std::swap(p, q); | |
} | |
if (m == 1) | |
return n - 1; | |
static thread_local std::vector<size_t> buf; | |
buf.resize(std::max(3 * m, buf.size())); | |
size_t* prevPrevDist = buf.data(); | |
size_t* prevDist = buf.data() + m; | |
size_t* dist = buf.data() + 2 * m; | |
// j = 0 | |
for (size_t i = 0; i != m; ++i) | |
dist[i] = i; | |
// j = 1 | |
std::swap(dist, prevDist); | |
dist[0] = 1; | |
for (size_t i = 1; i != m; ++i) { | |
const size_t substCost = (p[i - 1] == q[0]) ? 0 : 1; | |
dist[i] = std::min(prevDist[i - 1] + substCost, std::min(prevDist[i], dist[i - 1]) + 1); | |
} | |
for (size_t j = 2; j != n; ++j) { | |
std::swap(prevDist, prevPrevDist); | |
std::swap(dist, prevDist); | |
dist[0] = j; | |
{ | |
const size_t substCost = (p[0] == q[j - 1]) ? 0 : 1; | |
dist[1] = std::min(prevDist[0] + substCost, std::min(prevDist[1], dist[0]) + 1); | |
} | |
for (size_t i = 2; i != m; ++i) { | |
const size_t substCost = (p[i - 1] == q[j - 1]) ? 0 : 1; | |
const size_t transCost = (p[i - 1] == q[j - 2] && p[i - 2] == q[j - 1]) ? 1 : 2; | |
dist[i] = std::min( | |
std::min(prevDist[i - 1] + substCost, prevPrevDist[i - 2] + transCost), | |
std::min(prevDist[i], dist[i - 1]) + 1); | |
} | |
} | |
return dist[m - 1]; | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment