From 49c105dfc1fc3a0334d03de4d361abf23a6f1898 Mon Sep 17 00:00:00 2001 From: Chris Dyer Date: Sat, 6 Sep 2014 02:55:29 -0400 Subject: ld commit --- mteval/levenshtein.h | 29 +++++++++++++++++++++++++++++ 1 file changed, 29 insertions(+) create mode 100644 mteval/levenshtein.h (limited to 'mteval') diff --git a/mteval/levenshtein.h b/mteval/levenshtein.h new file mode 100644 index 00000000..13a97047 --- /dev/null +++ b/mteval/levenshtein.h @@ -0,0 +1,29 @@ +#ifndef _LEVENSHTEIN_H_ +#define _LEVENSHTEIN_H_ + +namespace cdec { + +template +inline unsigned LevenshteinDistance(const V& a, const V& b) { + const unsigned m = a.size(), n = b.size(); + std::vector edit((m + 1) * 2); + for (unsigned i = 0; i <= n; i++) { + for (unsigned j = 0; j <= m; j++) { + if (i == 0) + edit[j] = j; + else if (j == 0) + edit[(i % 2) * (m + 1)] = i; + else + edit[(i % 2) * (m + 1) + j] = std::min(std::min( + edit[(i % 2) * (m + 1) + j - 1] + 1, + edit[((i - 1) % 2) * (m + 1) + j] + 1), + edit[((i - 1) % 2) * (m + 1) + (j - 1)] + + (a[j - 1] == b[i - 1] ? 0 : 1)); + } + } + return edit[(n % 2) * (m + 1) + m]; +} + +} + +#endif -- cgit v1.2.3