Mercurial > dedupe

comparison CachedEditDistance.hpp @ 78:9744ec195be3

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Encapsulate EditDistance with caching.
author Tom Fredrik Blenning Klaussen <bfg@bfgconsult.no>
date Thu, 10 Oct 2013 01:07:52 +0200
parents
children
comparison
equal deleted inserted replaced
77:a827f3687c4a 78:9744ec195be3
1 #ifndef CACHEDEDITDISTANCE_HPP
2 #define CACHEDEDITDISTANCE_HPP
3
4 #include "DBCache.hpp"
5
6 #include "EditDistance.hpp"
7
8 template<typename Value>
9 struct InsertRegulator<OrderedPair<UniqueString>, Value >
10 {
11 uint n;
12 void start()
13 {
14 n = 0;
15 HuffmanString::getSet().setAutoRebuild(false);
16 }
17
18 static void finish()
19 {
20 HuffmanString::getSet().rebuild();
21 HuffmanString::getSet().setAutoRebuild(true);
22 }
23
24 void next()
25 {
26 if (++n == 2048)
27 HuffmanString::getSet().rebuild();
28 }
29 };
30
31 class CachedEditDistance {
32 protected:
33 typedef DBCache<OrderedPair<UniqueString>, int, true> cacheType;
34 public:
35 static int Compute(QString a, QString b, bool removeDiacritics = false);
36 static void removeDiacriticsNoCopy(QString& in)
37 {
38 EditDistance::removeDiacriticsNoCopy(in);
39 }
40 static QString removeDiacritics(const QString& in)
41 {
42 return EditDistance::removeDiacritics(in);
43 }
44
45 static cacheType* cache;
46 };
47
48 #endif //CACHEDEDITDISTANCE_HPP