Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(355)

Side by Side Diff: pkg/analysis_services/lib/src/correction/levenshtein.dart

Issue 417263003: Use a Levenshtein calculating algorithm with a threshold. (Closed) Base URL: https://dart.googlecode.com/svn/branches/bleeding_edge/dart
Patch Set: Fixes for review comments Created 6 years, 5 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch | Annotate | Revision Log
OLDNEW
1 library levenshtein; 1 library levenshtein;
2 2
3 import 'dart:math'; 3 import 'dart:math' as math;
4 4
5 /// Levenshtein algorithm implementation based on: 5 /**
6 /// http://en.wikipedia.org/wiki/Levenshtein_distance#Iterative_with_two_matrix_ rows 6 * The value returned by [levenshtein] if the distance is determined
7 /// 7 * to be over the specified threshold.
8 /// Implementation: https://github.com/conradkleinespel/levenshtein-dart 8 */
9 int getLevenshteinDistance(String s, String t, {bool caseSensitive: true}) { 9 const int LEVENSHTEIN_MAX = 1 << 20;
10
11 const int _MAX_VALUE = 1 << 10;
12
13 /**
14 * Find the Levenshtein distance between two [String]s if it's less than or
15 * equal to a given threshold.
16 *
17 * This is the number of changes needed to change one String into another,
18 * where each change is a single character modification (deletion, insertion or
19 * substitution).
20 *
21 * This implementation follows from Algorithms on Strings, Trees and Sequences
22 * by Dan Gusfield and Chas Emerick's implementation of the Levenshtein distance
23 * algorithm.
24 */
25 int levenshtein(String s, String t, int threshold, {bool caseSensitive: true}) {
26 if (s == null || t == null) {
27 throw new ArgumentError('Strings must not be null');
28 }
29 if (threshold < 0) {
30 throw new ArgumentError('Threshold must not be negative');
31 }
32
10 if (!caseSensitive) { 33 if (!caseSensitive) {
11 s = s.toLowerCase(); 34 s = s.toLowerCase();
12 t = t.toLowerCase(); 35 t = t.toLowerCase();
13 } 36 }
14 37
15 if (s == t) { 38 int s_len = s.length;
16 return 0; 39 int t_len = t.length;
40
41 // if one string is empty,
42 // the edit distance is necessarily the length of the other
43 if (s_len == 0) {
44 return t_len <= threshold ? t_len : LEVENSHTEIN_MAX;
17 } 45 }
18 if (s.length == 0) { 46 if (t_len == 0) {
19 return t.length; 47 return s_len <= threshold ? s_len : LEVENSHTEIN_MAX;
20 } 48 }
21 if (t.length == 0) { 49 // the distance can never be less than abs(s_len - t_len)
22 return s.length; 50 if ((s_len - t_len).abs() > threshold) {
51 return LEVENSHTEIN_MAX;
23 } 52 }
24 53
25 List<int> v0 = new List<int>.filled(t.length + 1, 0); 54 // swap the two strings to consume less memory
26 List<int> v1 = new List<int>.filled(t.length + 1, 0); 55 if (s_len > t_len) {
27 56 String tmp = s;
28 for (int i = 0; i < t.length + 1; i < i++) { 57 s = t;
29 v0[i] = i; 58 t = tmp;
59 s_len = t_len;
60 t_len = t.length;
30 } 61 }
31 62
32 for (int i = 0; i < s.length; i++) { 63 // 'previous' cost array, horizontally
33 v1[0] = i + 1; 64 List<int> p = new List<int>.filled(s_len + 1, 0);
65 // cost array, horizontally
66 List<int> d = new List<int>.filled(s_len + 1, 0);
67 // placeholder to assist in swapping p and d
68 List<int> _d;
34 69
35 for (int j = 0; j < t.length; j++) { 70 // fill in starting table values
36 int cost = (s[i] == t[j]) ? 0 : 1; 71 int boundary = math.min(s_len, threshold) + 1;
37 v1[j + 1] = min(v1[j] + 1, min(v0[j + 1] + 1, v0[j] + cost)); 72 for (int i = 0; i < boundary; i++) {
73 p[i] = i;
74 }
75
76 // these fills ensure that the value above the rightmost entry of our
77 // stripe will be ignored in following loop iterations
78 _setRange(p, boundary, p.length, _MAX_VALUE);
79 _setRange(d, 0, d.length, _MAX_VALUE);
80
81 // iterates through t
82 for (int j = 1; j <= t_len; j++) {
83 // jth character of t
84 int t_j = t.codeUnitAt(j - 1);
85 d[0] = j;
86
87 // compute stripe indices, constrain to array size
88 int min = math.max(1, j - threshold);
89 int max = math.min(s_len, j + threshold);
90
91 // the stripe may lead off of the table if s and t are of different sizes
92 if (min > max) {
93 return LEVENSHTEIN_MAX;
38 } 94 }
39 95
40 for (int j = 0; j < t.length + 1; j++) { 96 // ignore entry left of leftmost
41 v0[j] = v1[j]; 97 if (min > 1) {
98 d[min - 1] = _MAX_VALUE;
42 } 99 }
100
101 // iterates through [min, max] in s
102 for (int i = min; i <= max; i++) {
103 if (s.codeUnitAt(i - 1) == t_j) {
104 // diagonally left and up
105 d[i] = p[i - 1];
106 } else {
107 // 1 + minimum of cell to the left, to the top, diagonally left and up
108 d[i] = 1 + math.min(math.min(d[i - 1], p[i]), p[i - 1]);
109 }
110 }
111
112 // copy current distance counts to 'previous row' distance counts
113 _d = p;
114 p = d;
115 d = _d;
43 } 116 }
44 117
45 return v1[t.length]; 118 // if p[n] is greater than the threshold,
119 // there's no guarantee on it being the correct distance
120 if (p[s_len] <= threshold) {
121 return p[s_len];
122 }
123
124 return LEVENSHTEIN_MAX;
46 } 125 }
126
127 void _setRange(List<int> a, int start, int end, int value) {
128 for (int i = start; i < end; i++) {
129 a[i] = value;
130 }
131 }
OLDNEW
« no previous file with comments | « pkg/analysis_services/lib/src/correction/fix.dart ('k') | pkg/analysis_services/test/correction/levenshtein_test.dart » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698