1+ import javax .rmi .CORBA .Util ;
2+ import java .lang .reflect .Array ;
3+ import java .util .*;
4+ import java .util .Arrays ;
5+
6+
7+ import static java .lang .Integer .max ;
8+ import static java .lang .Integer .toBinaryString ;
9+ import static java .lang .Math .abs ;
10+ import static java .lang .Math .min ;
11+
12+ public class EditDistance implements EditDistanceInterface {
13+
14+ int c_i , c_d , c_r ;
15+ int n , m ;
16+ static int MAX = Integer .MAX_VALUE ;
17+ static int UNDEF = -1 ;
18+
19+ public EditDistance (int c_i , int c_d , int c_r ) {
20+ this .c_i = c_i ;
21+ this .c_d = c_d ;
22+ this .c_r = c_r ;
23+ }
24+
25+ public int [][] getEditDistanceDP (String s1 , String s2 ) {
26+
27+ int n = s1 .length ();
28+ int m = s2 .length ();
29+
30+ int [][] d = new int [n + 1 ][m + 1 ];
31+
32+ int factor_if_equal ;
33+
34+ //compute d_0_j and d_i_0
35+ for (int i = 0 ; i < n + 1 ; i ++)
36+ d [i ][0 ] = i * this .c_d ;
37+ for (int j = 0 ; j < m + 1 ; j ++)
38+ d [0 ][j ] = j * this .c_i ;
39+
40+ //compute the other values according to the lines
41+ for (int i = 1 ; i < n + 1 ; i ++) {
42+ for (int j = 1 ; j < m + 1 ; j ++) {
43+
44+
45+ if (s1 .charAt (i - 1 ) == s2 .charAt (j - 1 ))
46+ factor_if_equal = d [i - 1 ][j - 1 ];
47+ else
48+ factor_if_equal = d [i - 1 ][j - 1 ] + this .c_r ;
49+ d [i ][j ] = min (min (factor_if_equal , d [i - 1 ][j ] + this .c_d ), d [i ][j - 1 ] + this .c_i );
50+
51+ }
52+ }
53+
54+ return d ;
55+ }
56+
57+
58+ public List <String > getMinimalEditSequence (String s1 , String s2 ) {
59+ HashMap <Integer , Integer > d = new HashMap ();
60+ HashMap <Integer , String > op = new HashMap ();
61+ int [] aux_value = new int [3 ];
62+ String [] aux_string = new String [3 ];
63+ int X ;
64+ int arg ;
65+ int bound ;
66+ LinkedList <String > ls = new LinkedList <>();
67+
68+ int n = s1 .length ();
69+ int m = s2 .length ();
70+ this .n = n ;
71+ this .m = m ;
72+
73+
74+ X = 1 ; // upper bound
75+ while (true ) {
76+ bound = abs (n - m ) + X ;
77+
78+ d .clear ();
79+
80+ op .clear ();
81+
82+
83+ //compute d_0_j and d_i_0
84+ for (int i = 0 ; i < n + 1 ; i ++) {
85+ d .put (cantor (i , 0 ), i * this .c_d );
86+ op .put (cantor (i , 0 ), "delete(" + (i - 1 ) + ")" );
87+
88+ }
89+ for (int j = 0 ; j < m + 1 ; j ++) {
90+ d .put (cantor (0 , j ), j * this .c_i );
91+ }
92+ op .put (cantor (0 , 0 ), "nothing" );
93+ for (int j = 1 ; j < m + 1 ; j ++) {
94+ op .put (cantor (0 , j ), "insert(" + (j - 1 ) + "," + s2 .charAt (j - 1 ) + ")" );
95+ }
96+
97+
98+ //compute the other values according to the lines
99+ for (int i = 1 ; i < n + 1 ; i ++) {
100+
101+ for (int j = max (1 , i - bound ); j <= min (m , i + bound ); j ++) {
102+
103+
104+ if (i == j - bound )
105+ aux_value [0 ] = Integer .MAX_VALUE ;
106+ else {
107+ aux_value [0 ] = d .get (cantor (i - 1 , j )) + this .c_d ;
108+ aux_string [0 ] = "delete(" + (i - 1 ) + ")" ;
109+ }
110+ if (j == i - bound )
111+ aux_value [1 ] = Integer .MAX_VALUE ; // we do not care about value outside of the diagonal, so we set them to infinity to make sure they're not chosen
112+ else {
113+ aux_value [1 ] = d .get (cantor (i , j - 1 )) + this .c_i ;
114+ aux_string [1 ] = "insert(" + (j - 1 ) + "," + s2 .charAt (j - 1 ) + ")" ;
115+ }
116+ if (s1 .charAt (i - 1 ) == s2 .charAt (j - 1 )) {
117+ if ((j == i - bound ) || (i == j - bound ))
118+ aux_value [2 ] = Integer .MAX_VALUE ;
119+ else {
120+ aux_value [2 ] = d .get (cantor (i - 1 , j - 1 ));
121+ aux_string [2 ] = "nothing" ;
122+ }
123+
124+ }
125+ else {
126+ if ((j == i - bound ) || (i == j - bound ))
127+ aux_value [2 ] = Integer .MAX_VALUE ;
128+ else {
129+ aux_value [2 ] = d .get (cantor (i - 1 , j - 1 )) + this .c_r ;
130+ aux_string [2 ] = "replace(" + (i - 1 ) + "," + s2 .charAt (j - 1 ) + ")" ;
131+ }
132+
133+ }
134+
135+
136+ arg = argmin (aux_value );
137+ d .put (cantor (i , j ), aux_value [arg ]);
138+ op .put (cantor (i , j ), aux_string [arg ]);
139+
140+
141+ }
142+ }
143+
144+
145+
146+ if (X * (this .c_i + this .c_d ) >= d .get (cantor (n , m )))
147+ break ;
148+ else
149+ X *= 2 ;
150+
151+ }
152+
153+
154+ // read op in reverse to find the sequence we're looking for
155+
156+ Stack stack = new Stack (); //we need to read the insert instruction in the reverse order compared to the delete instructions
157+ int i = n ;
158+ int j = m ;
159+ while (i >= 0 && j >= 0 ) {
160+
161+ char first_letter = op .get (cantor (i , j )).charAt (0 );
162+ if (first_letter == "n" .toCharArray ()[0 ]){
163+ i --;
164+ j --;
165+ }
166+ if (first_letter == "d" .toCharArray ()[0 ]) {
167+ ls .add (op .get (cantor (i , j )));
168+ i --;
169+
170+ }
171+ if (first_letter == "i" .toCharArray ()[0 ]) {
172+ stack .push (op .get (cantor (i , j )));
173+ j --;
174+ }
175+ if (first_letter == "r" .toCharArray ()[0 ]){
176+ ls .add (op .get (cantor (i , j )));
177+ i --;
178+ j --;
179+ }
180+
181+ }
182+
183+ while (! stack .empty ())
184+ ls .add ((String ) stack .pop ()); //we need to read the insert instruction in the reverse order compared to the delete instructions
185+
186+ return ls ;
187+ }
188+
189+ public int argmin (int [] arr ) {
190+ int n = arr .length ;
191+ int res = 0 ;
192+
193+ for (int i = 1 ; i < n ; i ++) {
194+ if (arr [i ] < arr [res ])
195+ res = i ;
196+ }
197+
198+
199+ return res ;
200+
201+ }
202+
203+ public int cantor (int i , int j ){
204+ return i * (this .m + 1 ) + j ;
205+ }
206+
207+
208+ }
209+ // }
210+ // if (c_r == 6) {
211+ // ls.add("delete(1)");
212+ // ls.add("delete(1)");
213+ // ls.add("insert(2,c)");
214+ // ls.add("insert(3,b)");
215+ // }
216+ // else {
217+ // ls.add("replace(1,d)");
218+ // ls.add("replace(3,b)");
219+ // }
220+ // return ls;
221+ // /* Code block to be removed ends. */
222+ // }
223+
224+ //
225+ // for (int i=0; i<n+1; i++) {
226+ // for (int j = 0; j < m + 1; j++) {
227+ // System.out.print(op[i][j]);
228+ // }
229+ // System.out.println("");
230+ // }
0 commit comments