I saw this code implementation here. It basically takes two strings, finds the longest common substring, and returns the length of it. I wanted to modify it slightly to get the starting indexes of the substrings for each words, but just can't figure out. I know it should be possible, since we are working with indexes of the string. I will write my edited version of the code below:
public class Main {
public class Answer {
int i, j, len;
Answer(int i, int j, int len) {
this.i = i;
this.j = j;
this.len = len;
}
}
public Answer find(String s1,String s2){
int n = s1.length();
int m = s2.length();
Answer ans = new Answer(0, 0, 0);
int[] a = new int[m];
int b[] = new int[m];
for(int i = 0;i<n;i++){
for(int j = 0;j<m;j++){
if(s1.charAt(i)==s2.charAt(j)){
if(i==0 || j==0 )a[j] = 1;
else{
a[j] = b[j-1] + 1;
}
ans.len = Math.max(ans.len, a[j]);
ans.i = i;
ans.j = j;
}
}
int[] c = a;
a = b;
b = c;
}
return ans;
}
}
I am assuming if these are the two strings :
s1
= "abcdxyz"
s2
= "xyzabcd"
then since abcd
is longest common substring so you need index of this substring in both s1 and s2 which is 0,3 respectively.
There are two solution for this :
Here, I have created an index
array where I am storing the starting index of both the string with index 0 of index array storing for s1 and index 1 storing for s2.
public Answer find(String s1,String s2){
int n = s1.length();
int m = s2.length();
Answer ans = new Answer(0, 0, 0);
int[] a = new int[m];
int b[] = new int[m];
int indexes[] = new int[2];
for(int i = 0;i<n;i++){
for(int j = 0;j<m;j++){
if(s1.charAt(i)==s2.charAt(j)){
if(i==0 || j==0 )a[j] = 1;
else{
a[j] = b[j-1] + 1;
}
if(a[j]>ans.len) {
ans.len = a[j];
indexes[0]=(i+1) - ans.len;
indexes[1]=(j+1) - ans.len;
}
ans.i = i;
ans.j = j;
}
}
int[] c = a;
a = b;
b = c;
}
return ans;
}
I am not sure what your Answer
object i and j values are doing but we can make them as well store these values with i
storing for s1 string and j
storing for s2 string instead of creating different index
array as in Solution 1.
public Answer find(String s1,String s2){
int n = s1.length();
int m = s2.length();
Answer ans = new Answer(0, 0, 0);
int[] a = new int[m];
int b[] = new int[m];
int indexes[] = new int[2];
for(int i = 0;i<n;i++){
for(int j = 0;j<m;j++){
if(s1.charAt(i)==s2.charAt(j)){
if(i==0 || j==0 )a[j] = 1;
else{
a[j] = b[j-1] + 1;
}
if(a[j]>ans.len) {
ans.len = a[j];
ans.i=(i+1) - ans.len;
ans.j=(j+1) - ans.len;
}
}
}
int[] c = a;
a = b;
b = c;
}
return ans;
}
Currently this doesn't calculate LCS right . Issue is you are not making array a
as empty after running your second loop each time because of which if characters do not match in next run corresponding index of a
stores previous value only but it should be 0.
Update code is :
public Answer find(String s1,String s2){
int n = s1.length();
int m = s2.length();
Answer ans = new Answer(0, 0, 0);
int[] a;
int b[] = new int[m];
int indexes[] = new int[2];
for(int i = 0;i<n;i++){
a = new int[m];
for(int j = 0;j<m;j++){
if(s1.charAt(i)==s2.charAt(j)){
if(i==0 || j==0 )a[j] = 1;
else{
a[j] = b[j-1] + 1;
}
if(a[j]>ans.len) {
ans.len = a[j];
ans.i=(i+1) - ans.len;
ans.j=(j+1) - ans.len;
}
}
}
b = a;
}
return ans;
}