暴力查询即可,外面一层遍历target字符串,内层对source进行遍历,直到当前的内层遍历source中已无可匹配字符,在进行下一次source遍历,直到外层遍历结束。最终统计进行了几次内层遍历。
优化的方法:将内层遍历改为二分查找,进而衍生的需要用map存储source的每个字符对应的下标的列表。
public class test {
public int shortestWay(String source, String target){
Map<Character, List<Integer>> map = getCharIndexs(source);
int res = 1;
int index = -1;
for(int i=0; i<target.length(); i++){
char ch = target.charAt(i);
int nextIndex = bsearch(map, ch, index);
if(nextIndex == -1 && index == -1){
return -1;
}else if(nextIndex == -1){
res++;
index = -1;
i--;
}else{
index = nextIndex;
}
}
return res;
}
private Map<Character, List<Integer>> getCharIndexs(String src){
Map<Character, List<Integer>> map = new HashMap<>();
for(int i=0; i<src.length(); i++){
char ch = src.charAt(i);
map.computeIfAbsent(ch, k -> new ArrayList<>()).add(i);
}
return map;
}
private int bsearch(Map<Character, List<Integer>> map, char ch, int index){
if(!map.containsKey(ch)){
return -1;
}
List<Integer> array = map.get(ch);
int l=0;
int r=array.size()-1;
while(l<r){
int mid = l + (r-l)/2;
if(array.get(mid) <= index){
l = mid + 1;
}else{
r = mid;
}
}
return array.get(l) > index ? array.get(l) : -1;
}
}