从字符串数组中获取包含指定字符的字符串列表

Question

I just had an interview, and the interviewer asked me the following question:

From an array of String (English words), write a function that returns all the words that contain a character.

I initially proposed something like:

public static List<String> find(String[] array, char c) {
     if(null ==  array || array.length == 0)
          return null;

     List<String> res = new ArrayList<String>();
     for(String s : array) {
          if(s.indexOf(c) > -1) {
               res.add(s);
          }
     }
     return res;
}

Then, the interviewer asked me to optimize my solution, what kind of data structure could be used. (She said my solution is brute-force)

I still don't see how I can optimize this. Any ideas?

Answer 1

You cannot optimize this function further, but I suppose interviewer wanted to hear something like this:

If we need to perform this operation several times on the same array, we can use following class (writing in pseudo code):

class FastStringArrayContainsChar {
    private Map<Char, Set<String>> index = new HashMap<Char, Set<String>>();
    public FastStringArrayContainsChar(String[] input) {
        for (String s: input) {
            for (int i=0; i<s.length; s++) {
                char c = s.charAt(i);
                if (index.contains(c))
                    index.get(c).put(s);
                else
                    index.put(c, new HashSet<String>(){{this.put(s);}});
            }
        }
    }
    public List<String> containsChar(char c) {
        return new ArrayList<String>(index.get(c));
    }
}

Answer 2

To check that the ith string contains character c , you check it letter by letter and return once found.

boolean check(String s, char c)
{   for(int i = 0, n = s.length(); i < n; i++)
       if(s[i] == c)
          return true;
    return false;
}

This does not change the worst-case time but might run fast in practice. Other than this I don't see how you can further improve code (i'm assuming that no other information was given regarding words in array).

Answer 3

Perhaps input array contains duplicates. It's using more memory though

public static List<String> find(String[] array, char c) {
     if(array == null || array.length == 0) {
        return null;
     }
     Map<String, Boolean> map = new HashMap<String, Boolean>();
     List<String> res = new ArrayList<String>();
     for(String s : array) {
        if (map.containsKey(s)) {
            if (map.get(s)) {
                res.add(s);
            }
        } else {
            if(s.indexOf(c) > -1) {
                 map.put(s, true);
                 res.add(s);
            } else {
                map.put(s, false);
            }
        }

     }
     return res;
}