UVa #494 - regex [^a-zA-z]+ to split words using Java

Question

I was playing with UVa #494 and I managed to solve it with the code below:

import java.io.BufferedReader;
import java.io.IOException;
import java.io.InputStreamReader;

class Main {    
    public static void main(String[] args) throws IOException{
        BufferedReader in = new BufferedReader(new InputStreamReader(System.in));
        String line;
        while((line = in.readLine()) != null){
            String words[] = line.split("[^a-zA-z]+");
            int cnt = words.length;
            // for some reason it is counting two words for 234234ddfdfd and words[0] is empty
            if(cnt != 0 && words[0].isEmpty()) cnt--; // ugly fix, if has words and the first is empty, reduce one word
            System.out.println(cnt);
        }
        System.exit(0);
    }
}

I built the regex "[^a-zA-z]+" to split the words so for example the strings abc..abc or abc432abc should be splitted as ["abc", "abc"]. However, when I try the string 432abc, I have as a result ["", "abc"] - the first element from words[] is just an empty string but I was expecting to have just ["abc"]. I can't figure out why this regex gives me the first element as "" for this case.

Answer 1

Check the split reference page: split reference

Each element of separator defines a separate delimiter character. If two delimiters are adjacent, or a delimiter is found at the beginning or end of this instance, the corresponding array element contains Empty. The following table provides examples.

Since you have several consecutive delimiter characters, you get empty array elements