SCJP6 study - regex string.split - strange results, can anyone explain please?

[Nigel Shrin]

eg based on a whizlabs q:

 	public static void main(String[] args) {

		String str = "111211111121111"; 
		String[] s=str.split("1{2}"); // delimiter is '11'
		System.out.println("The length of s is:\t" + s.length + "\n");
		
		for(int i=0; i<s.length; i++){ 
			System.out.printf("%d %s\n", i, s[i]); 

The length of s is: 5

0
1 12
2
3
4 2

I was expecting even numbers of 1's ie 11 to disappear, so I would have 122 remaining.
It has removed 6 pairs of 11. How do you account for the way it is storing nulls?
>

[Henry Wong]

Start with the original string...

"111211111121111"

Do a regex find to get the delimiters...

"11 12 11 11 11 2 11 11"

Add "X" to show the blank tokens...

"X 11 12 11 X 11 X 11 2 11 X 11 X "

Remove the trailing blanks, because you used the version of split() that removes them...

"X 11 12 11 X 11 X 11 2"

So, you get a final result of 5 tokens, which are blank, 12, blank, blank, and 2.

Henry

[Ninad Kulkarni]

@Henry
Great explaination.

[Nigel Shrin]

Hi Henry, I really appreciate the reply, the penny has dropped!

I've created a more complex example and managed to predict the result.

Cheers!