Regex for splitting a string to get operators

Question

I have a String = A+B+CC-D0. I want to use a regular expression to split it apart to get an array of of the ops {+,+,-}

I have tried the regular expression:

"[^+-]"

But the output gave me a array of { ,+,+, ,+}, why?

    String s1 = "A+B+C-D";
    String s2 = "A0+B0";
    String[] op1 = s1.split([^+-]);
    String[] op2 = s2.split([^+-]);
    for(int i = 0; op1.length; i++){
        System.out.println(op1[i]);
    }

Here's the output:

Output of op1:
""
"+"
"+"
""
"-"

Output of op2:
""
""
"+"

just remove the empty elements from the array. – Avinash Raj Mar 01 '15 at 09:55 — Avinash Raj, Mar 01 '15 at 09:55
Yes I could, but then I need extra computations – Eagle Mar 01 '15 at 09:56 — Eagle, Mar 01 '15 at 09:56
Instead of splitting you could start matching. – HamZa Mar 01 '15 at 10:17 — HamZa, Mar 01 '15 at 10:17

Avinash Raj · Accepted Answer · 2015-03-01T10:08:59.357

2

Replace all the characters other than operators with empty string then split the resultant string according to the boundary which exists between two operators.

String s1 = "A+B+C-D";
String[] op1 = s1.replaceAll("[^+-]","").split("(?<=.)(?=.)");
for(String i: op1){
    System.out.println(i);
}

Output:

+
+
-

(?<=.) positive lookbehind which asserts that the match must be preceded by a character.
(?=.) Positive lookahead which asserts that the match must be followed by a character.

edited Mar 01 '15 at 10:08

answered Mar 01 '15 at 09:58

Avinash Raj

172,303
28
230
274

Maybe you should explain what `split("(?<=.)(?=.)")` is doing exactly. It's quite cryptic. Still +1 because this works quite well. – peter.petrov Mar 01 '15 at 10:08

Rohit Jain · Answer 2 · 2015-03-01T10:16:14.910

2

The problem is, you're splitting on single character, that is not + or -. When you split a string - ABC, it will get split 3 times - on A, B and C respectively, and hence you'll get an array - ["", "", "", ""].

To avoid this issue, use quantifier on regex:

String s1 = "A+B+C-D";
String s2 = "A0+B0";
String[] op1 = s1.split("[^+-]+");
String[] op2 = s2.split("[^+-]+");
System.out.println(Arrays.toString(op1));
System.out.println(Arrays.toString(op2));

this is splitting on "[^+-]+".

Now to remove the empty element at the beginning of array, you've to get rid of first delimiter from string, using replaceFirst() may be:

String[] op1 = s1.replaceFirst("[^+-]+", "").split("[^+-]+");
System.out.println(Arrays.toString(op1)); // [+, +, -]

edited Mar 01 '15 at 10:16

answered Mar 01 '15 at 09:59

Rohit Jain

209,639
45
409
525

1

I don't think it's fixed in Java 8. I just tried it. Maybe I'm missing something. – peter.petrov Mar 01 '15 at 10:06
this also gives me an extra empty element. – Avinash Raj Mar 01 '15 at 10:07
Yes I'm running java 8 and there is still a blank in the array – Eagle Mar 01 '15 at 10:08
@peter.petrov Yup, haven't been fixed. I guess I updated the fix in wrong answer. :( – Rohit Jain Mar 01 '15 at 10:12
It has been fixed (or "fixed" is not the right word since this behaviour is expected, "improved" is better word) in Java 8 but only for zero-width delimiters, like `""` or look-around. – Pshemo Mar 01 '15 at 10:13

Regex for splitting a string to get operators

2 Answers2