拆分整数值，但不拆分浮点值

Question

I have a java program where I need to split on integer values but not floating point values

ie. "1/\\\\2" should produce: [1,/\\\\,2] but "1.0/\\\\2.0" should produce: [1.0,/\\\\,2.0]

does anybody have any ideas?

or could anybody point me in the direction of how to split on the specific strings "\\\\/" and "/\\\\" ?

UPDATE: sorry! one more case! for the string "100 /\\ 3.4e+45" I need to split it into: [100,/\\,3.4,e,+,45]

my current regex is (kind of really ugly):

line.split("\\s+|(?<=[-+])|(?=[-+])|(?:(?<=[0-9])(?![0-9.]|$))|(?:(?<![0-9.]|^)(?=[0-9]))|(?<=[-+()])|(?=[-+()])|(?<=e)|(?=e)");

and for the string: "100 /\\ 3.4e+45" is giving me: [100,/\\,3.4,+,45]

Answer 1

You could try something like this:

    String regex = "\\d+(.\\d+)?", str = "1//2";
    Matcher m = Pattern.compile(regex).matcher(str);
    ArrayList<String> list = new ArrayList<String>();

    int index = 0;
    for(index = 0 ; m.find() ; index = m.end()) {
        if(index != m.start()) list.add(str.substring(index, m.start()));
        list.add(str.substring(m.start(), m.end()));
    }

    list.add(str.substring(index));

The idea is to find number using regex and Matcher , and also add the strings in between.

Answer 2

This regex should do it:

(?:(?<=[0-9])(?![0-9.]|$))|(?:(?<![0-9.]|^)(?=[0-9]))

It's two checks, basically matching:

1. A digit not followed by a digit, a decimal point, or the end of text.
2. A digit not preceded by a digit, a decimal point, or the start of text.

It will match the empty space after/before the digit, so you can use this regex in split() .

See regex101 for demo.

---

Follow-up

could anybody point me in the direction of how to split on the specific strings "\\/" and "/\\""

If you want to split before a specific pattern, use a positive lookahead : (?=xxx) . If you want to split after a specific pattern, use a positive lookbehind : (?<=xxx) . To do either, separate by | :

(?<=xxx)|(?=xxx)

where xxx is the text \\/ or /\\ , ie the regex \\\\/|/\\\\ , and doubling for Java string literal:

"(?<=\\\\/|/\\\\)|(?=\\\\/|/\\\\)"

See regex101 for demo.