java字符串拆分正则表达式保留定界符

Question

Give an input string such as

"abbbcaababbbcaaabbca"

I want to split such a string into an array of groups " bca " " ab " " a " and " b "

So the above example would return

"ab", "b", "bca", "ab", "ab", "b", "bca", "a", "ab", "bca".

I have a 29 line piece of code of nested loops that accomplish this task (returns ArrayList). However, it would be nice to get this done with a one line regular expression.

Can this task be accomplished using the following method?

stringVar.split("regEX") 

Answer 1

Not an one-liner, but you can do it using Matcher.find with a loop.:

ArrayList<String> result = new ArrayList<String>();
String s = "abbbcaababbbcaaabbca";
Matcher m = Pattern.compile("bca|ab|a|b").matcher(s);
while (m.find())
    result.add(m.group());

DEMO

Answer 2

It can be accomplished using lookaround assertions , but @falsetru's answer is preferred over split ting.

String[] ss = "abbbcaababbbcaaabbca".split("(?<=bca|ab)|(?<=a(?=ab))|(?<=b(?=bca))");
System.out.println(Arrays.toString(ss)); //=> [ab, b, bca, ab, ab, b, bca, a, ab, bca]

If the string contains letters only, you could shorten this using a backreference.

String[] ss = "abbbcaababbbcaaabbca".split("(?<=bca|ab)|(?<=(.)(?=\\1))")

Answer 3

It looks like you are trying to split between identical characters. In that case you can use

stringVar.split("(?<=(\\w))(?=\\1)") 

but it will result in ab, b, bca, abab, b, bca, a, ab, bca , which means that abab will not be split.

If you want you can manually add case where you can decide that after ab or bca you also want to split via

stringVar.split("(?<=(\\w))(?=\\1)|(?<=ab|bca)") 

which now will return ab, b, bca, ab, ab, b, bca, a, ab, bca