Regular expression, get an array from captured group Java

Question

I have some text that occurs in a particular format as shown below Each line starts with a + followed by a space and some text It then has a bunch of lines stuck together that start with a minus sign or @ or % or * and space and some text following it. I would like to capture each block separately from below using Regular expressions.

+ you rock
- I rock and rule.

+ you rule
- I rock and rule.
- That is a perfect artificial entity.

+ you made a mistake
- That is impossible. I never make mistakes.
- I guess so, something must have gone wrong.

Output

Block 1 + you rock - I rock and rule.

Block 2 + you rule - I rock and rule. - That is a perfect artificial entity.

This is my current regular expression

(^\+.*$)(?:\r?\n)(?:(^[-%@\*].*$)(?:\r?\n)?)+

In the above expression, Group 1 = (^+.$) that captures the statement following a +, group 2 = (^[-%@*].$) that captures the second part but notice that there may be more than one statement which has a - at the start.

When I run a while loop in Java code

Pattern pattern = Pattern.compile(regex, Pattern.MULTILINE);
Matcher matcher = pattern.matcher(contents);
while (matcher.find()) {
    // This gives me the item following +
    System.out.println(matcher.group(1));
    // This ONLY gives me the last item following -, how do I get all
    System.out.println(matcher.group(2));
}

How do I get all the statements that have a minus sign in front of them as an array?

Basically, you can't do it. You'll need split or a loop. But I want to question your premise: Why is it important to solve this with a regular expression? What's wrong with other types of solutions? Are they somehow uncool? I've been trying to understand this for years--why so many programmers seem to assume that they need to find the magical regex to solve something. Second question: Do you think that code like (^\+.*$)(?:\r?\n)(?:(^[-%@\*].*$)(?:\r?\n)?)+ makes a program more or less readable? — ajb
– ajb, Commented Sep 5, 2016 at 16:53
nice point but honestly i am not aware of what other solutions may work for this problem, i have a file with around 300000 lines of text in this format and would like to parse them to blocks and extract and process information from them — PirateApp
– PirateApp, Commented Sep 5, 2016 at 16:57
btw how can I split or loop the second part which has a - sign — PirateApp
– PirateApp, Commented Sep 5, 2016 at 16:57
Hint: since Java 8 we have \R which represents line separators (including \r\n so you can use it instead of \r?\n). Anyway, why do you want to have these parts in array? What are you trying to achieve? Maybe reading this text line-by-line, while checking if you found empty line would be easier? — Pshemo
– Pshemo, Commented Sep 5, 2016 at 16:58
"Not aware of other solutions"? Can't you just read one line at a time and use charAt() or startsWith() to see if the first character is + or -? — ajb
– ajb, Commented Sep 5, 2016 at 17:00

Andriy Kryvtsun · Accepted Answer · 2016-09-06 16:04:57Z

2

Using this regexp ^\+[^+]* with m and g modifiers gives you needed result
https://regex101.com/r/bH1aQ9/1

On your test data result will be 3 groups start with + character.

The solution idea is to treat all you lines like one big line and split it on groups is started with + and haven't + inside them.

Update

To take into account @Pshemo note about possible + characters inside lines

Regexp ^\+.*?(^$|\Z) with with s, m and g modifiers
https://regex101.com/r/bH1aQ9/1

edited Sep 6, 2016 at 16:04

answered Sep 5, 2016 at 17:20

Andriy Kryvtsun

3,3933 gold badges33 silver badges45 bronze badges

Sign up to request clarification or add additional context in comments.

10 Comments

Andriy Kryvtsun Over a year ago

@PirateApp [^+]* means 0 or more characters in brackets (it defines * postfix) and ^+ inside the brackets means any character but not (^) plus.

Andriy Kryvtsun Over a year ago

@PirateApp thx! But be careful with regexps regex.info/blog/2006-09-15/247

Pshemo Over a year ago

@PirateApp Just be aware that this solution will not accept any + inside your group (except first one). So for instance + I like + operator will stop at + I like .

Andriy Kryvtsun Over a year ago

@Pshemo you are right, but in source data there was not such corner case. I would update the task conditions with this note.

Andriy Kryvtsun Over a year ago

@Pshemo added the second solution bases on your note

|

Collectives™ on Stack Overflow

Regular expression, get an array from captured group Java

1 Answer 1

10 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

10 Comments

Your Answer

Sign up or log in

Post as a guest

Related