Java - all substring for a given regex

Question

I have a pattern like the following

API:ADD|TYPE:ABC...MATCH:TRUE

[LOTS OF OTHER LOG LINES]

API:ADD|TYPE:ABC...MATCH:TRUE

[LOTS OF OTHER LOG LINES]

API:ADD|TYPE:DEF...MATCH:TRUE

I tried the following regex:

(API:.*MATCH:(TRUE|FALSE))

while (matcher.find()) {
    System.out.println(i + " occurence");
    i++;
    matches.add(matcher.group());
}

It matches from first "API" to last "TRUE" and hence only one substring is returned! I want three substrings (in this scenario) starting from "API" till either "TRUE" or "FALSE".

Appreciate your help in this regard.Thanks.

Edit:

-------------------------------------------------------
20:31:57    CALL    add     35  
-------------------------------------------------------
20:31:57    REASON  API:ADD|TYPE:ABC|ErrorType:VALIDATION|Error Message:User already has|MATCH:FALSE

You need to know how many groups you want to capture with regex. Do you expect that you will have a fixed number? — Tim Biegeleisen, Oct 28 '15 at 07:04
How are you reading the input?. Line by line?. or all those lines are in one String?. — TheLostMind, Oct 28 '15 at 07:05
Please show us an entire sample line along with what you want to extract. — Tim Biegeleisen, Oct 28 '15 at 07:05
I receive it as plain text from HTML response and have the entire content in a String. It can be any number of times - wherever the pattern occurs in the String,need them in a list. — user104309, Oct 28 '15 at 07:12
`.*` wouldn't match across new lines so I believe OP has buffered the content into a string without the new lines. — Ravi K Thapliyal, Oct 28 '15 at 07:13
@user104309 If you print out the HTML response do you see any new lines? — Ravi K Thapliyal, Oct 28 '15 at 07:13
You need to show us an entire line so we know what to work with here. — Tim Biegeleisen, Oct 28 '15 at 07:17
@TimBiegeleisen Please see my edit. I guess all new lines are removed while storing in String (jsoup) — user104309, Oct 28 '15 at 08:49
How are you reading in this log file? From your example, it appears that you are reading in all 3 lines into a single string. If this be the case, then the regex isn't the problem. Please clarify this point for us. — Tim Biegeleisen, Oct 28 '15 at 08:53
@user104309 The solution is very simple. If you (or jsoup) aren't preserving new lines use `.*?`, otherwise make sure the newlines are not stripped and your current regex should then work without any issues provided your pattern never occurs more than once per line. — Ravi K Thapliyal, Oct 28 '15 at 08:59

Ravi K Thapliyal · Answer 1 · 2015-10-28T07:04:46.640

1

Use a non-greedy quantifier .*? and the regex will start matching as little as possible.

(API:.*?MATCH:(TRUE|FALSE))

edited Oct 28 '15 at 07:04

answered Oct 28 '15 at 07:04

Ravi K Thapliyal

51,095
9
76
89

score 1 · Answer 2 · edited May 23 '17 at 12:29

1

This is because you use .* which is a greedy quantifier.

You should try to use .*? which is reluctant -- that means that it matches the smallest possible substring.

For more info about greedy vs. reluctant see the excellent answers here: Greedy vs. Reluctant vs. Possessive Quantifiers

edited May 23 '17 at 12:29

Community

1
1

answered Oct 28 '15 at 07:05

Stefan Winkler

3,871
1
18
35

Java - all substring for a given regex

2 Answers2