Java: Using Regular Expressions to Extract various number of values in a String

Question

I want to write a function to extract various number of values from a String according to a regex pattern:

Here is my function code:

/**
 * Get substrings in a string using groups in regular expression.
 * 
 * @param str
 * @param regex
 * @return
 */
public static String[] regexMatch(String str, String regex) {
    String[] rtn = null;
    if (str != null && regex != null) {
        Pattern pat = Pattern.compile(regex);
        Matcher matcher = pat.matcher(str);
        if (matcher.find()) {
            int nGroup = matcher.groupCount();
            rtn = new String[nGroup];
            for (int i = 0; i < nGroup; i++) {
                rtn[i] = matcher.group(i);
            }
        }
    }
    return rtn;
}

When I test it using:

String str = "nets-(90000,5,4).dat";
String regex = "(\\d+),(\\d+),(\\d+)";
String[] rtn = regexMatch(str, regex);

I get:

rtn: [90000,5,4,90000,5]

How can I get rtn to be [90000,5,4] as I expected?

You could restructure your solution a little by removing everything except commas and numbers and then split on commas. However, this depends on the other strings you process having a similar structure. — awksp
– awksp, Commented May 15, 2014 at 22:10

Pshemo · Accepted Answer · 2014-05-15 22:45:36Z

1

Your array currently store

[0] -> 90000,5,4
[1] -> 90000
[2] -> 5

That is why you are seeing as output [90000,5,4,90000,5]. It is because group(0) represents entire match so it returns 90000,5,4.

What you need is match from groups 1, 2 and 3.

(\\d+),(\\d+),(\\d+)
   1      2      3

So change

rtn[i] = matcher.group(i);

to

rtn[i] = matcher.group(i+1);

edited May 15, 2014 at 22:45

answered May 15, 2014 at 22:11

Pshemo

125k25 gold badges194 silver badges280 bronze badges

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

Comments

GKproggy · Accepted Answer · 2014-05-15 22:14:55Z

1

First, I would start the for loop with 1 so you can get the grouping you are declaring in your regex. The loop should look like this:

for (int i = 1; i <= nGroup; i++) {
            rtn[i] = matcher.group(i);
        }

Group 0 is known to be the entire matching string for your regex. The grouping is from:

String regex = "(\\d+),(\\d+),(\\d+)";

You would say matcher.group(1), matcher.group(2), and matcher.group(3) will give you what you want.

answered May 15, 2014 at 22:14

GKproggy

1881 silver badge12 bronze badges

4 Comments

Pshemo Over a year ago

If OP will start loop with 1 then returned array will not fill [0] and you will try to fill [length]. Unless you will write it as rtn[i-1] = matcher.group(i); which will be variation of my answer.

Reimeus Over a year ago

this would produce an IndexOutOfBoundsException

Changwang Zhang Over a year ago

I should start from 0 to nGroup-1 I feel.

GKproggy Over a year ago

true dat, matcher.groupCount() doesnt include index 0 as part of it.

Collectives™ on Stack Overflow

Java: Using Regular Expressions to Extract various number of values in a String

2 Answers 2

Comments

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related