Java regular expressions - what is wrong with this code

Question

I am trying to extract the name of a property referenced in a string using the $() construct. For instance, if bb=xo-xo, then "aa$(bb)aa" expands to "aaxo-xoaa".

Here is the code:

public static void main(String[] args) {
  final String PROPERTY_NAME_REGEX = "\\w+(?:\\.\\w+)*";
  final String PROPERTY_REFERENCE_REGEX = "\\$\\((" + PROPERTY_NAME_REGEX + ")\\)";
  Pattern pattern = Pattern.compile(PROPERTY_REFERENCE_REGEX);
  String value = "hhh $(aa.bbcc.dd) @jj $(aakfd) j";
  Matcher matcher = pattern.matcher(value);
  StringBuffer sb = new StringBuffer();
  while (matcher.find()) {
    System.out.println(String.format("\"%s\" at [%d-%d)",
      matcher.group(),
      matcher.start(),
      matcher.end()));
    for (int i = 0; i < matcher.groupCount(); ++i) {
      System.out.println(String.format("group[%d] = %s", i, matcher.group(i)));
    }
  }
}

And it displays:

"$(aa.bbcc.dd)" at [4-17)
group[0] = $(aa.bbcc.dd)
"$(aakfd)" at [22-30)
group[0] = $(aakfd)

But I was hoping to get the following output:

"$(aa.bbcc.dd)" at [4-17)
group[0] = aa.bbcc.dd
"$(aakfd)" at [22-30)
group[0] = aakfd

What am I doing wrong?

"Some people, when confronted with a problem, think 'I know, I'll use regular expressions.' Now they have two problems." - Jamie Zawinski — nfechner
– nfechner, Commented Mar 4, 2012 at 19:08
take a look at the for loop. The groups index begin with 1 for the parenthesis enclosed matches — Omnaest
– Omnaest, Commented Mar 4, 2012 at 19:18
Sorry for being a smartass. You should use a parser for this, not regex.. You could take a look at the ExpandProperties class from ant, which does exactly this type of replacements. — nfechner
– nfechner, Commented Mar 4, 2012 at 19:22
Agreed. I have another question -stackoverflow.com/questions/9558411/… — mark
– mark, Commented Mar 4, 2012 at 20:14

Amber · Accepted Answer · 2012-03-04 19:23:35Z

2

To answer your specific problem, you should be looking at group[1], not group[0].

The Matcher.groupCount() method does not include group[0] in the count, thus your for loop is never showing you the group[1] matches because i < matcher.groupCount() is false.

Change your condition to i <= matcher.groupCount() and your output will be more enlightening.

That said, there are better ways of doing this than writing your own regex - e.g. http://api.dpml.net/ant/1.6.4/org/apache/tools/ant/filters/ExpandProperties.html

answered Mar 4, 2012 at 19:23

Amber

532k89 gold badges643 silver badges558 bronze badges

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

1 Comment

mark Over a year ago

Agreed. I have another question -stackoverflow.com/questions/9558411/…

erikxiv · Accepted Answer · 2012-03-04 19:27:55Z

0

Group 0 is always the entire match, regardless of any specified capturing groups. To top this off, Matcher.groupCount() returns the number of capturing groups, excluding the entire match. To get the result you were after, change your for loop to the following (notice that it starts at 1, and continues one step further due to the added equal sign):

for (int i = 1; i <= matcher.groupCount(); i++) {

answered Mar 4, 2012 at 19:27

erikxiv

4,0952 gold badges26 silver badges22 bronze badges

Collectives™ on Stack Overflow

Java regular expressions - what is wrong with this code

2 Answers 2

1 Comment

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related