Regular expression matches too often

Question

I have a string which contains certain class names + method calls. I'd like to extract these but the regex I wrote to do this returns more matches than intended.

#!/usr/bin/perl



$someInput = "Lorem ipsum  dolor sit CLASS3.aMethod.anotherMethod amet, consetetur sadipscing CLASS1.bMethod elitr, sed diam nonumy eirmod";



@matches = ($someInput =~ /((CLASS1|CLASS2|CLASS3)(\.[A-Za-z0-9_]+){1,})/g);

foreach my $match (@matches) {

    print $match . "\n";

}

Instead of one match in the form "CLASS.methodOne.methodTwo..." it also matches the class name and the methods separately. Please have a look at the example result:

CLASS3.aMethod.anotherMethod

CLASS3

.anotherMethod

CLASS1.bMethod

CLASS1

.bMethod

What I actually want:

CLASS3.aMethod.anotherMethod

CLASS1.bMethod

I would appreciate it if anyone could help me with this. A short explanation would also be great :) Thanks!

Answer 1

The regular expression

((CLASS1|CLASS2|CLASS3)(\.[A-Za-z0-9_]+){1,}))

defines 3 capture groups, each delineated by a balanced pair or parentheses:

1: ((CLASS1|CLASS2|CLASS3)(\.[A-Za-z0-9_]+){1,})
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2: ((CLASS1|CLASS2|CLASS3)(\.[A-Za-z0-9_]+){1,})
     ^^^^^^^^^^^^^^^^^^^^

3: ((CLASS1|CLASS2|CLASS3)(\.[A-Za-z0-9_]+){1,})
                           ^^^^^^^^^^^^^^^

and so when one expression is successfully matched against this expression, the result will be 3 matches.

Use the (?:...) syntax to keep an expression together as a logical unit in a regular expression but to designate it as a non-capturing group. This expression will only return one match on success:

((?:CLASS1|CLASS2|CLASS3)(?:\.[A-Za-z0-9_]+){1,})

Regular expression matches too often

Sign up