Question

I’m a beginner in Perl. Could someone help me on how to extract data correctly from the script below?

#####################################################################
#! /usr/bin/perl
$text = "Name: Anne Lorrence Name: Burkart Name: Claire Name: Dan" ;
$match = 0 ;
while ($text =~ /Name: \b(\S+)\s+(\S+)\b/g || /Name: \b(\S+)\b/g) {
    ++ $match ;
print "Match number $match is $1 $2\n" ;
}
######################################################################

I wanted my output be something like this:

Match number 1 is Anne MLorrence
Match number 2 is Burkart
Match number 3 is Claire 
Match number 4 is Dan

but in fact, my script gives me this:

Match number 1 is Anne MLorrence
Match number 2 is Burkart Name

May I know what is going wrong?

Answer 1

$text = "Name: Anne Lorrence Name: Burkart Name: Claire Name: Dan" ;
$match = 0 ;
while ($text =~ /Name: (.+?)(?= Name:|$)/g) {
    ++ $match ;
    print "Match number $match is $1\n" ;
}

It uses a non-greedy capture and a zero-width positive lookahead to delimit the fields.

Match number 1 is Anne Lorrence
Match number 2 is Burkart
Match number 3 is Claire
Match number 4 is Dan

The |$) part is an alternate. An easier example to understand would be (ABC|DEF), which means “match either ‘ABC’ or ‘DEF'”. The $ is simply the symbol for end-of-line.

The zero-width positive lookahead is explained in the perlre docs, but I’ll try to summarize here. It’s part of a class of patterns called “Look-Around Assertions”, and the name is quite accurate. Imagine the regex engine “looking around” at the point in the string. The one employed here “looks ahead” in the string for a positive match. It’s called zero-width because it doesn’t consume any of the string in the pattern matching process.

So, the pattern /Name: (.+?)(?= Name:|$) says:

1. Match “Name: “
2. Match, and capture, as little as possible
3. Until you see that the following characters are, either, ” Name:” or the EOL

There are probably better ways of solving your task, but this is short and clear and gives you an insight into some lesser used parts of the regex language. Look-Arounds are extremely useful and well worth learning about.