CODEWITHSUNDEEP
regexperl
Jahnavi
Jahnavi

Reputation: 11

perl regex to handle and preserver single and multiple words into a variable

I am writing a perl script to read the full name of a member and save it to variables firstname and lastname like below:

my ($firstname, $lastname) = $member =~ m/^(\w+.*?) +(\w+)$/;
my $member_name = $firstname.' '.$lastname;

The value for $member comes from an upstream service which would be like for example "Jane Doe" Now the code above cannot handle when the service sends $member value like "Jane". The regex fails to handle a single word in that code line. I need it to handle both multiple and single words. I cannot implement a new code functionality so I am looking to add to the existing regex so that there is minimal change and that it can handle both the conditions.

So far this is what I am testing with in the command line but so far no luck:

perl -e 'my ($firstname, $lastname) = "Jane Doe" =~ m/^(\w+.*?) +(\w+)$/|m/^(\w+)$/; print "$firstname\n$lastname";'

When I substitute "Jane Doe" with "Jane", nothing prints. I want the code to be in this format though. like if the value is multiple words it should print them both, otherwise just the single word.

Your help will be greatly appreciated.

Upvotes: 0

Views: 46

Answers (1)

simbabque
simbabque

Reputation: 54323

There is a syntax error in your Perl code. You terminated the pattern too early.

#    /                 /   /      /   
#                      V
    m/^(\w+.*?) +(\w+)$/|m/^(\w+)$/

This will lead to the | being interpreted as a bit-wise or. Since there's another m// behind it, the | will take the return values of both m// operations and do its magic. The second m// will just match against the topic $_.

What you actually want is to merge both patterns. 

my ($firstname, $lastname) = "Jane Doe" =~ m/^(?:(\w+.*?) +)?(\w+)$/;

You need to make the first name optional with a non-capture group (?:), followed by a ? none-or-one quantifier.

You cannot have three capture groups, like you probably intended, because the third one would go to $3, and not $1.

However, the above solution uses the last name, which you then assign to the $firstname variable. Your full name pattern allows for names with any characters in them, like Jean-Luc Picard. But if you pass in just Jean-Luc, the match will fail. So if you want only the first name, you should use the correct pattern to make it consistent.

A simple way of doing that is to make the last name optional instead.

my ($firstname, $lastname) = "Jane" =~ m/^(\w+.*?)(?: +(\w+))?$/;

Remember that this will set $lastname to undef, which doesn't matter so much in your command line example, but in a proper program with strict and warnings (which you of course have turned on, right?) it will complain if $lastname is used as a string while it's undef.

I suggest you read this article about names.

Upvotes: 1

Related Questions