Parse CSV based on random header layout perl - regex

Question

I am trying some code using Tie::Handle::CSV, however I suppose this could be done using other modules or none at all.
What I want to do is take a file of random layout and match it to what I predict the headers will be. Then I want to arrange it to my table structure.
DATA
First Name,last name,date of birth
Jim,Johansen,08/25/1989
OR
2nd Name,1st Name,D.O.B
Johansen,Jim,08/25/1989
OR
2nd Name,1st Name,D.O.B,city,county
Johansen,Jim,08/25/1989,milwaukee,N/A
As you can see, I have varying data structures. I want it to arrange them to my schema, even if the fields I request are empty. The way I want to do this is by searching my input with my column variables.
Here is what I am trying.
Code
use Tie::Handle::CSV;
my $name1 =qr/First Name|Name|1st Name/i;
my $name2 =qr/Last Name|Maiden Name|2nd Name/i;
my $date_of_birth =qr/date of birth|D.O.B/i;
my $city =qr/city|town/i;
my $csv_fh = Tie::Handle::CSV->new('list.txt', header => 1);
while (my $csv_line = <$csv_fh>)
{
print $csv_line->{'$date_of_birth'}.",".$csv_line->{'$name1'}." ".$csv_line->{'$name2'}.",".$csv_line->{'$city'}.\n"; ##note I am searching for the column {$'colummn regex'} instead of {'column'} to see if my input file matches any of the header options.
}
close $csv_fh;
My output is blank since this module is not understanding the regex I am implimenting. However, my output would contain the columns specified if I used their literal names, i.e.
The out put I want would be:
Scenario 1
Date of Birth,Name,City ##my implemented header
08/25/1989,Jim Johansen, ##noting also that if there is no 'city' in the input data, leave blank.
Scenario 2
Date of Birth,Name,City ##my implemented header
08/25/1989,Jim Johansen,
Scenario 3
Date of Birth,Name,City ##my implemented header
08/25/1989,Jim Johansen,milwaukee
Perhaps there is a better option than a module or even my regex variables. Has anyone had to parse csvs in ever changing layouts?

Answer 1

You never use $name1, $name2, etc, much less in a match (or substitution) operator, so you ever execute any regex match.
my $field_names = $csv_fh->header();
my ($name1_header) = grep /First Name|^Name$|1st Name/i, #$field_names;
my ($name2_header) = grep /Last Name|Maiden Name|2nd Name/i, #$field_names;
my ($dob_header ) = grep /date of birth|D\.O\.B/i, #$field_names;
my ($city_header ) = grep /city|town/i, #$field_names;
my #recognized_fields = ( $name1_header, $name2_header, $dob_header, $city_header );
my %recognized_fields = map { $_ => 1 } #recognized_fields;
my #other_headers = grep !$recognized_fields{$_}, #$field_names;
while (my $row = <$csv_fh>) {
my $name1 = $name1_header ? $row->{$name1_header} : undef;
my $name2 = $name2_header ? $row->{$name2_header} : undef;
my $dob = $dob_header ? $row->{$dob_header } : undef;
my $city = $city_header ? $row->{$city_header } : undef;
my #other_fields = #$row{#other_headers};
...
}