When processing CSV files, I usually use Text::CSV to read in the first line to get the column headings and then I use that information to index into the rest of the data by column names (I know what the columns are called, but I don't know what column number they are). Currently, I use the following code, which does not feel very Perl-ish.
use warnings;
use strict;
use Text::CSV;
# setup a CSV parser
my $csvParser = Text::CSV->new({ binary => 1, sep_char => ',', empty_i
+s_undef => 1 }) or die "Cannot create new CSV parser: $!\n";
open my $csvHandle, "<", $csvFile or die "could not open $csvFile: $!"
+;
# figure out the column headings from the first line
my @csvColumns = ();
foreach my $colName (@{$csvParser->getline($csvHandle)}) {
push @csvColumns, $colName;
}
# reverse that to allow lookup by name
my %csvColumns;
for my $colNum (0 .. scalar(@csvColumns)-1) {
$csvColumns{$csvColumns[$colNum]} = $colNum;
}
# access some column in the CSV file by name
while (my $line = $csvParser->getline($csvHandle)) {
print "Column Foo has value $$line[$csvColumns{Foo}];
}
Is there a more perl-like way to reverse an array into a hash that could then be used to get the index into the data?