G'day fishy,

I work with biological data that's often of comparable size — I have a 2GB CSV file that I use for volume testing. I would never attempt to slurp an entire file of this size; instead, I would read and process line-by-line. Apart from the huge memory overhead, you're reading the data twice: once to slurp; again to process.

I don't have Raku available. Here's a quick-and-dirty example in Perl5. I've mostly used the same variable names as your code; hopefully, a conversion to Raku would not be too difficult for you.

Input CSV:

$ cat dummy.csv
a,b,c
d,e,f
g,h,i
[download]

Perl5 script column_extract.pl:

#!/usr/bin/env perl

use strict;
use warnings;
use autodie;
use constant TARGET_COL => 1;

use Text::CSV;

my $file_name = 'dummy.csv';
my @columns;
my $csv = Text::CSV::->new();
open my $fh, '<', $file_name;

while (my $row = $csv->getline($fh)) {
    push @columns, $row->[TARGET_COL];
}

print "@columns\n";
[download]

Output:

$ ./column_extract.pl
b e h
[download]

$ cat test.raku
#!raku

use Slang::Tuxic;
use Text::CSV;

say "All of it";
my @aoh1 = csv (in => "test.csv", out => Hash);
dd @aoh1;

say "Fragment col 2";
my @aoh2 = csv (in => "test.csv", out => Hash, fragment => "col=2");
dd @aoh2;

$ cat test.csv
"No.","Name"
"1","hello"
"2","bye"

$ raku test.raku
All of it
Array @aoh1 = [{:Name("hello"), "No." => "1"}, {:Name("bye"), "No." =>
+ "2"}]
Fragment col 2
Array @aoh2 = [{:Name("hello")}, {:Name("bye")}]
[download]

If you want a ; as separator, just add sep => ";".

You are allowed to use other ways to write it. Just read the docs.

I don't use Rakudo. But the number of lines hints to me that (maybe) you should start looking into using a database.

You didn't really tell us what you are trying to accomplish (the "big picture"), only that you are working with budget data. I'm assuming here that the general rule of "the data amount will exponentially grow over time" holds true for your project as well. A modern relational database like PostgreSQL, MySQL, Oracle or even (oh god oh god, i don't believe i'm saying this) Microsoft SQL Server (please kill me!) could serve you better in the long run than doing your data crunching/reporting/whatever directly on some CSV file.

my @dta = csv(in => "file.csv");
[download]

Thank you everybody for your comments!

Here are the whole use case with script and sample input data.

Greetings.