UNIX for Dummies Questions & Answers

Hi I would like to sort a csv file. It has 50 fields and approx 1400000 lines. I want to sort by three columns as follows. Say sort on coulmn 5, if entries are equal sort on column 3 if this is also equal sort on column 6.

It seems that this is possible for 5 then 3 i.e.

Code:

sort -t "," -k 5,5 -k 3,3 <filename>

but it doesn't seem to work for 5 then 3 then 6 i.e.

Code:

sort -t -k 5,5 -k 3,3 - k 6,6 <filename>

(that is assuming this is doing what I think).

I know in perl the subroutine to do this would be, this works for smaller files but when I try to run my large file it crashes due to memory issues.

Code:

sub tableSorter($$){
    my ($row1,$row2) = @_;
 
    my $column6comparison = $row1->[5] cmp $row2->[5];
    if($column6omparison != 0) {
        return $column10comparison;
    }
 
    my $column3omparison = $row1->[3] cmp $row2->[3];
    if($column3comparison != 0) {
        return $column3comparison;
    }
 
    return $row1->[6] <=> $row2->[6];
 
}

my questions
Would I neeed to write a subroutine to do this ?

or can I do it in a single sort line ?

Any help would be greatly appreciated

Check you man page for "sort". The "-T pathname" option allows you to specify a larger filesystem that sort can use. When sorting larger files you can sometimes run out of space in the default filesystem sort uses for sorting temp space.