reading from an arbitrary number of files using anonymous file handles

downer has asked for the wisdom of the Perl Monks concerning the following question:

I wish to create a script that reads from n files until i have m total unique lines. In order to do this, I would like to use anonymous file handles. my code just gives me warnings which suggest that lines are not being read. can anyone give me a pointer as to what is wrong with my code?

#!/usr/bin/perl
use strict;
use warnings;
my $wantedLines = shift; # number of unique instances wanted

my $numFiles = 0;
my @files = ();
while( my $file = shift)
{
    open $files[$numFiles++], '<', $file or die "can't open $file, $!\
+n";
}

my $totalLines = 0;
my $ct = 0;
my %unique = ();

while($totalLines <= $wantedLines)
{
    my $num = $ct++;
    $num = $num%$numFiles;
    print "$num\n";

    my $this = <$files[$num]>;
    unless(exists $unique{$this})
    {
        print $this;
        $totalLines++;
    }
}
[download]

Comment on reading from an arbitrary number of files using anonymous file handles Download Code

Replies are listed 'Best First'.
Re: reading from an arbitrary number of files using anonymous file handles by JadeNB (Chaplain) on Dec 05, 2008 at 23:21 UTC
The cheap answer (cheap because there should be a better one, but I don't know it) is that `<$handle>` seems to work, but `<$array_of_handles[0]>` doesn't. Changing `my $this = <$files[$num]>;` [download] to `my $handle = $files[$num]; my $this = <$handle>;` [download] should work. By the way, you read one line too many. Your `while` test should be `$totalLines < $wantedLines`. UPDATE: Thanks to Tanktalus for pointing me to the documentation. I/O Operators has this to say: Read more... (948 Bytes) As the documentation suggests, and almut and Perlbotics point out below, `readline` will probably serve better.	[reply] [d/l] [select]
Re: reading from an arbitrary number of files using anonymous file handles by almut (Canon) on Dec 05, 2008 at 23:30 UTC
...as JadeNB said. Or use `readline`: `my $line = readline $files[$num];` [download]	[reply] [d/l] [select]
Re: reading from an arbitrary number of files using anonymous file handles by Perlbotics (Archbishop) on Dec 05, 2008 at 23:40 UTC
You can use readline instead of the <angle brackets> `... # my $this = <$files[$num]>; my $this = readline($files[$num]); ...` [download] In order to take track of unique lines, you might want to complement: `unless( ... ) { ... $unique{$this}++; ... }` [download]	[reply] [d/l] [select]
Re: reading from an arbitrary number of files using anonymous file handles by Not_a_Number (Prior) on Dec 06, 2008 at 00:11 UTC
Just for fun(?): `use strict; use warnings; my $wantedLines = shift; my %seen; while ( <> ) { print unless $seen{$_}++; last unless --$wantedLines; }` [download] Update: Sorry, I posted the wrong version. Replace the contents of the `while` loop with: `unless ( $seen{$_}++ ) { print; last unless --$wantedLines; }` [download]	[reply] [d/l] [select]
Re: reading from an arbitrary number of files using anonymous file handles by eye (Chaplain) on Dec 06, 2008 at 07:11 UTC
Although not part of the OP's question, I think it is a bad idea to open all the files in advance of reading from any. This makes the code more fragile than it needs to be.	[reply]