The code you've posted is perfectly legal (although you really need to check the return value of open: open(FILE, "filetoread") or die "Error opening file: $!\n";). I suspect your error is higher up in the code and perl just happens to error out at that line.

$ cat test1.pl 
#!/usr/bin/perl -w
use strict;

open( FILE, "filetoread.something" ) or die "can't open filetoread.som
+ething:$!\n";
while ( <FILE> ) { 
        print;
}
close(FILE);


$ cat filetoread.something 
blah blah blah
blah blah blah
blah blah blah
blah blah blah
blah blah blah
blah blah blah


$ ./test1.pl 
blah blah blah
blah blah blah
blah blah blah
blah blah blah
blah blah blah
blah blah blah

$ perl -v

This is perl, v5.8.0 built for cygwin-multi-64int
[download]

#--GETAVATARS--#
 sub getavatars {
     my $id = $_[0];
     my $filename = "./avatars/$id/avatar.dat";
     open (FILE, "$filename");
     my @avatars = <FILE>;
     close(FILE);
     return(FILE);
 }
[download]

Content-type: text/html 
ERROR

Bareword "FILE" not allowed while "strict subs" in use at ./modules/ad
+min/users.cgi line 161. Compilation failed in require at admin.cgi li
+ne 54.
[download]

Why would you want to return a closed file handle?

return(FILE);
[download]

That's the line causing the error. I'm not sure what that subroutine is supposed to do, as it is reading the file into a lexical array, local to the subroutine, and not doing anything with it afterwards. Maybe you want to:

return @avatars;
[download]

or:

return \@avatars;
[download]

?

Liz

EDIT: Oh my god, I see the error. I'm not sure why I put "return(FILE), but that's apparently what's causing it. Thanks for your help :)

Incidentally, if you really meant to return the close handle you would either have used IO::File in the first place or you'd return the glob ala return \*FILE;.

I guess you need to figure out how to go to specified line numbers in the text editor that you use for perl programming. When perl reports an error at a specific line number, the first thing to try is to go directly to that line number in the file.

sub mysub {
 my $fh = \do { local *FILE};
 open($fh, ">myfile")...
...
}
[download]

sub mysub {
 open(my $fh, ">myfile")...
...
}
[download]

use strict;
use IO::File;

my $f = new IO::File "filetoread.something", "r"
           or die "Can not open input file";
while (<$f>)
{
  # do stuff ...
}

undef $f; # close the file
          # note that you don't even have to do this
          # this is optional because the file will be
          # closed when Perl script finishes. I am just
          # forcing an early file closure here.
[download]

Why not use open? Why is it "antique"?

Roger, you've pimped the use of IO::File over open once before in Re: Removing duplicates from list yet you never state why you think IO::File is better. I'm honestly curious as to why you prefer one over the other. Is it just because IO::File uses OO syntax?

-- vek --

Since Perl 5.6.0 we can perform file operations using a scalar - which is quite convenient to pass around in function calls. For example -



open my $fh, ">output.txt"
    or die "Can't create output: $!";
print {$fh} "Hello world!\n";
close $fh;  # if don't close, then it will be closed
            # automatically when $fh is out of scope.
[download]

However the above code is not valid in version of Perl prior to 5.6.0, which I am still using on some of the production boxes. However the following will work on Perl prior to 5.6.0.

my $fh = IO::File "output.txt", "w"
   or die "Can not create output: $!";
print $fh "Hello world!\n";
undef $fh;
[download]

My preference of IO::File over open is on the Perl version compatibility and also ease of use.

This is only my personal perference however.