Something like this will do?

#!/usr/bin/perl -w

use strict;

sub checkit {
    my $file = shift;
    return (-e $file && -w _);  # the _ is a shortcut for checking on 
+the same file instead of making another stat call
}

my $ret = checkit("/home/monk/hello");
if ($ret) { print "File exists with write perm\n" }
else { print "File does not exist or no write perm\n"; }
[download]

will let you know shortly why your code does not work. In the mean time check out perldoc -f -X

Update: First off why do you need to check for exists and write perm? -w would take care of that.

Why do you think your -e,-w works?  if (-e,-w $file) does not do  if (-e $file && -w $file) The comma operator just returns the last value evaluated. So it basically will do -w $file and tell you the answer. It might work out but if you turn on warnings you will notice -e works on uninit value as it is looking for $_. If you have set $_ elesewhere it makes things worse as it is not doing what you wanted

 $_[0] && $_[1] in your if statement just checks if the values  $_[0] and $_[1] for true/false. It does not evaluate the expression. Besides  if ($_[0] $_[1]) is not a valid syntax

Hope this helps!

cheers

SK

The following does what you would like, although I'd not recommend this as anything other than a learning exercise. In particular if (-e $filename && -w _) is more efficient than hitting on the file system for each test.

use strict;
use warnings;

my $filename = "c:\\myfile.txt";

checkit ("-e,-w", $filename)
    ? print "File $filename exists and is writeable"
    : print "File $filename does not exist or is not writeable";

sub checkit
{
my ($tests, $filename) = @_;
my @testList = split ",", $tests;

for (@testList)
    {
    return 0 if ! eval ("$_ '$filename'");
    }
    
return 1;
}
[download]

Note that eval is used to perform the actual test and split is used to generate a list of tests to be performed.

More mods to Grandfather's code to avoid the ugly eval:

sub checkit {
    my ($tests, $filename) = @_;
    my %TestCode=(
        "-e" => sub{-e $_[0] } ,
        "-w" => sub{-w $_[0] } ,
        "-d" => sub{-d $_[0] } ,
    );
    my @testList = split ",", $tests;
    for (@testList){
       exists $TestCode{$_}       or return 0;
       $TestCode{$_}->($filename) or return 0; 
    }
    return 1; # Passed all tests
}
[download]

(Untested)

Yes, as others have shown, more optimization is possible to avoid multiple stat calls for the same file - the point of THIS post was to avoid the eval.

     "Man cannot live by bread alone...
         He'd better have some goat cheese and wine to go with it!"

Just 2c,

On unix, the same test on a directory name will indicate you can create a file in the directory. You need write permission on the directory inode to insert a new file's metadata. You may want to test -d where -e is tested above, just to verify it's a directory.

I don't know whether the win32 file tests emulate that or not.

After Compline,
Zaxo

Modified GrandFather's code a bit so that you don't have to loop through to test

sub checkit {
    my ($tests, $filename) = @_;
    my @testList = split ",", $tests;
    my $teststr  = shift(@testList) . " '" . $filename . "'" . 
                   " && " . join (' _ && ', @testList) . " _";
    return eval ("($teststr)");
    return 1;
}
[download]

Not very clean but at least it saves you too many stat calls.