pump a large file of many Gigs to the other side of the wall

[Anton van der Steen]

Hey fellow Perl addicts,

Since long I struggled with the problem to pump large files of many many
Gigs to a different place on a server or netwerk.

Well it took some research, but here is the solution:

1.use perl
2. rename the extension of the file to something exoctic like sex. This is
by far out an unkowm extension not used by Mikrosoft.
3. Use script see below and look at the packages of perl that have to be
installed, it is only one namely File::Find:

#!perl/bin/perl

use File::Find;


my @files = find(\&cleanup, "e:\\");

sub cleanup {


if ((/\.sex$/) ) #|| (/\.fun$/)
{
my $doc_1 = substr($File::Find::name, 0, 2 );
my $doc_2 = substr($File::Find::name, 4 );
my $doc_3 = join "",$doc_1,"/",$doc_2;


open FILE ,"<$doc_3" or die $!;

my $file_out = "e:/dump7/resultaat.txt";
open(OUT,">>$file_out");

while (<FILE>)
{

print OUT ;
}
close FILE;

}
}

4. Eureka !!! a file has born on the other side off the wall.

Otherwise you have to buy expensive tools which can do this,
but no Perl comes to the rescue.

5. Have fun.

Anton Vandersteen
Eindhoven/The Netherlands

 

[DJ Stunks]

Hey fellow Perl addicts,

I have a feeling you're addicted to something other than Perl.

-jp

 

[Tad McClellan]

3. Use script see below

I don't think that that is a very good idea.

my $doc_3 = join "",$doc_1,"/",$doc_2;

Why the obfuscation?

my $doc_3 = "$doc_1/$doc_2";

open FILE ,"<$doc_3" or die $!;

Good. You check the return value.

open(OUT,">>$file_out");

Bad. You didn't check the return value.

 

[Ala Qumsieh]

Since long I struggled with the problem to pump large files of many many
Gigs to a different place on a server or netwerk.

I might be overlooking something, but how is this different from 'copy'?

--Ala

 

[news reader]

Hi Anton,


I'm lost, what you try to achieve, but however some questions /
comments. (assuming you're not trolling)


1.) Why do you search the whole drive e: for your file suffix and thus
potentially multiple files? You'd really like to find all of them?
Couldnt't you just specify the filename(s) on the command line or drag
and drop them onto your perl script icon, which should (though I never
tried myself) pass the file names to @ARGV.

2.) Do you really always want to append to your result file?

3.) How come, that your source and destination file are on different
'sides' of the 'wall' (whatever it means) if source and destination are
located under 'e:'

4.) Why do you once write 'e:\\' and once write 'e:/' ?

5.) Do you know for 100%, that the source files are always text files?
If not and you have an unfortunate files without cr/lf characters your
copy loop might slurp in the entire source file in one shot, which may
explode your memory as you are talking about files of a size of multiple GB.

6.) Why do you close FILE, but not OUT?

7.) What is this 'wall' and the 'side'?

8.) Any context could help to see, why and how perl made the world a
better place for you.

9.) Are you really sure, that there's really no 'exotic' application
with the file suffix '\.[e][x]$' . There are so many weird
applications out there. Who knows? ;-)

10.) Don't you expect, that many members of the newsgroup will filter
out messages containing your choice of suffix in a perl related news group?

11.) Just out of curiousity: What would be the expensive tool, that
would perform the same task?


tot ziens and have fun


N.

 

[axel]

Anton van der Steen wrote:

I might be overlooking something, but how is this different from 'copy'?

The script would not work with binary files.

Axel

 

[John Bokma]

2. rename the extension of the file to something exoctic like sex.
This is by far out an unkowm extension not used by Mikrosoft.

So Microsoft left you out the beta test of Microsoft Porn 1.0?

 

[John Bokma]

So Microsoft left you out the beta test of Microsoft Porn 1.0?

<http://filext.com/detaillist.php?extdetail=SEX>