PERL count files in a dir

Asteroid · 07-08-2009, 03:23 AM

Hi Guys,

I need to count files in a dir which were updated yesterday.
ls -lth | grep -i 'Jul 7' | wc -l

The dir holds files of last 15 days and total count is as 2067476.
Is it efficient to count the files using perl? I have developed the following perl script making use of system().

Can anybody coment, any other way without using system()

Code:

	

#!/usr/bin/perl 
  
 
@months = (Jan,Feb,Mar,Apr,May,Jun,Jul,Aug,Sep,Oct,Nov,Dec); 
($sec,$min,$hour,$monthday,$mon,$year,$wday,$yday,$isdst) = localtime(time); 
  
  
$month =$months[$mon]; 
 
############################################# 
# ls -lth | grep -i 'Jul 7' | wc -l  
############################################# 
 
$command = "ls -lth  | grep -i '" . $month . "  " . $monthday -1 . "' | wc -l"; 
  
$count = system($command); 
print "$count \n";

Thank you
Regards
@Asteroid

PMP · 07-08-2009, 04:22 AM

Use perl built in functions opendir to open the directory and then list files

Code:

#!/usr/bin/perl -w

$dirname = "/tmp";

opendir ( DIR, $dirname ) || die "Error in opening dir $dirname\n";
while( ($filename = readdir(DIR))){
     print("$filename\n");
}
closedir(DIR);

Asteroid · 07-08-2009, 05:24 AM

Quote:

Originally Posted by PMP

Use perl built in functions opendir to open the directory and then list files

Code:

#!/usr/bin/perl -w

$dirname = "/tmp";

opendir ( DIR, $dirname ) || die "Error in opening dir $dirname\n";
while( ($filename = readdir(DIR))){
     print("$filename\n");
}
closedir(DIR);

But What about printing files from specific date. Listing file is not a problem, rather listing and counting files update on a specfic date is required here.

PMP · 07-08-2009, 05:34 AM

use perl stat function to find the mofification time of a file, filter out the files using that

http://perldoc.perl.org/functions/stat.html

Sergei Steshenko · 07-08-2009, 10:22 AM

Quote:

Originally Posted by PMP

Use perl built in functions opendir to open the directory and then list files

Code:

#!/usr/bin/perl -w

$dirname = "/tmp";

opendir ( DIR, $dirname ) || die "Error in opening dir $dirname\n";
while( ($filename = readdir(DIR))){
     print("$filename\n");
}
closedir(DIR);

Wrong - since you'll also count directories.

Also wrong because a file or directory named "0" will yield FALSE and your while loop will stop iterating.

Also wrong because no strict is used.

jan61 · 07-08-2009, 01:38 PM

Moin,

without perl:

Code:

find . -maxdepth 1 -type f -mtime -1 -print | wc -l

Jan

PMP · 07-09-2009, 12:51 AM

Thanks for the correction Sergei Steshenko.
Just to explain

Wrong - since you'll also count directories.
Just read the second post of mine which says

Quote:

use perl stat function to find the mofification time of a file, filter out the files using that

So when I am saying file it means files only not directories.

Also wrong because a file or directory named "0" will yield FALSE and your while loop will stop iterating.

This code is just to explain the opendir function, you can say it an idea of using the function.

Also wrong because no strict is used.

Well, this program compiles without complaining that strict is not being used, and up to my understanding using strict pragma is not compulsory. It is just a safe side of writing code.

Thank You

chrism01 · 07-09-2009, 02:09 AM

You'll need

Code:

  next if $file =~ /^\.\.?$/;     # skip . and ..

inside the loop to skip curr & parent dirs

micxz · 07-09-2009, 02:31 AM

How about:

Code:

#!/usr/bin/perl

# Path to directory you want to count files in on your server.
$dir = "/path/some/where/";

$directory_count = 0;
$file_count=0;

opendir(DIR, $dir);
LINE: while($FILE = readdir(DIR)) {
next LINE if($FILE =~ /^\.\.?/);

## check to see if it is a directory
if(-d "$FILE"){
$directory_count++;
}
else {
$file_count++;
}
}
closedir(DIR);

print "Directories: $directory_count\n";
print "Files: $file_count\n";

exit;

Sergei Steshenko · 07-09-2009, 06:30 AM

Quote:

Originally Posted by PMP

...
Well, this program compiles without complaining that strict is not being used, and up to my understanding using strict pragma is not compulsory. It is just a safe side of writing code.

Thank You

Of course - as long as you believe you do not make mistakes.

In production I ruthlessly eradicate Perl scripts without 'use strict;' - years of experiences have shown that real life scripts without 'use strict;' do have bugs which 'use strict;' shows.

Sergei Steshenko · 07-09-2009, 06:34 AM

Quote:

Originally Posted by micxz

How about:

Code:

#!/usr/bin/perl

# Path to directory you want to count files in on your server.
$dir = "/path/some/where/";

$directory_count = 0;
$file_count=0;

opendir(DIR, $dir);
LINE: while($FILE = readdir(DIR)) {
next LINE if($FILE =~ /^\.\.?/);

## check to see if it is a directory
if(-d "$FILE"){
$directory_count++;
}
else {
$file_count++;
}
}
closedir(DIR);

print "Directories: $directory_count\n";
print "Files: $file_count\n";

exit;

Without going into details - why do you have quotes in

Code:

"$FILE"

?

And if you wrote

Code:

if(-d $FILE)

it would still be a mistake (not because of absence/presence of quotes).

And since '.' and '..' are directories, checking for being a directory would suffice.

And yet another Perl script without warnings and strict enabled, with all the variables being global ...

PMP · 07-09-2009, 06:37 AM

Well that what I said, Its a practice that one needs to follow but not necessary. And saying that wrong is wrong in itself

micxz · 07-09-2009, 12:59 PM

Quote:

Originally Posted by Sergei Steshenko

Without going into details - why do you have quotes in

Code:

"$FILE"

I'm not sure, trial and error. It's works so I didn't think it was wrong.

Quote:

Originally Posted by Sergei Steshenko

And if you wrote

Code:

if(-d $FILE)

it would still be a mistake (not because of absence/presence of quotes).

And since '.' and '..' are directories, checking for being a directory would suffice.

And yet another Perl script without warnings and strict enabled, with all the variables being global ...

I didn't think checking for . or .. was necessary as in most cases. Hence the:
=~ /^\.\.?/);

Is this acceptable in your book:

Code:

#!/usr/bin/perl

use strict;

my $dir = "./";

my $directory_count = 0;
my $file_count=0;

opendir(DIR, $dir);
LINE: while(my $FILE = readdir(DIR)) {
next LINE if($FILE =~ /^\.\.?/);

if(-d $FILE){
$directory_count++;
}
else {
$file_count++;
}
}
closedir(DIR);

print "Directories: $directory_count\n";
print "Files: $file_count\n";

exit;