Fetching unique owner name

[Anonymous Monk]

On my Solaris 7 unix.
I am trying to fetch a unique name of html or htm owners. Basically get names of all our web authors and a total count.

From my "ls -l" I just need the owner of the html or htm file and get a person's name just once where I dont see two of the same names. For example if I get these names for all html or htm owners:

Jones
Jones
Smith
Smith
Richards

I only need unique names such as:

Jones
Smith
Richards

My attempt to do it in Perl:

#!/usr/local/bin/perl -w
open STDERR, "> /dev/null" || die "Cant redirect standrd error: $!";
use File::Find;
use strict;
print "Web owners\n";
print "----------\n\n";
my $ct = 0;
sub search
{
   if( $_ =~ /\.html?$/)
   {
       $datetime = scalar(localtime(time));
       my $name = $File::Find::name;
       open ( F, $name ) or die "$!: $name\n";
           while(my $owner = <F>)
             {
             ($ownerid, $groupid) = (stat( "$owner" ))[4];
             $ownerName = getpwuid($ownerid);
                 {
                 $ownerName{$1} = 1;
                 }
             } 
  close F;
   }
}
my $dirs = '/thisdirectory/directory';

find( \&search, $dirs));

foreach (keys %ownername)
{
      print "$_\n";
}  
print "$ct\n";

[JaWi]

Hi,

update: Actually tested the code and refined it a bit...
update2: Crafted a better perl solution, not calling perl in Perl...

Couldn't you just do something like:

 my ( @temp, %authors );
 @temp = sort map { my @b=split; $b[2]; } `ls -l`;
 foreach ( @temp )
 {
   $authors{$}=1 if not $authors{$_} and length $_;
 }
 print "$_\n" for keys %authors;'

IIRC, uniq is a standard unix command, which is also available under Solaris 7.

Greets,

-- JaWi

"A chicken is an egg's way of producing more eggs."

[Anonymous Monk]

HOw would this work? Am I in the right direction??

         @temp = '/mydirectory/directory';
         my ( @temp, %authors );
         @temp = sort map { my @b=split; $b[2]; } `ls -l`;
         foreach ( @temp )
         {
           $authors{$}=1 if not $authors{$_} and length $_;
         }
         print "$_\n" for keys %authors;'

[JaWi]

Hi,

Sorry, I (wrongly) assumed you were an advanced Perl monk. I'll explain my solution step by step:

my ( @temp, %authors );
@temp = sort map { my @b=split; $b[2]; } `ls -l /path/to/authors`;

(NB: note the added path, I forgot that in my original answer!)
You should read this from right to left: each line of the output of the ls command is split and a map is created of each second element (ie. the user names). The array returned by map is sorted and assigned to @temp.
The array @temp now contains all authors of the files in your directory, including doubles!

foreach ( @temp )
{
  $authors{$}=1 if not $authors{$_} and length $_;
}

This will loop through @temp and make a hash with the author names as key. The value is only stored if it doesn't exists already and has a length of > 0.
I do this to ensure the hash only contains unique names.

print "$_\n" for keys %authors;

This loops through the keys of the hash %authors and prints them. To get an array of the unique authors of the files in your directory you would do something like my @unique_authors = keys %authors;

Hope this helps,

-- JaWi

"A chicken is an egg's way of producing more eggs."

[Callum]

mirod's post on Re: creating unique list from elements in an array is an excellent description of what you need to do -- far better than my dodgy first attempt!