1: This code came from my desire to use perl for something
   2: besides CGI, which is my current perl-y focus.
   3: I expected slightly different behavior from
   4: File::Find::file, though this may be my Win background
   5: shining through.
   6: Put simply, this will traverse a filesystem, and determine
   7: the # of files based on extention. Again, this was more to
   8: keep the programming muscles limber; I plan on using this as a shell for more complex things. I do appreciate any
   9: comments you have, however.
  10: <code>
  11: #use warnings;
  12: use strict;
  13: use File::Find;
  14: 
  15: my %hash;
  16: my @temp;
  17: 
  18: File::Find::find (\&wanted,"c:\\");
  19: my @keys = keys (%hash);
  20: @keys = sort_by_value (\%hash);
  21: 
  22: 
  23: print "-"x20,"\n";
  24: foreach my $key(@keys) {if ($key =~ /^$/) {print "no ext\t:\t"}else{  print "$key\t:\t"}; print $hash{$key},"\n"}
  25: 
  26: sub sort_by_value {
  27:     my %hash = %{shift()};
  28:     sort {$hash {$a} <=> $hash{$b}  } keys %hash;
  29: }
  30: 
  31: 
  32: sub wanted {
  33: my @file ;
  34: my $fname;
  35:     #   don't attempt to read a directory.
  36:     if (-d $_) {return};
  37:     $fname = $File::Find::name;
  38:     #   turn any '/' into a '\'
  39:     #   then make sure that any '\\' get turned into '\'
  40:     $fname =~ s#\/#\\#g;
  41:     $fname =~ s#\\\\#\\#;
  42:     @temp= split  /\./, $_;
  43:     #   mmm. autovivification
  44:     $hash{@temp[1]}++
  45: }
  46: </code>

Replies are listed 'Best First'.
Re: file ext counter.
by chromatic (Archbishop) on Nov 23, 2000 at 10:35 UTC
    While I applaud your use of references to create complex data structures, you can save some memory and speed by operating on the hash directly: 
    sub sort_by_value {
    my $hash = shift;
    sort {$hash->{$a} <=> $hash->{$b}  } keys %$hash;
}

    [download]
    I won't ask why you declare @keys and then immediately clobber it. :) (I often forget to delete test code, too.)
[reply]
[d/l]
Re: file ext counter.
by lemming (Priest) on Nov 22, 2000 at 21:17 UTC
    This line:
      use warnings;
  use strict;
  use File::Find;

  my %hash;
  my @temp;
 
  my $dir = shift @ARGV;
  die "Give me a directory to search\n" unless (-d "$dir");
  File::Find::find (\&wanted,"$dir");
 
  print "-"x20,"\n";
  foreach my $key (sort keys %hash) {
    print $key, "\t:\t",$hash{$key},"\n";
  }
 
sub wanted {
#   don't attempt to read a directory.
  return if (-d $_);
  my $fname = $File::Find::name;
  @temp = split(/[\/\\]/, $fname);
  $fname = $temp[-1];
# Make sure there is an extension. We check for a
# trailing period even though on Win32 it shouldn't happen
  if (($fname =~ /\./) && !($fname =~ /\.$/)) {
    $fname = lc($fname);
    @temp= split(/\./, $fname);
    $hash{$temp[-1]}++;
  }
  else {
    $hash{"no ext"}++;
  }
}

    [download]

    Update:
    I Changed a few more things around. I went right for the filename since a "." in a directory could mess things up as well.
    $array[-1] so that we don't have troubles with files like "blah.txt.sh".
    And we now account for no extension at all.
    This code could make use of File::Basename as well.
[reply]
[d/l]
[select]
      ah, I see why you'd want to use  $hash{@temp[-1]}++ but this exhibits bad behavior for files with no extention at all! update

      your 

        if (($fname =~ /\./) && !($fname =~ /\.$/)) {
    $fname = lc($fname);
    @temp= split(/\./, $fname);
    $hash{$temp[-1]}++;
  }
  else {
    $hash{"no ext"}++;
  }

      [download]
      seems a little clumsy. consider 
          if ($#temp >0)
        {$hash{@temp[-1]}++}
    else
        {$hash{""}++}

      [download]
      which sees if there's more than 1 bucket in @temp. If there is, we know the lastmost one's the extention. If there isn't we know the filename has no extention. in fact, I think this can be shortened to <code> if ($#temp) <code> but that's not a big deal. I kept the hash named "" for the highly special case of a file, e.g. "foo.no ext"
[reply]
[d/l]
[select]

Back to Craft