Just for fun, I want to post a script that does without the loops. So with map/ no while or for. What do you think?

undef $/;                        #just read the whole STDIN in once:
$_=<>;

tr/ \t\n\r//d;                   #remove all spaces
map{ $count{ $_ }++;} split //;  #the actual count
foreach $key (sort(keys %count)) #OK, one loop to print
{
  print "$key\t$count{$key}\n";
  $total += $count{$key};
}
print "Total\t$total\n";
[download]

Well, I tested it, and this is the tail of the output, run with time perl .... (too lazy for benchmarking):

ó       18
ô       93
õ       12
ö       16
÷       12
ø       25
ù       2
ú       3
û       1
ü       2
ý       6
þ       2
ÿ       6
0.26user 0.05system 0:00.62elapsed 49%CPU (0avgtext+0avgdata 0maxresid
+ent)k
0inputs+0outputs (275major+797minor)pagefaults 0swaps
[jeroen@rulffk] time: 11:49:31 1 jobs  wd: ~ (12.011Mb)
$ ls -l test.txt
-rw-rw-r--   1 jeroen   jeroen     150271 Aug 17 11:40 test.txt
[jeroen@rulffk] time: 11:50:04 1 jobs  wd: ~ (12.011Mb)
[download]

And it looks to me like a 150kb file that has been counted in just 0.62 s. Looks pretty impressive to me. I don't consider myself to be a decent perl coder, so I would appreciate any comments to this....

Have fun,

Jeroen

I was dreaming of guitarnotes that would irritate an executive kind of guy (FZ)

PP: I just took a .txt file, but it wasn't a text-file after all. Maybe some word (ugh!) document or whatever.

Update Removed square brackets from tr///d, as pointed out by merlyn. And changed a funny typo (s/while/whole/). Justifies my remark on decent coding, I'm afraid :-<. Another update: added the total number.