On Mar 26, 12:27 pm, [EMAIL PROTECTED] (Tom Phoenix) wrote:
> On Wed, Mar 26, 2008 at 8:18 AM, <[EMAIL PROTECTED]> wrote:
> > I have two sorted files (one string per line).
> > [I'd also like to know how to sorvle this if the lists weren't sorted
> > (as complimented sets)].]
> > I want to output the List1 items not found in the List2 file.
> > grep is too slow.
> > diff gets stuck because list2 has millions of items.
>
> If the lists aren't sorted, it's probably best to read the second list
> (the list of filters) into a hash. But since they're sorted, and
> because you have many filters, it's more efficient to read the files
> in parallel.
>
> My first draft of this program used this line to implement the inner loop:
>
> $current_filter = <FILTERS> while $item gt $current_filter;
>
> ... but then I realized that the second file could run out of filters
> before the first one runs out of data, so it had to become more
> complex:
>
> #!/usr/bin/perl
>
> use strict;
> use warnings;
>
> die "huh?" unless @ARGV == 2;
> my($data_file, $filters) = @ARGV;
>
> open DATA_FILE, $data_file or die "Can't read '$data_file': $!";
> open FILTERS, $filters or die "Can't read '$filters': $!";
>
> my $current_filter = '';
>
> # outer loop reads a line at a time
> DATA_LINE:
> while (my $item = <DATA_FILE>) {
>
> # inner loop updates the filter, if needed
> # This inner loop would be just this line:
> ### $current_filter = <FILTERS> while $item gt $current_filter;
> # .... except that we have to allow for the filters to run out.
> while ($item gt $current_filter) {
> if (defined($current_filter = <FILTERS>)) {
> # a filter was read from the file: normal case
> } else {
> # No more filters; print everything else
> print $item;
> print while <DATA_FILE>;
> last DATA_LINE;
> }
> }
>
> # the inner loop has now updated $current_filter
> print $item unless $item eq $current_filter;
> }
>
> Hope this helps!
>
> --Tom Phoenix
> Stonehenge Perl Training
Works great, thanks.
One more thing if I may:
How do I mod the code to function as is with two args (perlscr
list1.txt list2.txt)
or accept stdin as data_file when only one arg is given? (cat
list1.txt | perlscr list2.txt)
Thanks Again
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
http://learn.perl.org/
