Re: There is more than one way (and mine is not the best)

Updated with a little more code and comments.

Is there some reason you are not using File::Find for this?

I have added some sample, untested code for you below. I have no idea what your code is doing, but I hope the following will help you do it better.

#!/usr/bin/perl -w

use strict;
use File::Find;

my $dir = "/Wherever my .txt files are stored";

my ($qnum, $code, $ctext, $ntext, $stext) = ('', '', '', '', '');

find(\&processfile, ("$dir/"));

sub processfile {
  if (/(.*)\.txt$/i) {
    open QIN, "$_" or die $!;
    open QOT, ">$1.txt" or die $!;
    while (<QIN>) {
      chomp;
      s/\222/'/g;
      s/\226/-/g;
      if (/^Col/i) {
        s/^col.//gi;
        s/\(net\)//gi;
        $ntext = $_;                         
        print QOT "net1$ntext (net);unl1\n";
      } elsif (/^\(S/i) { 
        (undef, $stext) = split(/\s/, $_, 2);      
        print QOT "net2$stext (subnet);unl1\n";
      } elsif (/^[0-9]/) {
        ($code, $ctext) = split(/\s/, $_, 2);           
        print QOT "n01$ctext;c=(&txt($code.ge.1)\n";
      }
    }
    close QIN;
    close QOT;
    
    open (QIN, "$1.tmp") or die $!;
    open (QOT, ">$1.axe") or die $!;
    while (<QIN>) {
      if (/n01\s+nothing|n01\s+don''t know|n01\s+other/i) {
        $_ =~ s/\n/;nz;nosort\n/g;
      }
      if (/\s+[\w;]/) {                  
        s/n01\s+/n01/g;            
        s/net1\s+/net1/g;          
        s/net2\s+/net2/g;
      }
      if (/^net1misc/i) {                 
        s/unl1/unl1;nz;nosort/g;
      }
      unless (/^n01;c=|^net. /i) {        
        print QOT $_;             
      }
    }
    close QOT;
    
    
  }
}
[download]

Comment on Re: There is more than one way (and mine is not the best) Download Code

Replies are listed 'Best First'.
Re: Re: There is more than one way (and mine is not the best) by graff (Chancellor) on May 26, 2004 at 04:43 UTC
Is there some reason you are not using File::Find for this? Well, it seems that one very good reason is that the OP doesn't show any need for recursive descent of a directory tree. All the files are in a single directory, and File::Find is overkill for that -- and somewhat harder to grok (because of all the unrelated things it does), compared to the simple glob example in the first reply, or the basic "opendir...; readdir..." functions, which are also much easier to learn than File::Find -- e.g.: `opendir( D, "." ) or die "WTF? $!"; for my $file ( grep /\.txt$/, readdir D ) { # do whatever it is the OP is doing... }` [download] For that matter, the OP didn't say one way or the other, but maybe there are subdirectories containing "*.txt" files, and he might actually prefer to leave those alone -- just process the ones in the current directory. In that case, File::Find would cause real trouble (or at least extra work to avoid trouble).	[reply] [d/l]

Replies are listed 'Best First'.

Re: Re: There is more than one way (and mine is not the best)
by graff (Chancellor) on May 26, 2004 at 04:43 UTC

Is there some reason you are not using File::Find for this?

Well, it seems that one very good reason is that the OP doesn't show any need for recursive descent of a directory tree. All the files are in a single directory, and File::Find is overkill for that -- and somewhat harder to grok (because of all the unrelated things it does), compared to the simple glob example in the first reply, or the basic "opendir...; readdir..." functions, which are also much easier to learn than File::Find -- e.g.:

opendir( D, "." ) or die "WTF? $!";

for my $file ( grep /\.txt$/, readdir D ) {
   # do whatever it is the OP is doing...
}
[download]

[reply]
[d/l]