Re^2: Radoteur

Great many thanks for your input, I have started rewriting my code before even reading all your post, so here is my take on your advice:

#!/usr/bin/perl

#this is a perl implementation of the radoteur from the book "Théorie 
+du Bordel
#Ambiant" authored by the french researcher Roland Moreno
#It will take a word list as input and output giberrish-y words from i
+t, after
#a while it will cycle, this program has been specially written to ide
+ntify
#the relationship between the wordlist and the size of the cycles
#next thing would be to do some stat works on the word size, their num
+bers
#inside the wordlist to find the equation that models the word list - 
+cycle
#size the most accurately

use strict;
use warnings;
use autodie;
my($evt,$niter);
my $cycled = 0;
$niter= 0;

#this is a double hash, the first level keys are line numbers, the sec
+ond
#its values are another hash which keys are letter coordinates inside 
+the 
#line's word, if we have the same letter in the same word on the same 
+line
#appear twice then we have cycled
my %letter_lines;
if(!defined $ARGV[0]){
    die "to use me, give me the word list file as argument\n";
}
while($cycled == 0){
    open my $fh, '<',$ARGV[0];
    FILE_ITER: while(<$fh>){
            #here I take in the new word and split it into its letters
            #I also count the iteration number
        my $word = $_;
        $niter++;
        my @letters = split("", $word);

            #if evt is undefined, that's because we were looking for a
+ newline
            #in the previous event or its the first iteration, in any 
+case
            #the new event will be the first letter of the current wor
+d
        if(!defined($evt)){

            if(defined $letter_lines{$.}{0}){
                            warn "we have cycled: doing line $. letter
+ 0
                            again\n";
                $cycled = 1;
                last FILE_ITER;
            }

            else{
                $letter_lines{$.}{0} = 1;
                            warn "adding to hash line $. letter 0\n";
            }
            $evt = shift @letters;
            print $evt;
            next FILE_ITER;
        }
        else{
            WORD_ITER:foreach my $i (0 .. $#letters){
                if($evt eq $letters[$i]){
                    my $j = $i+1;
                    $evt = $i < $#letters?$letters[$j]:undef;
                    if(defined $evt){
                        if(defined $letter_lines{$.}{$j}){
                            warn "we have cycled :
                            line $. letter $j \n";
                            $cycled = 1;
                            last FILE_ITER;
                        }
                        else{
                            warn "adding to hash line $.
                            letter $j\n";
                            $letter_lines{$.}{$j} = 1;
                            print $evt;
                        }
                    }
                    last WORD_ITER;
                }
            }
        }
    }
    close($fh);
    warn "going back to the beginning\n";
}
warn "cycled in $niter iterations\n";
[download]

Comment on Re^2: Radoteur Download Code

Replies are listed 'Best First'.
Re^3: Radoteur by stevieb (Canon) on Oct 19, 2015 at 14:46 UTC
Your loop labels are not required here (and rarely ever are). You can remove them in both the loops, and the `last` and `next` calls. For example, replace: `WORD_ITER:foreach(...)` [download] with: `foreach(...)` [download] and: `next WORD_ITER; last WORD_ITER;` [download] with just: `next; last;` [download] You can do this in both your `while` and `for` statements/blocks.	[reply] [d/l] [select]
Re^4: Radoteur by QuillMeantTen (Friar) on Oct 20, 2015 at 08:57 UTC
I know they are not needed, yet I read in Perl Best Practices that whenever last or next are used adding labels helps will the readability and future maintanability of the code It seemed to make sense so I adopted the practice. update: typo	[reply]
Re^5: Radoteur by BrowserUk (Patriarch) on Oct 20, 2015 at 09:08 UTC
I read in Perl Best Practices that whenever last or next are used adding labels helps will the readability If you are prepared to pay a 15% performance penalty for that asinine piece of advice; continue: `cmpthese -1,{ a=>q[ LOOP:for(1..1000){ next LOOP; } ], b=>q[ for(1..1000){ next; } ] };; Rate a b a 7862/s -- -14% b 9195/s 17% --` [download] With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday' Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". I knew I was on the right track :) In the absence of evidence, opinion is indistinguishable from prejudice.	[reply] [d/l]
Re^6: Radoteur by QuillMeantTen (Friar) on Oct 20, 2015 at 12:18 UTC
Re^7: Radoteur by Athanasius (Archbishop) on Oct 21, 2015 at 09:06 UTC


Perl: the Markov chain saw
	PerlMonks