Two inconsistent hashes from same input text file in Perl - perl

Question

I've created a hash that lists each unique word from a text file and the number of times each word appears. The resulting output from this hash indicates it has read the entire file correctly.
However, later efforts to parse text from the same file appear to only capture some of the contents.
For illustrative purposes, a second hash designed to capture each word from the infile as a hash value and that word's relative ordering in the document as the hash key captures only a small fraction of all the words in the document.
Any insights as to the cause of this phenomenon?
#!/usr/bin/perl
use strict;
use warnings;
my $path = "U:/Perl";
chdir($path) or die "Cant chdir to $path $!";
# Starting off computing a simple word count for each word in the 10-K.
my %hash;
my $word;
my %words;
while (<>) {
my #words = split;
# Here creating an index of each word in the 10-K and the number of times
# it appears. This hash works correctly.
foreach my $i (0 .. $#words) {
my $word = $words[$i];
++$hash{$word};
# Here trying to create a hash where each word from the infile is a value,
# and the word's order in the doc is the key.
$words{$i} = $word;
}
}
# The code below simply sorts each hash and prints it to an external file.
my #keys = sort { "\L$a" <=> "\L$b" } keys %words;
open(my $fh2, '>', 'report2.txt');
foreach (#keys) {
print $fh2 "$_ \t $words{$_}\n ";
}
close $fh2;
#keys = sort {
"\L$hash{$a}" <=> "\L$hash{$b}" or
"\L$a" cmp "\L$b"
} keys %hash;
open(my $fh3, '>', 'report3.txt');
foreach (#keys) {
print $fh3 "$_ \t $hash{$_}\n ";
}
close $fh3;

Answer 1

I presume you're talking about the %words hash? You are keying that hash by the position of each word on each line, so it will only ever have as many entries as the longest line in the file has words.
When processing this data
a b c d e f
g h i j k
l m n o
p q r
s t
u
your program will build %hash with 21 elements, with the keys a to u and all the values equal to 1 as they are all different.
But the hash %words will have six elements -- the number of words in the longest line -- and the values will be overwritten by the last word at that position in the line. So your hash will look like
{ 0 => 'u', 1 => 't', 2 => 'r', 3 => 'o', 4 => 'k', 5 => 'f' }
as the last word in the first position on the line is u, the last one in the second position is t, etc.
Surely you don't want a hash indexing the words across the whole file? That would be an array!
Update
Thank you for explaining your intention. I don't think putting all the words from the file into a hash one by one is going to help you to count all the unique six-word sequences.
In any case, using numeric keys from 1 to N, where N is the number of words in the file, is misusing a hash, and as I intimated above what you really want is an array that is meant to be indexed by integers.
I think you should keep an array that holds the current six-word sequence. If you add each word to the end of the array and drop them from the beginning then it will always hold the most recent six words from the file.
Something like this, perhaps
use strict;
use warnings;
my #sequence;
my %sequences;
while (<DATA>) {
for (split) {
push #sequence, $_;
if (#sequence >= 6) {
shift #sequence while #sequence > 6;
++$sequences{"#sequence"};
}
}
}
use Data::Dump;
dd \%sequences;
__DATA__
a b c d e f
g h i j k
l m n o
p q r
s t
u
output
{
"a b c d e f" => 1,
"b c d e f g" => 1,
"c d e f g h" => 1,
"d e f g h i" => 1,
"e f g h i j" => 1,
"f g h i j k" => 1,
"g h i j k l" => 1,
"h i j k l m" => 1,
"i j k l m n" => 1,
"j k l m n o" => 1,
"k l m n o p" => 1,
"l m n o p q" => 1,
"m n o p q r" => 1,
"n o p q r s" => 1,
"o p q r s t" => 1,
"p q r s t u" => 1,
}