search for string in file using an aray elements perl - arrays

Question

I have a array which contains set of unique elements my_array= [aab, abc def, fgh,]
I have a file which containing these elements(repeated also)
I want to count each unique element has how many repetitions if no repetition then count is 1
example of file :
i want to have aab but no i dont want abc
i want to have aab but no i dont want def
output should be
aab - 2
abc - 1
def - 1
I tried to search first and print it its not woking
use strict;
use warnings;
my #my_array;
#my_array =("abc", "aab", "def");
open (my $file, '<', 'filename.txt') or die;
my $value;
foreach $value (#my_array) {
while(<$file>) {
if ($_ =~ /$value/){
print "found : $value\n";
}
}
}
Also tried 2nd method
use strict;
use warnings;
my #my_array;
#my_array =("abc", "aab", "def");
open (my $file, '<', 'filename.txt') or die;
while (<$file>) {
my $k=0;
if ($_ =~ /$my_array[$k]/) {
print "$my_array[$k]”;
}
}

Answer 1

Sample input data does not specify if lookup words repeat in the line or not.
Following demo code assumes that lookup words do not repeat in the line.
If this statement above does not true then the line should be split into tokens and each token must be inspected to get correct count of lookup words.
use strict;
use warnings;
use feature 'say';
use Data::Dumper;
my(%count,#lookup);
#lookup =('abc', 'aab', 'def');
while( my $line = <DATA> ) {
for ( #lookup ) {
$count{$_}++ if $line =~ /\b$_\b/;
}
}
say Dumper(\%count);
exit 0;
__DATA__
i want to have aab but no i dont want abc
i want to have aab but no i dont want def
Output
$VAR1 = {
'aab' => 2,
'abc' => 1,
'def' => 1
};

Answer 2

I'm a fan of the Algorithm::AhoCorasick::XS module for performing efficient searches for multiple strings at once. An example:
#!/usr/bin/env perl
use warnings;
use strict;
use Algorithm::AhoCorasick::XS;
my #words = qw/abc aab def/;
my $aho = Algorithm::AhoCorasick::XS->new(\#words);
my %counts;
while (my $line = <DATA>) {
$counts{$_}++ for $aho->matches($line);
}
for my $word (#words) {
printf "%s - %d\n", $word, $counts{$word}//1;
}
__DATA__
i want to have aab but no i dont want abc
i want to have aab but no i dont want def
outputs
abc - 1
aab - 2
def - 1
The $counts{$word}//1 bit in the output will give you a 1 if that word doesn't exist in the hash because it wasn't encountered in the text.

Answer 3

Can build an alternation pattern from the keywords and so match all that are on the line in one regex run, then populate a frequency hash with the matches
use warnings;
use strict;
use feature 'say';
use Data::Dumper;
my #keywords = qw(aab abc def fgh);
my $re_w = join '|', #keywords;
my %freq;
while (<>) {
++$freq{$_} for /($re_w)/g
}
say Dumper \%freq;
The <> operator reads line by line the files with names given on the command line, so the program is used as prog.pl file. (Or open the file "manually" in the program.)
The for loop imposes list context on its expression, so that regex returns the list of matches (captures), as the match operator does in the list context, and the ++$freq{$_} expression works with them one at a time.
The code counts all instances of keywords that repeat on a line. If that's not desired please clarify (can add a call to List::Util::uniq before feeding the list of matches to the for loop).
There are a number of other details that may need closer attention.
One example: if there are overlapping keywords, which one takes precedence? For instance, with keywords the and there, once the word there is encountered in the text should it be matched by there or by the? If it is there then keywords in the alternation pattern should be ordered from longest to shortest,
my $re_w = join '|', sort { length $b <=> length $a } #w;
Please clarify if there are additional considerations.