How to grep word from file - regex

Question

I want to grep some word inside a file from another file. My code is able to grep the word on last line of the file but not the word before it. I have no idea why and hope can get help here. Below is the perl script i using:
open(FILE1,"file1.txt") or die "Error, File1 could not open\n";
open(FILE2,"file2.txt") or die "Error, File2 could not open\n";
open(FILE3, ">file3.txt") or die "Error, File3 could not open\n";
use strict;
use warnings;
use List::MoreUtils qw(uniq);
my #file1=<FILE1>;
my #file2=<FILE2>;
my $j =0;
my $i =0;
my $zone =0;
for ($j=0; $j<=$#file2; $j++){
$zone = $file2[$j];
unless ( $zone =~ m/#(.*?)/ ) {
print "$zone";
my #fid = grep /$zone/ , #file1;
#fid = uniq(#fid);
s{^\s+|\s+$}{}g foreach #fid; #cancel leading space
for ($i=0; $i<=$#fid; $i++){
print FILE3 "$fid[$i]\n";
}
##fid=();
}
}
close(FILE3);
My file1.txt is something like this:
i am a dog
i am a cat
we are the fish
he is a boy
she is a girl
My file2.txt is like this:
is
am
But my file3 can only show those sentence contain am but no is, if i put is in second line and am in first line then my file3 only contain the sentences with is. I not very sure why my code can only grep the last row in my file2. Thanks for the help.

Answer 1

When reading from a file, the final newline is part of each line read. You can remove the newlines from the pattern array by chomping:
chomp( my #file2 = <FILE2> );

Answer 2

You can already do this with egrep :
egrep -f file2.txt file1.txt

Answer 3

The root of this problems is chomp - you're not removing linefeeds, so the matches aren't working.
But aside from that, there's a few problems with your code that could do with addressing:
opening files, you should use 3 arg open with lexical file handles, as it's better style: open (my $file1, '<', 'file1.txt' ) or die $!;
rather than a loop of loops, you'd probably be better off compiling up a 'match regex'.
Instead of reading all of a file into an array, you can iterate line by line, and you don't need to use the memory.
If you're iterating a loop, and only using the index to acccess the current element, you're far better off using foreach my $line ( #things ) { type syntax.
So your code actually could be reduced to something like:
#!/usr/bin/env perl
use strict;
use warnings;
open(my $data, '<',"file1.txt") or die $!;
open(my $search, '<', "file2.txt") or die $!;
open(my $output, '>', "file3.txt" ) or die $!;
chomp ( my #search_terms = <$search> );
#quotemeta is needed to avoid 'special' regex characters doing things.
my $search_regex = join "|", map { quotemeta }, #search_terms;
#note - '\b' denotes word boundary, which may not be what you want.
#means 'is' won't match 'fish'
#so get rid of them if that's not what you want.
$search_regex = qr/\b($search_regex)\b/;
print "Using: $search_regex\n";
select $output; #default print destination
while ( <$data> ) {
print if m/$search_regex/;
}
Output (in 'file3.txt'):
i am a dog
i am a cat
he is a boy
she is a girl

Answer 4

please try this .
use strict;
use warnings;
use List::MoreUtils qw(uniq);
open(FILE1,"file1.txt") or die "Error, File1 could not open\n";
open(FILE2,"file2.txt") or die "Error, File2 could not open\n";
open(FILE3, ">file3.txt") or die "Error, File3 could not open\n";
my #file1=<FILE1>;
my #file2=<FILE2>;
my $j =0;
my $i =0;
foreach my $main_line(#file1){
chomp($main_line);
foreach my $line(#file2){
chomp($line);
if ($main_line =~ /$line/i) {
print FILE3 "$main_line\n";
}
}
}
close(FILE3);
thanks,
praveenzx~