Splitting regex matches by newlines in perl - regex

Question

I am trying to glob files from a directory and print out regexp matches,
Trying to match
<110>
*everything here*
<120>
My matches would be
SCHALLY, ANDREW V.
CAI, REN ZHI
ZARANDI, MARTA
However when i try to split this by newline and join using "|", I am not getting the desired output that is
Applicant : SCHALLY, ANDREW V. | CAI, REN ZHI | ZARANDI, MARTA
My current output is only
| ZARANDI, MARTA
Can someone see any obvious mistakes?
#!/usr/bin/perl
use warnings;
use strict;
use IO::Handle;
open (my $fh, '>', '../logfile.txt') || die "can't open logfile.txt";
open (STDERR, ">>&=", $fh) || die "can't redirect STDERR";
$fh->autoflush(1);
my $input_path = "../input/";
my $output_path = "../output/";
my $whole_file;
opendir INPUTDIR, $input_path or die "Cannot find dir $input_path : $!";
my #input_files = readdir INPUTDIR;
closedir INPUTDIR;
foreach my $input_file (#input_files)
{
$whole_file = &getfile($input_path.$input_file);
if ($whole_file){
$whole_file =~ /[<][1][1][0][>](.*)[<][1][2][0][>]/s ;
if ($1){
my $applicant_string = "Applicant : $1";
my $op = join( "|", split("\n", $applicant_string) );
print $op;
}
}
}
close $fh;
sub getfile {
my $filename = shift;
open F, "< $filename " or die "Could not open $filename : $!" ;
local $/ = undef;
my $contents = <F>;
close F;
return $contents;
}
EDIT 1
I Ran Code on a single file
#!/usr/bin/perl
use warnings;
use strict;
use IO::Handle;
my $input_file = "01.txt-WO13_090919_PD_20130620";
my $input_path = "../input/";
my $whole_file = &getfile($input_path.$input_file);
if ($whole_file =~ /[<][1][1][0][>](.*)[<][1][2][0][>]/s ) {
print $1;
my #split_string = split("\n", $1);
my $new_string = join("|", #split_string) ;
print "$new_string \n";
}
sub getfile {
my $filename = shift;
open F, "< $filename " or die "Could not open $filename : $!" ;
local $/ = undef;
my $contents = <F>;
close F;
return $contents;
}
Output
Chen, Guokai
Thomson, James
Hou, Zhonggang
Hou, Zhonggang

Answer 1

Replace
$whole_file =~ /[<][1][1][0][>](.*)[<][1][2][0][>]/s ;
if ($1) {
with
if ($whole_file =~ /[<][1][1][0][>](.+)[<][1][2][0][>]/s) {
The problem with your original code is that $1 is unchanged (i.e. retained from the previous file) if the regexp doesn't match.
If that doesn't solve the problem, then double check and make sure that you have the correct value if $applicant_string. Your join + split line looks correct.

Answer 2

I run your code and get
|SCHALLY, ANDREW V. |CAI, REN ZHI| ZARANDI, MARTA
Which is pretty close. all you need to do is trim whitespace before you join. So replace this
my #split_string = split("\n", $1);
my $new_string = join("|", #split_string) ;
With this:
my #split_string = split("\n", $1);
my #names;
foreach my $name ( #split_string ) {
$name =~ s/^\s*(.*)\s*$/$1/;
next if $name =~ /^$/;
push #names, $name;
}
my $new_string = join("|", #names);

Answer 3

#pts is correct, the regex capture variables do not get reset to UNDEF
upon a negative match, looks like they retain their last value.
So his solution should work for you. Use the if ( $whole_file =~ // ) {} form.
Beyond that, you could clean up the operation a little by doing something like this
use strict;
use warnings;
$/ = undef;
my $whole_file = <DATA>;
if ( $whole_file =~ /[<][1][1][0][>](.*)[<][1][2][0][>]/s )
{
my $applicant_string = $1;
$applicant_string =~ s/^\s+|\s+$//g;
my $op = "Applicant : " . join( " | ", split( /\s*\r?\n\s*/, $applicant_string) );
print $op;
}
__DATA__
<110>
SCHALLY, ANDREW V.
CAI, REN ZHI
ZARANDI, MARTA
<120>
Output:
Applicant : SCHALLY, ANDREW V. | CAI, REN ZHI | ZARANDI, MARTA