perl array from csv file creating newline in unexpected place - perl

Question

Hi I have a few scripts that convert an xlsx file to a tab seperated file, which then remove any commas, duplicates and then splits it by commas. (i do this to make sure users have not put any commas in a colomn)
I then do some stuff. and then convert it back to an xlsx file. This has always worked fine. But instead of opening and closing files all the time i thought i would push the file to an array and then convert it to an xlsx at the end. Unfortunatly when i try and convert back to an xlsx file it is creating a newline in the space between the name. If i OUTPUT to a csv file then Open it and convert to an xlsx file it works fine.
#!/usr/bin/perl
use strict;
use warnings;
use Spreadsheet::BasicRead;
use Excel::Writer::XLSX;
local $" = "'\n'";
open( STDERR, ">&STDOUT" );
#covert to csv
my $xlsx_WSD = ( "C:\\Temp\\testing_file.xlsx"),, 1;
my #csvtemp;
if ( -e $xlsx_WSD ) {
my $ss = new Spreadsheet::BasicRead($xlsx_WSD) or die;
my $col = '';
my $row = 0;
while ( my $data = $ss->getNextRow() ) {
$row++;
$col= join( "\t", #$data );
push #csvtemp, $col . "\n" if ( $col ne "" );
}
}
else {
print " C:\\Temp\\testing_file.xlsx file EXISTS ...!!\n";
print " please investigate and use the restore option if required !..\n";
exit;
}
;
my #arraynew;
my %seen;
our $Header_row = shift (#csvtemp);
foreach (#csvtemp){
chomp;
$_ =~ s/,//g;
$_ =~ s/\t/,/g;
# print $_ . "\n" if !$seen{$_}++ ;
push #arraynew, $_ . "\n" if !$seen{$_}++ ; #remove any dupes
}
#covert back to xlsx
my $workbook = Excel::Writer::XLSX->new("C:\\Temp\\testing_filet.xlsx");
my $worksheet = $workbook->add_worksheet();
my ( $x, $y ) = ( 0, 0 );
while (<#arraynew>) {
my #list = split /,/;
foreach my $c (#list) {
$worksheet->write( $x, $y++, $c );
}
$x++;
$y = 0;
}
__DATA__
Animal keeper M/F Years START DATE FRH FSM
GIRAFFE JAMES LE M 5 10/12/2007 Y
HIPPO JACKIE LEAN F 6 11/12/2007 Y
ZEBRA JAMES LEHERN M 7 12/12/2007 Y
GIRAFFE AMIE CAHORT M 5 13/12/2012 Y
GIRAFFE MICKY JAMES M 5 14/06/2007 Y
MEERKAT JOHN JONES M 9 15/12/2007 v v
LEOPPARD JIM LEE M 8 16/12/2002
unexpected result
GIRAFFE JAMES
LE M 5 10/12/2007 Y
"
HIPPO" JACKIE
LEAN F 6 11/12/2007 Y
"
ZEBRA" JAMES
LEHERN M 7 12/12/2007 Y
"
GIRAFFE" AMIE
CAHORT M 5 13/12/2012 Y
"
GIRAFFE" MICKY
JAMES M 5 14/06/2007 Y
"
MEERKAT" JOHN
JONES M 9 15/12/2007 v v
"
LEOPPARD" JIM
LEE M 8 16/12/2002

Answer 1

Since you are running this on Windows, have you considered using Win32::OLE instead?
use strict;
use Win32::OLE;
my $app = Win32::OLE->GetActiveObject('Excel.Application')
|| Win32::OLE->new('Excel.Application', 'Quit');
my $wb = $app->Workbooks->Open("C:/Temp/testing_file.xlsx");
my $ws = $wb->ActiveSheet;
my $max_row = $ws->UsedRange->Rows->Count;
my $max_col = $ws->UsedRange->Columns->Count;
my ($row, %already) = (1);
while ($row <= $max_row) {
my ($col, #output) = (1);
while ($col <= $max_col) {
my $val = $ws->Cells($row, $col)->{Text};
if ($val =~ /[,\t]/) {
$val =~ tr/,//d;
$val =~ tr/\t/,/;
$ws->Cells($row, $col)->{Value} = $val;
}
#output[$col - 1] = $val;
$col++;
}
if ($already{join "|", #output}++) {
$ws->Rows($row)->EntireRow->Delete;
$max_row--;
} else {
$row++;
}
}
$wb->SaveAs("C:\\temp\\testing_filet.xlsx");

Answer 2

This is an issue with end of line characters.
There are three conventions for marking the end of a line: with \n on Unix, \r\n on Windows and \r on Mac. It looks as though your script is assuming the Mac convention while input and output use the Windows convention.
So after reading the input, a leading \n appears on all lines except the first. As long as this is also the case with the output lines prior to composing them with \r, you end up with an output file with perfectly \r\n-delimited lines. Clearly it's better to make your script wary of what line ending convention the input is using and ensure it uses the same for splitting the lines and composing the output.