http://www.perlmonks.org?node_id=788884


in reply to Re^3: remove line feed at the end
in thread remove line feed at the end

<ROOT> <FILE>sourcetag1</FILE> <NUMBER>00000 11111</NUMBER> <SOURCE>source1</SOURCE> <AUTHOR>author1 staff1</AUTHOR> <HEADLINE>DISPOSABLE DECOR: THE CUTTING EDGE DULLS FASTTYLE AT A SPEED USUALLY ASSOCIATED WITH WARDROBE ITEMS</HEADLINE> </ROOT> Sorry...It is not only for the line which has \ at the end. It should remove all the new line character anything which doesnot match {sometext} at the start. I mergeed the code with the sample which you provided. But it works for the line {HEADLINE} The merged code is below 
#!/usr/bin/perl
use strict;
use warnings;

my $output = '';
my $tag;
my $fh;
LINE: while ( my $line = <DATA>)  {
    chomp $line;
    # if line ended with a '\', remove '\' and save line for later out
++put
    if ( $line =~ s/\\$// || $line =~ s/\s$//) {
        # clean indentation
        $line =~ s/^\s+/ /;
        # save line
        $output .= $line;
        next LINE;
    }
    # we have previous line(s) to consider?
    if ( length $output ) {
        # clean indentation
        $line =~ s/^\s+/ /;
        $line = $output . $line;
        $output = '';
    }
  if($line =~ /^{(.*)}/) {
        $tag = $1;
    }else {
        if($tag eq 'FILE') {
                if(defined($fh)){
                print $fh "</ROOT>";
                close($fh);
                }
 my $filename = $line;
            open($fh, '>', "$filename.xml") or die "$filename: $!";
            print $fh '<?xml version="1.0"?>',"\n";
            print $fh "<ROOT>\n";
            print $fh "<FILE>$filename</FILE>\n";
        } elsif(defined($fh)) {
if($line ne ''){
$line =~ s/\\//gi;
            print "<$tag>$line</$tag>\n";
}
}
}
    print $line, $/;
}


__DATA__
{FILE}
sourcetag1
{NUMBER}
00000
11111
{SOURCE}
source1
{KEYWORD}
{AUTHOR}
author1
staff1
{HEADLINE}
DISPOSABLE DECOR: THE CUTTING EDGE DULLS FAST\
TYLE AT A SPEED USUALLY ASSOCIATED WITH WARDROBE ITEMS

{FILE}
sourcetag2
{NUMBER}
00002
{SOURCE}
sourcenam2
{KEYWORD}
{AUTHOR}
author2
staff2

[download]

Replies are listed 'Best First'.
Re^5: remove line feed at the end
by ig (Vicar) on Aug 15, 2009 at 16:27 UTC 
    use strict;
use warnings;

my ($output, $tag, $fh);

while (<DATA>) {
    chomp; s/^\s+//; s/\s+$//;
    if(/^{(.*)}$/) {    # {TAG} line
        $fh = output($output, $tag, $fh);
        $output = "";
        $tag = $1;
    } else {            # not a {TAG} line
        next unless($tag);
        next if(/^\s*$/);
        s/\\//g;
        $output .= ($output) ? " $_" : "<$tag>$_";
    }
}

$fh = output($output, $tag, $fh);
if($fh) {
    print $fh "</ROOT>\n";
    close($fh);
}
exit(0);


sub output {
    my ($output, $tag, $fh) = @_;
    if($output) {
        if($output =~ m/<FILE>(.*)/) {
            if($fh) {
                print $fh "</ROOT>\n";
                close($fh);
            }
            open($fh, '>', "$1.xml") or die "$1.xml: $!";
            print $fh "<?xml version=\"1.0\"?>\n<ROOT>\n";
        }
        print $fh "$output</$tag>\n";
    }
    return($fh);
}

__DATA__
^B^B^B^B^B^B
{FILE}
sourcetag1
{NUMBER}
00000

{SOURCE}
source1
{KEYWORD}
{AUTHOR}
author1
staff1
{HEADLINE}
DISPOSABLE DECOR: THE CUTTING EDGE DULLS FAST\
STYLE AT A SPEED
USUALLY ASSOCIATED WITH WARDROBE ITEMS.

{FILE}
sourcetag2
{NUMBER}
00002
{SOURCE}
sourcenam2
{KEYWORD}
{AUTHOR}
author2
staff2

    [download]
[reply]
[d/l]
Re^5: remove line feed at the end
by Anonymous Monk on Aug 15, 2009 at 14:14 UTC
    Your code seems that concatenating of lines is only for the last statement before the {FILE} tag. Figure out the way to solve it.
[reply]

In Section Seekers of Perl Wisdom