Anyway: PDF is based on Postscript. In PostScript, the string delimiters are parens. That's the cause of the problems, and why it doesn't break with brackets and braces. The way around it, is to escape the parens, any parens, if PDF supports it, with a backslash. The backslash itself must be special, too.

Let me pull in the PDF docs, from PDFReference.pdf:

3.2.3 String objects — literal strings
Any characters may appear in a string except unbalanced parentheses and the backslash,which must be treated specially. Balanced pairs of parentheses within a string require no special treatment.

That should be enough to get you on your way, I suppose. You can check the table at page 30 from the above PDF file (warning: huge file: 9MB), that is the 50th page in the document.

p.s. It seems that quite a few people are rather fond of this module, including myself. If the original author is no longer interested in supporting it, perhaps somebody else should pick it up. How does one go about, in order to do that? (BTW the best place to escape the backslashes and the parens, would be in the module.)

Great Job Bart.. could you post your additions to add color? Sounds great :)

could you post your additions to add color? Sounds great :)

Sure. Eventually, I'd like to get it on CPAN, but I'm not sure how long it can still take, so in the meantime, here's my add-on module for PDF::Create, which I gave the name PDF::Create::Extensions. All you need is this extra file, which adds a few methods to PDF::Create::Page, and they're available as soon as you do

use PDF::Create;
use PDF::Create::Extensions;
[download]

You can create two kinds of colour, "linecolor" (for lines) and "fillcolor" (for area fills). For each, you can set the color in the same way, in RGB ("RGB" or "RG": red, green, blue, each value ranging from 0 to 1, less is darker), grayscale ("GRAY", "GREY" or "G", 1 value from 0 (white) to 1 (black); and CMYK ("CMYK" or "K", 4 values from 0 to 1, more is darker), like this:

$page->linecolor(RGB => 0.6, 0.2, 0.2);
[download]

The color type is case insensitive.

The second thing you can set, is line width, line connection style ("join": styles "miter", "round", "bevel"), and line end style ("cap": styles "butt", "round", "square" (="projecting square")), like this:

$page->linewidth(0.8, 'join' => 'miter');
[download]

See the PDF documentation on what the possibilities represent, or just experiment with these options. Again, the names are case insensitive.

Save the next code as PDF/Create/Extensions.pm somewhere in your @INC

Read more... (2 kB)

use strict;
use Regexp::Common;

my @tests = (
  'foo(bar)',
  'foo bar',
  'foo(bar',
  'foo bar)',
  'foo((bar)',
  'foo(bar))',
  'foo((bar)))',

);

for(@tests) {
  printf "%12s", $_;
  
  s{ [()] }( )xg and tr/ //s
   unless m/ (?<! [(]) $RE{balanced}{-parens => '()'} (?! [)] ) /x;
  
  printf " - %s\n", $_;
}


__output__

    foo(bar) - foo(bar)
     foo bar - foo bar
     foo(bar - foo bar
    foo bar) - foo bar 
   foo((bar) - foo bar 
   foo(bar)) - foo bar 
 foo((bar))) - foo bar
[download]

_________
broquaint

abowley@krait:~$ (echo 'Madrid(Spain' && echo 'Madrid(Spain)') | perl 
+-pe 's/\((?![^)]*\))/ /g;'
Madrid Spain
Madrid(Spain)
[download]

s/
  \(      # an open bracket
  (?!     # _not_ followed by
    [^)]* # anything except brackets (e.g. 'Spain')
    \)    # followed by a bracket
  )
 / /gz;
[download]

This works if you are sure that you won't have any nested parentheses. If you do then it doesn't work at all, for example:

a(b(c))           becomes a b(c))
a (b (c) d (e) f) becomes a  b (c) d (e) f)

Well, you can do that with a regex, but it's a good degree more complicated.....

abowley@krait:~$ (echo 'Madrid(Spain(Europe)' && echo 'Madrid(Spain(Eu
+rope))' && echo 'Madrid(Spain Europe))' && echo 'a(b(c))' && echo 'a 
+(b (c) d (e) f)') | perl -mstrict -wpe 'BEGIN { $brackets = qr#\([^()
+]*(?:(??{$brackets})[^()]*)*\)# }; s#([^)]+?)($brackets)([^(]+)# ( my
+ $s = $1 ) =~ tr/(/ /; ( my $e = $3 ) =~ tr/)/ /; $s . $2 . $e #ge;'
Madrid Spain(Europe)
Madrid(Spain(Europe))
Madrid(Spain Europe) 
a(b(c))
a (b (c) d (e) f)
[download]

This code will correct the gross imbalance of parens without requiring to delete them all. As is, it works line by line, but would operate on a whole file if you slurped it.

#! perl -slw
use strict;

while(<DATA>) {
    my $count=0;
    chomp;
    print;
    $count += $1 eq '(' ? +1 : -1 while m[([()])]g ;
    s[(^.*?)\(]    [$1]   while $count-- > 0;
    s[(.*)\)(.*?$)][$1$2] while ++$count < 0;
    print;
}
__DATA__
test (test)
test (test
(test (test))
(test (test)
(() (()) (((()))))
( () (()) ( ( () ) ) ))
(( () (()) ( ( () ) ) )
((( test
test )))
(((test)
(test)))
[download]

All it does is count the number of opens and closes and delete enough of whichever is in excess from the beginning or end to correct any imbalance.

In most cases, that probably means that the resultant parenthesizing would not reflect the original intent, but it is difficult to see how to determine what that intent was without specific knowledge.

#!/usr/bin/perl -w
use strict;
use Text::Balanced qw (extract_bracketed);
my $text = "In (Madrid) (Spain";

my ($extracted, $remainder, $skipped);
my $newtext = "";
for (;;) {
   ($extracted, $remainder, $skipped) = extract_bracketed($text,'()','
+[^\(]*');
   if ($extracted) {
      $newtext .= $skipped;
      $newtext .= $extracted;
      $text = $remainder;
   } else {
      if ($@->{error} =~ /Unmatched opening/) {
         $remainder =~ s/\(/ /;
         $text = $remainder;
      } elsif ($@->{error} =~ /Did not find opening bracket/) {
         $newtext .= $remainder;
         last;
      } else {
         die "Unexpected text balance error: $@";
      }
   }
}

print "Now text is: *$newtext*\n";
[download]