Ascii "chr" function - ignore?

I've got the following script:
#!/usr/bin/perl
        my @a = ();
        while ($line = <>) {
        while (@a) {
        pop @a;
        }
        chomp $line;
        while ($line =~ m/^([\d]+)[\D]/) {
        push @a, $1;
        $line =~ s/($1).//;
        }
        push @a, $line;
        for ($x = 0; $x < $#a; $x++) { print chr($a[$x]); }
        print chr($a[$x]), "\n";
}

Which converts dot seperated ascii values like the following to plain-text
7.101.120.97.109.112.108.101 = example  (there are 7 chars in the word example)
and 7.101.120.97.109.112.108.101.10.115.101.114.118.105.99.101.49.48.49 equates to example service101 (as example is 7 chars long, and service101 is 10 chars)
The problem is that when the char_count is a "9" or a "10", in ascii those are Horizontal Tab and Line feed respectively. I'm actually very positive I won't need any ascii code below 32 (if there are any, aside from 9&10)
Actually we can eliminate the "char_count" seperator and replace it with a space in the output or as ascii before the conversion I guess.

So the output of  9.116.101.115.116.105.110.103.49.50.9.116.101.115.116.105.110.103.49.50  would convert to "testing12 testing12" rather than "testing12testing12" note the space placement.

Again I don't need any ascii value 0-31 to be translated via "chr" anything below 31 can simply be converted to a space.
9.116.101.115.116.105.110.103.49.50.9.116.101.115.116.105.110.103.49.50    (testing12 testing12)
10.116.101.115.116.105.110.103.49.50.51    (testing123)
11.100.105.100.45.105.116.45.119.111.114.10710.116.101.115.116.105.110.103.49.50.51
10.116.101.115.116.105.110.103.49.50.50.9.116.101.115.116.105.110.103.49.50
reference http://xinn.org/Css-Rosetta-Stone.html

thanks!
-rich