That doesn't look like a split file ... it looks like a substr file. And, of course, the best way to get fields from a fixed-width line is ... unpack.

while(<DATA>)
{
    my @fields = unpack "A7A6A*", $_;


    print join(':', map { sprintf "%.2f", $_ } @fields), "\n";
}

__DATA__
123.45 345.3 234.67
34.56  234.1 123.87
897.34 567.32-100.90
[download]

Of course, such a small sample of your data doesn't lend itself to perfect analysis, so the regexp's may be more suiting.