How about just specifying decimal precision on the sprintf:

> perl -e 'print sprintf("%.0f%", 1/30*100)'
3%
[download]

Update: Aside from sprintf formatting, if what you're really trying to do is compare two floating point numbers for Test::More, you might want to look at Test::Number::Delta, which I wrote to avoid that kind of roundabout way of comparing numbers. E.g.

use Test::More tests => 1;
use Test::Number::Delta within => .01;
my $percent = 1/30;
delta_ok( $percent, 0.03, 'values within 1%');
[download]

No no, I was happy with all three answers about dropping the 0. In fact, it was a testing situation that led me to post this question, but I am okay with my current Test::More solution, which winds up along the lines of

......code.....
ok($keyword_density >= $min_keyword_density,"Keyword density " . sprin
+tf("%2d%%",$keyword_density*100) . ", minimum " . sprintf("%2d%%",$mi
+n_keyword_density*100) );
[download]

This seems elegant enough for me, but as always I welcome suggestions.

1. As requested, a non-sprintf solution:
   $percent = int(100 * $one_thirtieth) . '%';
2. Think about what rounding behavior you need; if $f = 888/1000, then:
   int(100*$f) will truncate to 88,
   sprintf(  '%d',100*$f) will truncate to 88,
   sprintf('%.0f',100*$f) will round to 89.
3. You can merge formatting and concatenations into a single sprintf call:

   ok( 
     $keyword_density >= $min_keyword_density, 
     sprintf(
       'Keyword density %.0f%%, minimum %.0f%%',
       100 * $keyword_density,
       100 * $min_keyword_density,
     )
   );
   [download]

Here is the code I was testing with:

Read more... (1496 Bytes)