Jasmine
Administrator
Mar 28, 2001, 2:44 PM
Post #4 of 8
(644 views)
|
Holding true to TIMTOWDI (there's more than one way to do it), another way to count occurances of something in a string is by using tr///, which probably the most efficient way to go.
You can use any of the noted methods in a "normal" script, but if you're (let's say) counting the number of sentences in War and Peace, you'll probably want to use tr.
Whenever in doubt on what's the most efficient way to do something in Perl, use the Benchmark module:
#!/usr/bin/perl
use Benchmark;
$count = 500_000;
$string = 'Sentence 1. Sentence 2. Sentence 3.';
timethese($count, {
'split array' => sub {
$count = @tokens = split( /\./, $string );
},
'split into @_' => sub {
$count = split( /\./, $string );
},
'using tr' => sub {
$count = ( $string =~ tr/\.// );
},
}
);
#output
Benchmark: timing 500000 iterations of split array, split into @_, using tr...
split array: 3 wallclock secs ( 2.03 usr + 0.00 sys = 2.03 CPU)
split into @_: 1 wallclock secs ( 1.82 usr + 0.00 sys = 1.82 CPU)
using tr: 0 wallclock secs ( 0.72 usr + 0.00 sys = 0.72 CPU)
|
MAIN
SEARCH
WHO'S 
LOG 
. Please, never worry to correct me when I may be misunderstanding a question. After all, the goal is to provide assistance :)