fishbot_v2 has my vote for best solution within the constraints. However, if you were doing this repeatedly, using index might be an interesting approach. I've coded fishbot_v2's solution into the regex sub below. I did several runs and pasted a pretty typical result.
use Benchmark qw/:all/;
sub regex {
return ($_[1] =~ m/^$_[0]/) ? 1 : 0;
}
sub index_func {
return (index($_[1], $_[0]) == 0) ? 1 : 0;
}
cmpthese ( 500000, {
'regex match' => sub { regex('abbr', 'abbreviation') or die },
'index match' => sub { index_func('abbr', 'abbreviation') or die }
+,
'regex miss' => sub { regex('anna', 'abbreviation') and die },
'index miss' => sub { index_func('anna', 'abbreviation') and die
+},
});
__END__
Rate regex match regex miss index miss index match
regex match 680272/s -- -13% -36% -38%
regex miss 781250/s 15% -- -27% -29%
index miss 1068376/s 57% 37% -- -3%
index match 1101322/s 62% 41% 3% --
As you can see, using index for this is much faster, regardless of whether you are more likely to match or miss.
|