Hi,
at some other forum we had the discussion which is the fastest way to read all files of a directory
- opendir $DIR, $folder; while(my $file = readdir $DIR) ...
- foreach my $file (glob "$folder/*") ...
- opendir $DIR, $folder; my @files = grep { ... } readdir $DIR; foreach my $file (@files) ...
i prefered 1 over 3 because i thought 1 must be faster since 3 will walk all files (at least) twice.
so i made a benchmark:
#!/usr/bin/perl
use strict;
use warnings;
use Benchmark qw(:all);
my %folders = (
Temp1 => 10,
Temp2 => 100,
Temp3 => 1000,
Temp4 => 5000,
Temp5 => 10000,
# Temp6 => 20000,
);
foreach my $folder (sort keys %folders) {
print "Folder: $folder, Files: ", $folders{$folder}, "\n";
timethese(5, {
'Way 1' => sub {
opendir(my $DIR, $folder) or die "Error: couldn't open dir
+ '$folder': $!\n";
while(my $file = readdir $DIR) {
my $filename = "$folder/$file";
if(-f $filename) {
# print STDERR "$filename\n";
}
}
closedir($DIR);
},
'Way 2' => sub {
foreach my $file (glob "$folder/*") {
my $filename = $file;
if(-f $filename) {
# print STDERR "$filename\n";
}
}
},
'Way 3' => sub {
opendir(my $DIR, $folder) or die "Error: couldn't open dir
+ '$folder': $!\n";
my @files = map { "$folder/$_" }
grep { -f "$folder/$_" }
readdir($DIR);
closedir($DIR);
foreach my $file (@files) {
my $filename = $file;
# print STDERR "$filename\n";
}
},
});
}
the result does not make sence to me, but maybe you can tell me why
Folder: Temp1, Files: 10
Benchmark: timing 5 iterations of Way 1, Way 2, Way 3...
Way 1: 0 wallclock secs ( 0.00 usr + 0.00 sys = 0.00 CPU)
(warning: too few iterations for a reliable count)
Way 2: 0 wallclock secs ( 0.00 usr + 0.00 sys = 0.00 CPU)
(warning: too few iterations for a reliable count)
Way 3: 0 wallclock secs ( 0.00 usr + 0.00 sys = 0.00 CPU)
(warning: too few iterations for a reliable count)
Folder: Temp2, Files: 100
Benchmark: timing 5 iterations of Way 1, Way 2, Way 3...
Way 1: 0 wallclock secs ( 0.00 usr + 0.05 sys = 0.05 CPU) @ 108.70/s (n=5)
(warning: too few iterations for a reliable count)
Way 2: 0 wallclock secs ( 0.00 usr + 0.08 sys = 0.08 CPU) @ 63.29/s (n=5)
(warning: too few iterations for a reliable count)
Way 3: 0 wallclock secs ( 0.03 usr + 0.00 sys = 0.03 CPU) @ 161.29/s (n=5)
(warning: too few iterations for a reliable count)
Folder: Temp3, Files: 1000
Benchmark: timing 5 iterations of Way 1, Way 2, Way 3...
Way 1: 1 wallclock secs ( 0.19 usr + 0.63 sys = 0.81 CPU) @ 6.15/s (n=5)
Way 2: 1 wallclock secs ( 0.41 usr + 1.11 sys = 1.52 CPU) @ 3.30/s (n=5)
Way 3: 1 wallclock secs ( 0.16 usr + 0.67 sys = 0.83 CPU) @ 6.04/s (n=5)
Folder: Temp4, Files: 5000
Benchmark: timing 5 iterations of Way 1, Way 2, Way 3...
Way 1: 9 wallclock secs ( 1.20 usr + 7.86 sys = 9.06 CPU) @ 0.55/s (n=5)
Way 2: 19 wallclock secs ( 1.73 usr + 16.59 sys = 18.33 CPU) @ 0.27/s (n=5)
Way 3: 10 wallclock secs ( 0.92 usr + 7.91 sys = 8.83 CPU) @ 0.57/s (n=5)
Folder: Temp5, Files: 10000
Benchmark: timing 5 iterations of Way 1, Way 2, Way 3...
Way 1: 43 wallclock secs ( 3.17 usr + 39.39 sys = 42.56 CPU) @ 0.12/s (n=5)
Way 2: 85 wallclock secs ( 6.53 usr + 76.41 sys = 82.94 CPU) @ 0.06/s (n=5)
Way 3: 42 wallclock secs ( 3.19 usr + 38.06 sys = 41.25 CPU) @ 0.12/s (n=5)
-
Are you posting in the right place? Check out Where do I post X? to know for sure.
-
Posts may use any of the Perl Monks Approved HTML tags. Currently these include the following:
<code> <a> <b> <big>
<blockquote> <br /> <dd>
<dl> <dt> <em> <font>
<h1> <h2> <h3> <h4>
<h5> <h6> <hr /> <i>
<li> <nbsp> <ol> <p>
<small> <strike> <strong>
<sub> <sup> <table>
<td> <th> <tr> <tt>
<u> <ul>
-
Snippets of code should be wrapped in
<code> tags not
<pre> tags. In fact, <pre>
tags should generally be avoided. If they must
be used, extreme care should be
taken to ensure that their contents do not
have long lines (<70 chars), in order to prevent
horizontal scrolling (and possible janitor
intervention).
-
Want more info? How to link
or How to display code and escape characters
are good places to start.