Comments on: Day 3 – File operations

By: markel

markel — Wed, 25 May 2011 17:55:01 +0000

good tut!
thanks!

By: Moritz

Moritz — Mon, 13 Dec 2010 14:25:15 +0000

Note that the speed disadvantage comes from doing all IO operations (and keeping the result in memory) before any processing can happen.

In Perl 6, this is not the case. The IO operations happen as the list items are accessed. If the list is not consumed, no internal calls to readline() will be made.

Lazy lists are good for something after all :-)

By: Cd-MaN

Cd-MaN — Mon, 13 Dec 2010 14:13:45 +0000

ok, so the comment system ate my “less than” and “greater than” signs. I will replace them with [ and ]:

for my $file ([*]) {
…
}

By: Cd-MaN

Cd-MaN — Mon, 13 Dec 2010 14:12:40 +0000

opendir/readder has a speed advantage in Perl 5 when there are many files/directories in a certain tree. Ie. the following code:

opendir D, ‘foo’;
while (my $file = readdir D) {
…
}
closedir D;

is faster and more memory efficient than the following (prettier looking) version:

for my $file () {
}

Is there some kind of mechanism (like lazy lists) in Perl 6 to have the implementation perform the optimal way?

By: Moritz

Moritz — Sat, 04 Dec 2010 08:55:24 +0000

It excludes . and .. by default. Which means it saves you from writing a grep on your own in most use cases. Or put another way, it’s a convenience.

By: Shawn M Moore

Shawn M Moore — Sat, 04 Dec 2010 06:37:20 +0000

I’m curious – why implement grep again as the “test” parameter? Does it do anything better?

By: Jake

Jake — Sat, 04 Dec 2010 06:08:35 +0000

You should use glob then

By: Moritz

Moritz — Fri, 03 Dec 2010 09:27:29 +0000

Right, you can tell it was specced (and implemented in Rakudo) by somebody who has written lots of Perl 5 code :-)

By: Steven Haryanto

Steven Haryanto — Fri, 03 Dec 2010 09:25:29 +0000

Reading the above, I think it’s nice that dir() by default skips “.” and “..”, because in Perl 5 I always do this: next if /^\.\.?$/;