Fayland and Perl Programming

Foorum 0.2.7 release

noreply@blogger.com (Fayland) — Thu, 02 Oct 2008 03:06:00 +0000

Please download from http://foorum.googlecode.com/files/Foorum-0.2.7.tar.gz

it's a good code source to learn Catalyst+DBIx::Class+Template and others.

try http://www.foorumbbs.com/

Patches or suggestion is really appreciated.

Thanks.

My 3rd CPAN module Today: Sphinx::Control

noreply@blogger.com (Fayland) — Mon, 29 Sep 2008 11:35:00 +0000

the code and idea is directly from Lighttpd::Control and Nginx::Control.

Sphinx::Control controls Sphinx searchd.

Thanks.

two new CPAN modules

noreply@blogger.com (Fayland) — Mon, 29 Sep 2008 05:30:00 +0000

* Task::BeLike::FAYLAND

OK, it's not so useful for u I think. just ignore it.

* Catalyst::Plugin::CHI

mix the CHI and Catalyst. :)

Beijing Perl Workshop on Nov 08, 2008

noreply@blogger.com (Fayland) — Thu, 25 Sep 2008 01:56:00 +0000

It is jointly organized with the postgresql china that means there will be some postgresql talks as well. We have also secured few sponsorships to fly Jesse Vincent and two other postgresql hackers to the workshop.

If you are planning a trip to beijing in the early November, send a email to conference[at]perlchina.org to register the workshop, or even better come give a talk(20,40 minutes or lighting talk)!

From http://use.perl.org/~Qiang/journal/37527

==============

I'm not sure if I can go there, but I want to!

a busy release day today

noreply@blogger.com (Fayland) — Fri, 19 Sep 2008 05:38:00 +0000

I just kicked out 2 new modules and 1 refreshing module to CPAN.

2 new modules:
* Pod::Simple::Wiki::Googlecode
* Pod::From::GoogleWiki for http://use.perl.org/~schwern/journal/37476

and 1 refreshing module:
* Foorum

well, just enjoy!

Learn Moose 3: around

noreply@blogger.com (Fayland) — Thu, 18 Sep 2008 04:40:00 +0000

for example I want a shuffled list everytime when I can $t->shuffled_list;
I know I can use a sub to do the request. but with Moose, it's much more elegant.

the main code is like:

around 'shuffled_list' => sub {
    my $next = shift;
    
    my $ret = $next->(@_);
    $ret = [ shuffle( @$ret ) ];
    
    return $ret;
};

full source code can be found in http://fayland.googlecode.com/svn/trunk/script/learn_moose/002_around.pl

the output looks like:

1, 10, 6, 4, 2, 7, 9, 8, 3, 5
3, 4, 7, 2, 1, 9, 5, 6, 8, 10

Here we use 'around' instead of 'before' or 'after' because 'before' doesn't work according to this FAQ: http://search.cpan.org/~drolsky/Moose-0.57/lib/Moose/Cookbook/FAQ.pod#How_can_I_affect_the_values_in_@__using_before?

the example above is a case of 'deflate' as http://search.cpan.org/~drolsky/Moose-0.57/lib/Moose/Cookbook/FAQ.pod#How_can_I_get_Moose_to_inflate/deflate_values_in_the_accessor?

please read the FAQ and Cookbook for more details.

Thanks.

jrock's git repos

noreply@blogger.com (Fayland) — Wed, 17 Sep 2008 07:13:00 +0000

http://git.jrock.us/?o=age

well, it has a lot of beautiful code.

git clone git://git.jrock.us/Ernst

Thanks.

Foorum 0.2.2 release

noreply@blogger.com (Fayland) — Mon, 15 Sep 2008 05:32:00 +0000

0.2.2 Mon Sep 15 12:03:00
- use Catalyst::Plugin::PageCache '0.19' and key_maker
- split Model/Validation.pm into ResultSet and remove it
- use 'MooseX::TheSchwartz' and 'TheSchwartz::Simple' to replace 'TheSchwartz'

Please download it from http://foorum.googlecode.com/files/Foorum-0.2.2.tar.gz

Thanks.

MooseX::TheSchwartz

noreply@blogger.com (Fayland) — Thu, 11 Sep 2008 10:16:00 +0000

I just released a new version of MooseX::TheSchwartz today.

it's written on the way when I learn Moose. and most of the code is copied from TheSchwartz and TheSchwartz::Simple.
yet that's cool to write such things with Moose.

Moose rocks!

Learn Moose 2: trigger

noreply@blogger.com (Fayland) — Fri, 05 Sep 2008 05:10:00 +0000

I want a scoreboard on MooseX::TheSchwartz.
so I defined one like "has 'scoreboard' => ( is => 'rw', isa => 'Str' );", but that's not what I want for all, I want more.
I want it to be a temp file related to $$ when I set it as 1 or 'on'.
I want I can custom a temp file for it.

I know I can write a method like the one in TheSchwartz. but with Moose, it's much more elegant as what u want!

has 'scoreboard'  => (
    is => 'rw',
    isa => 'Str',
    trigger => sub {
        my ($self, $dir) = @_;
        
        return unless $dir;
        # no endless loop when it's a file
        if ($dir =~ /\/theschwartz\/scoreboard\./is) {
            # get the real dir from $dir regardless a file
            my (undef, $dir) = File::Spec->splitpath( $dir );
            unless (-e $dir) {
                mkdir($dir, 0755) or die "Can't create scoreboard directory '$dir': $!";
            }
            return;
        }

        # They want the scoreboard but don't care where it goes
        if (($dir eq '1') or ($dir eq 'on')) {
            $dir = File::Spec->tmpdir();
        }
    
        $dir .= '/theschwartz';
        unless (-e $dir) {
            mkdir($dir, 0755) or die "Can't create scoreboard directory '$dir': $!";
        }
        
        $self->{scoreboard} = $dir."/scoreboard.$$";
    }
);

now I have all I want.

just be careful, when u want to change itself in trigger like "$self->{scoreboard} = 'xxx'" or "$self->scoreboard('xxx');", u must make sure there isn't a endless loop since change itself will call trigger again.

test code can be found here: http://fayland.googlecode.com/svn/trunk/script/learn_moose/001_attribute_trigger.pl

Thanks.

Learn Moose 1: subtype

noreply@blogger.com (Fayland) — Thu, 04 Sep 2008 07:22:00 +0000

finally I wrote my first Moose script.
mainly I want to write a MooseX::Schwartz if I have enough time.

Moose is very powerful.
I want a common function like set a flag to debug, besides that, we can define debug logger sub ourself.
something like:

->verbose(1);
->verbose(0);
->verbose( sub {
    my $msg = shift;
    $msg =~ s/\s+$//;
    print STDERR "[CUS] $msg\n";
} );
# then call ->debug depends on ->verbose.

final code is here:
http://fayland.googlecode.com/svn/trunk/script/learn_moose/007_subtype.pl

I'm not a Moose guru, but Moose sounds pretty exciting!

upgrade pain

noreply@blogger.com (Fayland) — Wed, 16 Jul 2008 02:37:00 +0000

Recently I upgrade subversion to 1.5.0 on my WinXP box. but GOD, that's not working with svn+ssh at all. It's too painful to point out the plink is not working in svn 1.5.0 as before.

Sphinx 0.9.8 is out. but I think I would upgrade it after 1 week or more.

upgrade is painful.

Thanks.

tip: one line force install CPAN module

noreply@blogger.com (Fayland) — Mon, 07 Jul 2008 06:14:00 +0000

$> perl -MCPAN -e 'force("install","Foorum");'

that's it. Thanks.

3 day 3 release

noreply@blogger.com (Fayland) — Tue, 17 Jun 2008 01:55:00 +0000

CPAN generally means responsibility. I released 3 Foorum distribution in the last 3 days.

CPAN means better document. so I tried to use DBICx::AutoDoc to create docs for Foorum::Schema. it's a bit different from the original method in DBICx::AutoDoc. I changed a bit to write pod into Schema files directly.

code are:(briefly)

my $ad = DBICx::AutoDoc->new(
    schema => 'Foorum::Schema',
    output => File::Spec->catdir($Bin, '..', '..', 'docs'),
);

# rewrite the Schema pm POD
use Template;

my $tt2 = Template->new( { INCLUDE_PATH => $ad->include_path, POST_CHOMP => 0, PRE_CHOMP => 0 } );
my $vars = $ad->get_vars;

# first get the lists of all Foorum::Schema pm files
my @sources = @{ $vars->{sources} };
foreach my $source (@sources) {
    my $class = $source->{class}; # Foorum::Schema::User
    
    # make file dir
    my @parts_of_modules = split('::', $class);
    $parts_of_modules[-1] .= '.pm';
    my $file_dir = File::Spec->catfile( $Bin, '..', '..', 'lib', @parts_of_modules );
    
    my $output;
    $tt2->process('pod.html', { source => $source }, \$output)
        || die $tt2->error(), "\n";
    
    # replace POD in real module
    open(my $fh, '<', $file_dir);
    local $/ = undef;
    my $in = <$fh>;
    close($fh);
    
    my ($code, $pod) = split(/\n1;\n/, $in);
    open(my $fh2, '>', $file_dir);
    print $fh2 "$code\n1;\n__END__\n\n$output\n";
    close($fh2);
    
    print "working on $class\n";
}

Full code can be found in http://foorum.googlecode.com/svn/trunk/bin/misc/dbicx_autodoc.pl

Thanks.

Father's Day

noreply@blogger.com (Fayland) — Sun, 15 Jun 2008 08:45:00 +0000

http://cn.zorpia.com/fayland/album/1078303/album/page=all/

Today is my first Father's Day. :)

Foorum 0.1.5 release

noreply@blogger.com (Fayland) — Sun, 15 Jun 2008 07:55:00 +0000

Finally I release the Foorum 0.1.5 after 2 months later.

CHANGES:
1, Foorum::Search with backend Foorum::Search::Database and Foorum::Search::Sphinx
2, change DATETIME to INT(11) for tables.

it's a release for my little boy diudiu. :)

u can get it from http://foorum.googlecode.com/files/Foorum-0.1.5.tar.gz

and I uploaded to CPAN too.

Thanks.

use INT(11) instead of DATETIME for time

noreply@blogger.com (Fayland) — Fri, 23 May 2008 15:36:00 +0000

well, don't use DATETIME or DATE in database for time, use INT(11) UNSIGNED instead.

the benefits are:
1, u don't need care about MySQL or SQLite or others. NOW() is used by MySQL while CURRENT_TIMESTAMP is used by SQLite. that's important while u developer in MySQL and write test cases by SQLite.
2, compare. while in TT, if u want to compare topic.time with date.now, u need wrap date.now by String.Compare or convert topic.time to int. that's not so easy. and date.format can convert INT to any format u want.

I spent some hours to do the upgrade for Foorum.
and the upgrade pl is simple: create a temp column then update as UNIX_TIMESTAMP, then drop old column and rename temp column.

code is here: http://foorum.googlecode.com/svn/trunk/bin/upgrade/0.1.5/up.pl

Enjoy.

Catalyst trap: param and params

noreply@blogger.com (Fayland) — Fri, 11 Apr 2008 07:32:00 +0000

well, if we write code as follows:

sub test : Local {
   my ($self, $c) = @_;
     my $test = {
       a => $c->req->param('a'),
       b => $c->req->param('b'),
       c => $c->req->param('c'),
   };
   $c->res->body(Dumper(\$test));
}

then try to visit test?b=1, guess what we get? that's not what we expected.

$VAR1 = \{
           '1' => 'c',
           'a' => 'b'
         };

a weird result!

please use follows, that's much better.

sub test : Local {
    my ($self, $c) = @_;

    my $test = {
        a => $c->req->params->{'a'},
        b => $c->req->params->{'b'},
        c => $c->req->params->{'c'},
    };
    $c->res->body(Dumper(\$test));
}

and get what we want:

$VAR1 = \{
            'a' => undef,
            'b' => '1',
            'c' => undef
          };

so do remember that. and don't write dbix insert/update code like that.

my shell history

noreply@blogger.com (Fayland) — Fri, 11 Apr 2008 00:08:00 +0000

-bash-3.00$ history|awk '{a[$2]++ } END{for(i in a){print a[i] " " i}}' |sort -rn|head
290 perl
235 svn
67 ls
60 screen
46 vim
27 mysql
25 cp
23 rm
22 ps
22 exit

please support Perl in Google App Engine

noreply@blogger.com (Fayland) — Thu, 10 Apr 2008 04:06:00 +0000

http://code.google.com/p/googleappengine/issues/detail?id=34

star it please. so that we will have a free Perl host by Google!

I have a son

noreply@blogger.com (Fayland) — Mon, 07 Apr 2008 13:35:00 +0000

My son (nickname: diudiu 丢丢, formal name: 林锦喆) was born a week ago on April Fool's Day (4.1) and weight was 3kg. I'm pretty happy with that. I'll upload some photos later. :)

KunMing Trip

noreply@blogger.com (Fayland) — Fri, 21 Mar 2008 03:40:00 +0000

I'll be in KunMing next week with all my Zorpia colleges. it would be a nice trip. :)

memcached for Windows

noreply@blogger.com (Fayland) — Wed, 19 Mar 2008 13:13:00 +0000

http://www.splinedancer.com/memcached-win32/

have fun.

Catalyst Tip: bind STDOUT with res->body

noreply@blogger.com (Fayland) — Fri, 14 Mar 2008 10:22:00 +0000

sub test_body : Local {
    my ($self, $c) = @_;

    $c->res->body(\*STDOUT);
    print "a\n";
    print "b\n";
}

I think that's pretty useful when u are trying to use "Benchmark" module (or others use STDOUT).

Have fun!

MySQL tip TIMESTAMPDIFF

noreply@blogger.com (Fayland) — Wed, 05 Mar 2008 07:49:00 +0000

Thanks for my college CY.

UNIX_TIMESTAMP(birthday) is not working when birthday is before 1970.
I should use TIMESTAMPDIFF(DAY, '1900-01-01', birthday) to get the DAY or SECOND

mysql> select TIMESTAMPDIFF(DAY, '1900-01-01', '1998-05-01');
+------------------------------------------------+
| TIMESTAMPDIFF(DAY, '1900-01-01', '1998-05-01') |
+------------------------------------------------+
|                                          35914 |
+------------------------------------------------+

http://dev.mysql.com/doc/refman/5.0/en/date-and-time-functions.html#function_timestampdiff

in Perl, we can do:

use Date::Calc qw/Delta_Days/;
print Delta_Days(1900,1,1, 1998,5,1);
^Z
35914