As Thick As Two Short Planks

Last week I released an update for one of my older modules, Test::DatabaseRow. In the course of this update I completely rewrote the guts of the module, turning the bloated procedural module into a set of clearly defined Moose-like Perl classes.

Why update now?

Test::DatabaseRow had been failing it’s tests since Perl 5.13.3 (it was another victim of the changing stringification of regular expressions breaking tests.) We’re currently planning to upgrade our systems at work from 5.12 to 5.14 in the new year, and (embarrassing) one of the key modules that breaks our 5.14 smoke is Test::DatabaseRow. Oooops.

Since I had my editor open, I decided it might be a good idea to switch to a more modern build system. And, since I was doing that, I thought it might be a good idea to fix one of my long standing todos (testing all rows returned from the database not just the first.)

In other words, once I’d started, I found it hard to stop, and before I knew it I had a reasonably big task on my hands.

The decision to refactor

When I first wrote Test::DatabaseRow back in 2003, like most testing modules of the time, it sported a simple Exporter based interface. The (mostly correct) wisdom was that simple procedural interfaces make it quicker to write tests. I still think that’s true, but:

• Procedual programming ends up either with very long functions or excessive argument passing. The single function interface made the internals of Test::DatbaseRow quite difficult – to avoid having one giant function call I ended up passing all the arguments a multitude of helper functions and then passing the multiple return values of one function onto the next.

• Many of the calls you write want to share the same defaults For example, the database handle to use, if we should be verbose in our output, should we do utf–8 conversion?… These are handled reasonably well with package level variables as defaults for arguments not passed to the function (which isn’t such a big deal in a short test script) but the code to support those within the testing class itself isn’t particularly clean having to cope with defaults evaluation in multiple places.

• Only being able to return once from the function is a problem. Sometimes you might want to get extra data back after the test has completed. For example, when I wanted to allow you to optionally return the data extracted from the database I had to do it by allowing you to pass in the args to row_ok references to variables to be populated as it executes. Again, while this isn’t the end of the world from an external interface point of view, the effect it has on the internals (passing data up and down the stack) is horrible.

For the sake of the internals I wanted things to change. However: I didn’t want to break the API. I decided to split the module into two halves. An simple external facing module that would provide the procedural interface, and an internal object orientated module that would allow me to produce a cleaner implementation.

No Moose, but something similar

As I came to create Test::DatabaseRow::Object I found myself really wanting to write this new class with Moose. Now, Moose is a very heavyweight dependency; You don’t want to have to introduce a dependency on hundreds of modules just because you want to use a simple module to test your code. In fact, Test::DatabaseRow has no non-core dependencies apart from DBI itself, and I wanted to keep it this way with the refactoring. So, no Moose. No Mouse. No Moo. Just me and an editor.

In the end I compromised by deciding to code the module in a Moose “pull accessor” style even if I didn’t have Moose itself to provide the syntax to do this succinctly.

The approach I took was to put most of the logic for Test::DatabaseRow::Object – anything that potentially changes state of the object – into lazy loading read only accessors. Doing this allowed me to write my methods in a declarative style, relying entirely on the accessors performing the calculation to populate themselves the first time they’re accessed. For example. Test::DatabaseRow::Object has a read only accessor called db_results which goes to the database the first time it’s accessed and executes the SQL that’s needed to populate it (and the SQL itself comes from sql_and_bind which, unless explicitly stated in the constructor, is populated on first use from the where and table accessors and so on.)

Since I wasn’t using Moose this produced a lot more code than we’d normally expect to see, but because I was following a standard Moose conventions it’s still fairly easy to see what’s going on (I even went as far to leave a Moose style has accessor declaration in a comment above the blocks of code I had to write to sufficiently convey what I was doing.)

A results object

The second big architectural change I made was to stop talking directly to Test::Builder. Instead I changed to returning a results object which was capable of rendering itself out to Test::Builder on demand.

This change made the internals a lot easier to deal with. I was able to break the test up into several functions each returning a success or failure object. As soon I was able to detect failure in any of these functions I could return it to Test::DatabaseRow, but if I got a success - which now hadn’t been rendered out to Test::Builder yet - I could throw it away and move onto the next potentially failing test while I still had other things to check.

This made implementing my missing feature, the ability to report on all rows returned from the database not just the first one, much easier to implement.

Problems, problems, problems

After all this work, and spending hours improving the test coverage of the module, I still botched the release of 2.00. The old module tested the interface with DBI by checking against a database that was on my old laptop in 2003. Since I no longer had that laptop these tests weren’t being run (I actually deleted them since they were useless) and hence I didn’t notice when I’d broken the interface to DBI in my refactoring.

Ilmari pointed out I was being stupid a few minutes after I’d uploaded. Within ten minutes I’d written some DBI tests that test with SQLite (if you have DBD::SQLite installed) and released 2.01.

The biggest surprise was the next day where our overnight Hudson powered smokes failed at work, and the only thing that had changed was Test::DatabaseRow (we re-build our dependencies from the latest from CPAN every night, and it’d automatically picked up the changes.) Swapping in the old version passed. Putting the new version in failed. Had I missed something else in my testing?

No, I hadn’t.

After several hours of head scratching I eventually worked out that there was an extra bug in Test::DatabaseRow 1.04 that I’d not even realised was there, and I’d fixed it with 2.01. The tests were failing in our smokes but not because I’d broken the testing infrastructure, but because I’d fixed it and now I was detecting an actual problem in my Hudson test suite that had previously been unexposed.

What’s next?

Oh, I’m already planning Test::DatabaseRow 2.02. On github I’ve already closed a feature request that chromatic asked for in 2005. Want another feature? File a bug in RT. I’ll get to it sooner or later…

Nov 18, 2011 • Share / Permalink

...we are working with the Police, the intelligence services and industry to look at whether it would be right to stop people communicating via these websites and services when we know they are plotting violence, disorder and criminality.

Dear Duncan Hames, I write to you today to express my concerns regarding statements made by the prime minister with respect to restricting access to "social media". It should be fairly obvious when the chinese regime are praising our censorship plans that they are ill thought through and should be scrapped. However obvious I feel that I must still enumerate the ways that this plan is wrong on many levels. Firstly, your prime minister seems unable to distinguish between the medium and the message. As we move more and more into the digital age more and more communication will take new forms, these new forms will replace more traditional forms of communication in society. To seek to control over some forms of commutation is modern equivalent of the government seeking to control the ability for its citizens to write to newspapers or talk in the street. Secondly, the idea of the government silencing it's citizens from communicating with one another is chilling. While I can understand that some speech may be criminal by it's content, woe befall any government who tries to pre-emptively stop such speech, as these very same controls can be used, and abused, to control its citizens. Thirdly, the prime minister is seeking to put restrictions on people that have not been convicted of a crime (he said, I quote, "when we know they are plotting violence, disorder and criminality", but that is a matter for the courts not the "[the government,] the Police, the intelligence services and industry" to decide.) What safeguards are being proposed that I, a law abiding citizen, may not also be restricted from communication? Fourthly, and ironically, your prime minister is suggesting restricting the primary ability for communication with wider society by those individuals who he claims live outside of our society. Finally, I do not understand your prime minister's desire to push for further attention grabbing legislation when our police force can already wield the RIP Act to gather evidence from these new forms of communication. While I may not agree with the RIP Act, let our police forces use these powers to full effect before granting them new ones. As a member of your constituency I ask you to ensure that your prime minister is questioned about such blatant flaws in his proposals in parliment. Thanking you in advance for your help in this matter Yours sincerely, Mark Fowler

Aug 16, 2011 • Share / Permalink

Aug 5, 2011 • Share / Permalink

  eval {
    ...; # do something that might throw an exception
  };
  
  if ($@) {
    ...; # handle exception here
  }

  package SomethingComplex;
  sub new { return bless {}, shift };
  sub DESTROY {
    eval {
      ...; # some cleanup code that might throw an exception
    };
    if ($@) {
      ...; # handle exception in cleanup code
    }
  }
  
  package main;
  
  eval {
    my $thingy = SomethingComplex->new();
    ...; # do something that might throw an exception
  };
  if ($@) {  
    ...; # will never be executed even on exception
  }

  eval {
    ...; # do something that might throw an exception
    1;
  } or do {
    ...; # handle the exception here
  }

  use Try::Tiny;
  try {
    ...; # do something that might cause an exception.
  } catch {
    ...; # handle the exception stored in $_ here
  };

Jun 6, 2011 • Share / Permalink

The Problem I'm Trying to Solve

bash$ ssh sandbox1.dev.example.priv
ssh: Could not resolve hostname sandbox1.dev.example.prvi: nodename nor servname provided, or not known

bash$ hostify sandbox.dev.example.com

bash$ hostify -r

Using Net::DNS to do DNS lookups

my $res = Net::DNS::Resolver->new(
  nameservers => [qw(10.5.0.1)],
);

my $query = $res->search($hostname);

if ($query) {
  foreach my $rr ($query->answer) {
    next unless $rr->type eq "A";
    say "Found an A record: ".$rr->address;
  }
}

Using Tie::File to edit a file in place

tie my @file, 'Tie::File', "/etc/hosts"
  or die "Can't open /etc/hosts: $!";

# alter the 21st line in the file
$file[20] = "10.0.69.1 filestore.example.priv";

Rudimentary argument passing with -s

#!/usr/bin/env perl -s
if ($r) { print "Someone called us with -r\n" }

#!/usr/bin/env perl -s
use 5.12.0;
use warnings;
if ($::r && $::r) { say "Someone called us with -r" }

The Complete Script

#!/usr/bin/env perl -s

use 5.12.0;
use warnings;

use Net::DNS;
use Tie::File;

# look at the file
tie my @file, 'Tie::File', "/etc/hosts"
  or die "Can't open /etc/hosts: $!";

# did someone want to update all the cached entires?
if ($::r && $::r) {
  my $found = 0;
  foreach (@file) {
    # skip down until the comment in my /etc/hosts that
    # states that "cached entries are below this point"
    next unless $found ||= m/cached entries/; 

    # then replace each host entry
    s{\A\d+\.\d+\.\d+\.\d+\s+(?<host>.*)\z}{
       dns_lookup($+{host}) . " $+{host}";
     }e;
  }
  exit unless @ARGV;
}

my $host = shift // die "No hostname supplied\n";
my $ip = dns_lookup( $host );

# look for an existing entry and replace it
foreach (@file) {
  exit if s/\A\d+\.\d+\.\d+\.\d+\s+\Q$host\E\z/$ip $host/;
}

# not found?  Add it to the end
push @file, "$ip $host";

########################################################################

sub dns_lookup {
  my $hostname = shift;

  my $res = Net::DNS::Resolver->new(
    nameservers => [qw(10.5.0.1)],
  );

  my $query = $res->search($hostname);
  
  if ($query) {
    foreach my $rr ($query->answer) {
      next unless $rr->type eq "A";
      return $rr->address;
    }
    die "No A record for $hostname";
  }

  die "query for $hostname failed: ", $res->errorstring;
}

Jan 30, 2011 • Share / Permalink

Iterative Development

A worked example

#!/usr/bin/p
use strict;
use warnings;
use DBI;

my $dbh = DBI->connect(
  "DBI:mysql:database=live;host=livedb.twoshortplanks.com",
  "admin",
  "opensaysme",
  { RaiseError => 1 }
);

...

#!/usr/bin/perl

use strict;
use warnings;
use DBI;

sub connect_to_live_db {
  return DBI->connect(
    "DBI:mysql:database=live;host=livedb.twoshortplanks.com",
    "admin",
    "opensaysme",
    { RaiseError => 1 }
  );
}

my $dbh = connect_to_live_db();

...

package TwoShortPlanksUtils;

use strict;
use warnings;

use base qw(Exporter);
our @EXPORT_OK;

sub connect_to_live_db {
  return DBI->connect(
    "DBI:mysql:database=live;host=livedb.twoshortplanks.com",
    "admin",
    "opensaysme",
    { RaiseError => 1 }
  );
}
push @EXPORT_OK,"connect_to_live_db";

1;

#!/usr/bin/perl

use strict;
use warnings;

use TwoShortPlanksUtil qw(connect_to_live_db);
my $dbh = connect_to_live_db();

use FindBin;
use lib $FindBin::Bin;

#!/usr/bin/perl

use strict;
use warnings;

use FindBin;
use lib $FindBin::Bin;

use TwoShortPlanksUtil qw(connect_to_live_db);
my $dbh = connect_to_live_db();

Jan 22, 2011 • Share / Permalink

Pop Quiz, Hotshot

save $myobject;

• A function call to the save function, passing $myobject as the argument, i.e. the same as

  save($myobject)

• An indirect method call, calling the save method on $myobject, i.e. the same as

  $myobject->save()

Just say no

A Solution: The no indirect pragma

no indirect ':fatal';

bash$ perl
use strict;
use warnings;
no indirect ':fatal';

my $myobject = bless {}, "SomeClass";
save $myobject;
Indirect call of method "save" on object "$myobject" at - line 6.
bash$

Jan 14, 2011 • Share / Permalink

Our example PSGI application

#!/usr/bin/perl

use strict;
use warnings;

use Dancer;

get '/hello/:name' => sub {
   return "Why, hello there " . params->{name};
};

dance;

bash$ perl hellow.pl 
>> Dancer server 5758 listening on http://0.0.0.0:3000
== Entering the development dance floor ...

bash$ lwp-request http://127.0.0.1:3000/hello/world
Why, hello there world

Running this with a PSGI webserver and plackup

bash$ plackup -s Twiggy -p 4077 -a dancer.pl 
Twiggy: Accepting connections at http://0.0.0.0:4077/

bash$ lwp-request http://127.0.0.1:4077/hello/world
Why, hello there world

Configuring plackup to run on boot on a Mac

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
        <key>Disabled</key>
        <false/>
        <key>Label</key>
        <string>com.twoshortplanks.hellow</string>
        <key>ProgramArguments</key>
        <array>
                <string>/usr/local/bin/plackup</string>
                <string>-a</string>
                <string>Twiggy</string>
                <string>-p</string>
                <string>4077</string>
                <string>-a</string>
                <string>/usr/local/hellow/hellow.pl</string>
        </array>
        <key>OnDemand</key>
        <false/>
</dict>
</plist>

bash$ plutil -lint com.twoshortplanks.hellow.plist
com.twoshortplanks.hellow.plist: OK

bash$ sudo launchctl load /System/Library/LaunchDaemons/com.twoshortplanks.hellow.plist

bash$ sudo launchctl list | grep hellow
2074    -    com.twoshortplanks.hellow

bash$ lwp-request http://127.0.0.1:4077/hello/world
Why, hello there world

bash$ sudo launchctl unload /System/Library/LaunchDaemons/com.twoshortplanks.hellow.plist

bash$ sudo launchctl load /System/Library/LaunchDaemons/com.twoshortplanks.hellow.plist

Conclusion

Jan 7, 2011 • Share / Permalink

A new perl in five commands

First I need to setup perlbrew. This takes three commands (the latter two will be different on your system, but don't worry, in each case the previous command tells you what to do when you run it.)

   curl -L http://xrl.us/perlbrewinstall | bash
   /Users/mark/perl5/perlbrew/bin/perlbrew init
   echo >>~/.bashrc "source /Users/mark/perl5/perlbrew/etc/bashrc"

Now to install and switch to the latest development version of perl I only need type:

   perlbrew install perl-5.13.8
   perlbrew switch perl-5.13.8

And to install and switch to the latest stable perl I only need to type:

   perlbrew install perl-5.12.2
   perlbrew switch perl-5.12.2

Again, in excruciating detail

If you're anything like me, you've seen this kind of quickfire instructions on the web before and it leaves you with questions. Wouldn't it be nice if someone posted a complete dump of exactly what was going on when these commands are run? Yes, it would be...

Let's start from the top again. First, download and install perlbrew with the one liner:

  travis-4:~ mark$ curl -L http://xrl.us/perlbrewinstall | bash

Off it goes:

    % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                   Dload  Upload   Total   Spent    Left  Speed
  102   306  102   306    0     0    236      0  0:00:01  0:00:01 --:--:--  2756

  ## Download the latest perlbrew

  ## Installing
  The perlbrew is installed as:

      /Users/mark/perl5/perlbrew/bin/perlbrew

  You may trash the downloaded /private/tmp/perlbrew from now on.

  Next, if this is the first time you install perlbrew, run:

      /Users/mark/perl5/perlbrew/bin/perlbrew init

  And follow the instruction on screen.

  ## Done. (automatically removes downloaded /tmp/perlbrew)

We'd better follow its instructions, eh?

  travis-4:~ mark$ /Users/mark/perl5/perlbrew/bin/perlbrew init

This prints out some more stuff, including an instruction to modify our path:

  Perlbrew environment initiated, required directories are created under

      /Users/mark/perl5/perlbrew

  Well-done! Congratulations! Please add the following line to the end
  of your ~/.bashrc

      source /Users/mark/perl5/perlbrew/etc/bashrc

  After that, exit this shell, start a new one, and install some fresh
  perls:

      perlbrew install perl-5.12.1
      perlbrew install perl-5.10.1

  For further instructions, simply run:

      perlbrew

  The default help messages will popup and tell you what to do!

  Enjoy perlbrew at $HOME!!

Okay, let's modify our path to contain the instructions then:

  travis-4:~ mark$ echo >>~/.bashrc "source /Users/mark/perl5/perlbrew/etc/bashrc"

Now we need to restart bash. The easiest way to do that is to close the current terminal and open a new one (and stops us getting confused.) After this, installing a new Perl is a doddle, with just a single command.

  travis-4:~ mark$ perlbrew install perl-5.13.8

And it does all the hard work for us:

  Attempting to load conf from /Users/mark/perl5/perlbrew/Conf.pm
  Fetching perl-5.13.8 as /Users/mark/perl5/perlbrew/dists/perl-5.13.8.tar.gz
  Installing perl-5.13.8 into /Users/mark/perl5/perlbrew/perls/perl-5.13.8
  This could take a while. You can run the following command on another shell to track the status:

    tail -f /Users/mark/perl5/perlbrew/build.log

  (cd /Users/mark/perl5/perlbrew/build; tar xzf /Users/mark/perl5/perlbrew/dists/perl-5.13.8.tar.gz;cd /Users/mark/perl5/perlbrew/build/perl-5.13.8;rm -f config.sh Policy.sh;sh Configure -de '-Dprefix=/Users/mark/perl5/perlbrew/perls/perl-5.13.8' '-Dusedevel';make;make test && make install) >> '/Users/mark/perl5/perlbrew/build.log' 2>&1 
  Installed perl-5.13.8 as perl-5.13.8 successfully. Run the following command to switch to it.

    perlbrew switch perl-5.13.8

So, as it says, we can switch which perl we're using just by using the "perlbrew switch" command:

  travis-4:~ mark$ perl -v

  This is perl, v5.10.0 built for darwin-thread-multi-2level
  (with 2 registered patches, see perl -V for more detail)

  Copyright 1987-2007, Larry Wall

  Perl may be copied only under the terms of either the Artistic License or the
  GNU General Public License, which may be found in the Perl 5 source kit.

  Complete documentation for Perl, including FAQ lists, should be found on
  this system using "man perl" or "perldoc perl".  If you have access to the
  Internet, point your browser at http://www.perl.org/, the Perl Home Page.

  travis-4:~ mark$ perlbrew switch perl-5.13.8
  travis-4:~ mark$ perl -v

  This is perl 5, version 13, subversion 8 (v5.13.8) built for darwin-2level

  Copyright 1987-2010, Larry Wall

  Perl may be copied only under the terms of either the Artistic License or the
  GNU General Public License, which may be found in the Perl 5 source kit.

  Complete documentation for Perl, including FAQ lists, should be found on
  this system using "man perl" or "perldoc perl".  If you have access to the
  Internet, point your browser at http://www.perl.org/, the Perl Home Page.

We can switch back to the system perl with the "perlbrew off" command:

  travis-4:~ mark$ perlbrew off
  travis-4:~ mark$ perl -v

  This is perl, v5.10.0 built for darwin-thread-multi-2level
  (with 2 registered patches, see perl -V for more detail)

  Copyright 1987-2007, Larry Wall

  Perl may be copied only under the terms of either the Artistic License or the
  GNU General Public License, which may be found in the Perl 5 source kit.

  Complete documentation for Perl, including FAQ lists, should be found on
  this system using "man perl" or "perldoc perl".  If you have access to the
  Internet, point your browser at http://www.perl.org/, the Perl Home Page.

And that's about it. Very simple!

Jan 1, 2011 • Share / Permalink

An Example of the Problem

  package FacebookUser;
  use Moose;

  use LWP::Simple qw(get);
  use JSON::XS qw(decode_json);
  
  has api_access_token => ( isa => 'Str', is => 'ro', required => 1 );
  has username         => ( isa => 'Str', is => 'ro', required => 1 );
  has data             => ( isa => 'HashRef', is => 'ro', lazy_build => 1 );
  
  sub _build_data {
    my $self = shift;
    my $url = "https://graph.facebook.com/@{[ $self->username ]}?access_token=@{[ $self->api_access_token ]}";
    return decode_json(get($url));
  }
  
  sub get_firstname { return $_[0]->data->{first_name} }
  sub get_lastname  { return $_[0]->data->{first_name} }
  sub get_gender    { return $_[0]->data->{gender} }
  sub get_hometown  { return $_[0]->data->{hometown}{name} }

  my $user = FacebookUser->new(
    api_access_token => $access_token,
    username => "2shortplanks",
  );
  say "Hello ".$user->get_firstname;

  my $user = FacebookUser->new(
    access_token => $access_token,
    username => "2shortplanks",
  );
  say "Hello ".$user->get('firstname');

  Can't use a FacebookUser object as a URI at /Library/Perl/5.10.0/HTTP/Request/Common.pm line 103

The obscured parent method problem

  package FacebookUser;
  use base qw(Class::Accessor);

  # use Class::Accessor to create the api_access_token and username accessor methods
  __PACKAGE__->mk_accessors(qw(api_access_token username));

  use LWP::Simple qw(get);
  use JSON::XS qw(decode_json);

  sub data {
    my $self = shift;
    return $self->{data} ||= $self->_build_data;
  }

  sub _build_data {
    my $self = shift;
    my $url = "https://graph.facebook.com/@{[ $self->username ]}?access_token=@{[ $self->api_access_token ]}";
    return decode_json(get($url));
  }

  sub get_firstname { return $_[0]->data->{first_name} }
  sub get_lastname  { return $_[0]->data->{first_name} }
  sub get_gender    { return $_[0]->data->{gender} }
  sub get_hometown  { return $_[0]->data->{hometown}{name} }

  my $user = FacebookUser->new({
    access_token => $access_token,
    username => "2shortplanks",
  });
  say "Hello ".$user->get_firstname;

  Can't use a FacebookUser object as a URI at /Library/Perl/5.10.0/HTTP/Request/Common.pm line 103

  bash$ perl -MCarp::Always example.pl 
  Can't use a FacebookUser object as a URI at /Library/Perl/5.10.0/HTTP/Request.pm line 70
  	HTTP::Request::uri('HTTP::Request=HASH(0x10091c5f8)', 'FacebookUser=HASH(0x10091c1f0)') called at /Library/Perl/5.10.0/HTTP/Request.pm line 16
  	HTTP::Request::new('HTTP::Request', 'GET', 'FacebookUser=HASH(0x10091c1f0)') called at /Library/Perl/5.10.0/HTTP/Request/Common.pm line 103
  	HTTP::Request::Common::_simple_req() called at /Library/Perl/5.10.0/HTTP/Request/Common.pm line 20
  	HTTP::Request::Common::GET('FacebookUser=HASH(0x10091c1f0)') called at /Library/Perl/5.10.0/LWP/UserAgent.pm line 386
  	LWP::UserAgent::get('LWP::UserAgent=HASH(0x10083c840)', 'FacebookUser=HASH(0x10091c1f0)') called at /Library/Perl/5.10.0/LWP/Simple.pm line 36
  	LWP::Simple::get('FacebookUser=HASH(0x10091c1f0)', 'username') called at /System/Library/Perl/Extras/5.10.0/Class/Accessor.pm line 393
  	Class::Accessor::__ANON__('FacebookUser=HASH(0x10091c1f0)') called at FacebookUser.pm line 18
  	FacebookUser::_build_data('FacebookUser=HASH(0x10091c1f0)') called at FacebookUser.pm line 13
  	FacebookUser::data('FacebookUser=HASH(0x10091c1f0)') called at FacebookUser.pm line 22
  	FacebookUser::get_firstname('FacebookUser=HASH(0x10091c1f0)') called at example.pl line 15

Solving this: Deleting the subroutine when done

  delete $FacebookUser::{get};

  say "Hello ".$user->get('firstname');
  Can't locate object method "get" via package "FacebookUser" at example.pl line 15

  say "Hello ".$user->get_firstname;
  Hello Mark

  panic();
  sub panic() { say "Don't panic Mr Mannering!"; return }

Doing this automatically

  delete $FacebookUser::{get};

  use namespace::clean;

  package FacebookUser;
  use base qw(Class::Accessor);

  use LWP::Simple qw(get);
  use JSON::XS qw(decode_json);

  use namespace::clean;

  # use Class::Accessor to create the api_access_token and username
  # accessor methods
  __PACKAGE__->mk_accessors(qw(api_access_token username));

  ...

  package FacebookUser;
  use Moose;

  use LWP::Simple qw(get);
  use JSON::XS qw(decode_json);
  
  use namespace::clean -except => 'meta';
  
  ...

Conclusion

Jul 17, 2010 • Share / Permalink