About Kent Cowgill
Articles filed under...
abs ab_ripper andylester arms back baggyshorts bestpractices biceps bike birthday blog bugs bus calculator cardio catalyst cgi chart chest chinups code cpan datamodel dbi doctor documentation exercise exhaustion fitness flattire flat_tire google gps heart_rate helmet history home houston html humor journal kate kenpo kenpo_x kettlebell knees lazy legs lisa lisanne maps math matthew michaelmckenna mom montreal motivation movie mysql oops orm P90X pain park patellar_tendonitis patrick pdf perl phb photos physical_therapy plyometrics poor_gait presentation procrastination progress pullups pushups pyramid rabbits racecondition rant refactor rest ribs ride route running shoulders situps slides sore spike sql statistics syntax test testing textile timex training triceps ups versioncontrol video vim vimrc walk warren work workouts yapc yapcna2007 yoga youtube

A R C H I V E S

(3)
(1)
(3)
(2)
(7)
(15)
(16)
(25)
(3)
(4)
(2)
(4)
(11)
(1)
(1)
(3)
(2)
(2)
(10)
(5)
(2)
(3)
(4)
(9)
(21)
(3)
(3)
(1)
(6)
(4)
(1)
(4)
(3)
(2)
(1)


    Is Kent Cowgill Online?
    View Kent Cowgill's profile on LinkedIn
    Add to Technorati Favorites

    Recent Entries...

    Re: Re: Yoga kicks my butt

    Chris @ 46: Tatyana @ 21 – at best its a stop gap measure...

    Re: Vibram FiveFingers FTW

    Hats off to whoever wrote this up and potesd it....

    Re: A little more detail on using a new model

    百度 [url=http://www.sina.com]sina[/url] ...

    Re: Catching up through week 7

    testing video ...

    Re: Porting a non-Moose object to Moose

    Wow, look what I found, greedy genius ...

    Re: Porting a non-Moose object to Moose

    Kevin, You're right, that does seem a little confusing. ...

    Re: Porting a non-Moose object to Moose

    Wait. I'm confused. Moose isn't the tool to reach for. So...

    Re: Porting a non-Moose object to Moose

    You should switch to MooseX::Types to declare your Typed and...

    Porting a non-Moose object to Moose

    I'm currently working with a lot of legacy code in an envi...

    Testing strategy for mocking code

    I keep finding myself using the following idiom for writing ...
    weblog | `web·lôg -läg |
    noun
    Another term for BLOG
    ORIGIN 1990s: from web in the sense [World Wide Web] and log in the sense [regular record of incidents.]
    blog | bläg |
    noun
    A web site on which an individual or group of users produces an ongoing narrative.
    ORIGIN a shortening of WEBLOG.

    Tired eyes

    Kent Cowgill

    Posted by
    on Thursday, August 20 2009, 8:00pm

    It's so easy to miss things if you're just casually trying to familiarize yourself with some perl code that's brand new to you.

    It wasn't until I was creating a comprehensive set of unit tests and calculating code coverage metrics that I saw this not once, but three times in a row...

    if( condition ){
      die "some appropriate error message";
      return;
    }

    Of course, even when writing the tests, I didn't spot it until I was writing a test for the last if.

    Thankfully Devel::Cover quickly highlighted that no matter what I tried, I couldn't write test code to exercise that return; statement.

    Same thing with this one:

    sub method {
      my( $self, $args ) = @_;
      return unless $args->{ key };

      # more code ...

      if( $args ){
        # more code here that uses $args
      }
      # etc...
    }

    I couldn't understand why I couldn't manage to manipulate the inputs of method to cover all branches of that if.

    And then I realized it would never get that far with the early short-circuit return. Duh.

    Filed under: testing perl refactoring code cleanup
    Related Photos: code perl
    Comment on 'Tired eyes' [ no replies - permalink ]

    Main Page | Login

    Do you want to buy me ? Find more gift ideas at my wishlist