in reply to Global Vars

Melly:

For quick & dirty stuff, I don't mind globals a bit. They get to be a problem when the program starts to get too large and complicated. If you ever find yourself using grep to track down usage of a global variable, or when you have a quick reference sheet for your global variables, or when you track down a bug due to reusing a global variable name that you forgot, then you know that your program is past due for some reorganization/refactoring/etc.

I always wonder why quick & dirty programs last longer than the supposedly long-term ones?

...roboticus

Replies are listed 'Best First'.
Re^2: Global Vars
by tbone1 (Monsignor) on Sep 04, 2009 at 13:57 UTC
    What he said. I have sometimes created a global hash (called something like %my_global_values or something like that) so that 1) it's obvious what is a global variable, and 2) I'm far less likely to accidentally reuse a variable.

    And the quick and dirty programs probably last longer because expectations are lower, they're for (relatively) simple things, and developers are less likely to overthink them.

    --
    tbone1, YAPS (Yet Another Perl Schlub)
    And remember, if he succeeds, so what.
    - Chick McGee

[reply]
      tbone1:

      Nice technique. It seems to me that it's a handy step on the way to splitting things off into modules and/or object-oriented bits. Since all the references go through your global hash anyway, when you split off a function/module/object, the code in your subroutines wouldn't have to change much:

      BEFORE 

      my %my_global_values = (foo=>0, etc=>0);
my $rGlobals = \%my_global_values;
sub zigafoo { $rGlobals-> = 7 }

      [download]
      AFTER 
      package barbaz;
my %my_package_values = (foo=>0);
my $rPackage = \%my_package_values;
sub foobar { $rPackage->{foo} = 7; }

      [download]

      ...roboticus
[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom