Re: how would you work with this material

On Thu, Dec 26, 2002 at 10:21:39PM -0800, Michael R. Wolf wrote:
> 
> What do you think of material that has this material?  


That it is not making effective use of the time available.

Perl is a large language. There can never be "enough" time, so
you have to be hard-hearted about what you'll spend the time on...

Also, it presents error-prone ways of doing things when they
can easily be done more safely. ie. it seems to prefer "tricks"
rather than solid Software Engineering.


> How would you present it to students in an "Intro to Perl Programming"
> course?


"I don't recommend doing this, but you might see it in other people's
code, so I'll cover it quickly so you'll understand it if you see it."


> What would you do about it if you were employed by the company that
> wrote it?


Whine.


> What about if you contracted to them?


Replace it with something more valuable, or at least be
brisk and caveat it.


> Changing the $#array also changes the length of the array
> 
> Similar to using the pop function


TMTOWTDI is nice, but a "clear" way is better than an "obsure" way.

I'd teach the clear way and save the "obscure way" time 
for something else.


> @teams = ("redsox","yankees", "bluejays","rockies","giants","mariners");
> print "Last element int he Array: $#teams\n";


Errr, the "last element in the array" is "mariners", not "5"  :-)


> print "\@teams has " .($#teams + 1). " elements\n";

   print "\@teams has " . @teams . " elements\n";

will do the same thing, and avoid the possibility of the
famous "off by one" class of mistakes.


> This funciton is useful for loopping through the array


But not as useful as foreach.


> for($i=0; $i < $#teams+1; $i++) 


If you maintain the index yourself, you have the possibility
of making a mistake. If you use foreach instead, perl will do
the indexing for you.

Machines don't make mistakes, people do, so it's better to
leave it to a machine when possible.


> To grow or shrink the array, you simply change the value of the last
> index in the array


That is the obscure way. The clear way it to just put in into
the array (push/splice), and let perl grow it for you.

(unless you anticipate the array becoming "very large", but
 beware of premature optimization...
)


> Output:
> 
> Elements 11
> 1, 2, 3,  , 300, , , , , , 


Errr, where are all of the "uninitialized" warning messages?  <grin>


> - Get better performance behind the scenes by creating the array and
>   growing it to the required length first instead of building it
>   element by element.


Beware, as above.


> It is possible to empty an array by setting its length to a negative
> number

> @#array = -1;
  ^
  ^

I'll ignore the typo.

That's the obscure way.

@array = ();   # the clear way


Is "job security" the primary objective of the class?  <grin again>


> -- 
> Michael R. Wolf
>     All mammals learn by playing!


Some lessons are better learned later, when you time to kill.


-- 
    Tad McClellan                          SGML consulting
    [email protected]                   Perl programming
    Fort Worth, Texas