How Perl Works

To start with Perl you need the Perl interpreter. On any UNIX machine there is a 99.99-percent probability that it's already there. On a Windows machine or a Mac, you need to download the latest release of the language and install it on your machine. (See the links at the end of this article for more information.) Perl is widely available on the Web and is free.

Next, make sure you look in the DOCS directory that comes with Perl -- there will be user-manual-type stuff in there. At some point, it would not hurt to read through all of the documentation, or at least scan it. Initially it will be cumbersome, but after reading this article it will make much more sense.

Once you have Perl loaded, make sure you have your path properly set to include the Perl executable. Then, open a text editor and create a text file. In the file, place the following line:

print "Hello World!\n";

Name the file "test1.pl". At the command prompt, type:

perl test1.pl

Perl will run and execute the code in the text file. You should see the words "Hello World!" printed to stdout (standard out). As you can see, it is extremely easy to create and run programs in Perl. (If you are using UNIX, you can place a comment like #! /usr/bin/perl on the first line, and then you will not have to type the word "perl" at the command line.)

The print command prints things to stdout. The \n notation is a line feed. That would be more clear if you modified the test program to look like this (# denotes a comment):

# Print on two lines
   print "Hello\nWorld!\n";

Note that the print command understood that it should interpret the "\n" as a line feed and not as the literal characters. The interpretation occurred not because of the print command, but because of the use of double quotes (a practice called quoting in Perl). If you were to use single quotes instead, as in:

print 'Hello\nWorld!\n';

the \n character would not be interpreted but instead would be used literally.

There is also the backquote character: `. A pair of these imply that what is inside the quotes should be interpreted as an operating system command, and that command should be executed with the output of the command being printed. If you were to place inside the backquotes a command-line operation from the operating system, it would execute. For example, on Windows NT you can say:

print `cmd /c dir`;

to run the DIR command and see a list of files from the current directory.

You will also see the / character used for quoting regular expressions.

The print command understands commas as separators. For example:

print 'hello', "\n", 'world!';

However, you will also see a period:

print 'hello'. "\n". 'world!';

The period is actually a string concatenation operator.

There is also a printf operator for C folks.

PERL Note

In Windows NT, you cannot say:

print `dir`;

because dir is not a separate executable -- it's part of the command interpreter cmd. Type cmd /? at a DOS prompt for details.

Variables are interesting in Perl. You do not declare them, and you always use a $ to denote them. They come into existence at first use. For example:

   $s = "Hello\nWorld\n";
   $t = 'Hello\nWorld\n';
   print $s, "\n", $t;

Or:

   $i = 5;
   $j = $i + 5;
   print $i, "\t", $i + 1, "\t",  $j;     # \t = tab

Or:

   $a = "Hello ";
   $b = "World\n";
   $c = $a . $b;    # note use of . to concat strings
   print $c;

Since . is string concatenation, .= has the expected meaning in the same way that "+=" does in C. Therefore, you can say:

   $a = "Hello ";
   $b = "World\n";
   $a .= $b;
   print $a;

You can also create arrays:

   @a = ('cat', 'dog', 'eel');
   print @a, "\n";
   print $#a, "\n";  # The value of the highest index, zero based
   print $a[0], "\n";
   print $a[0], $a[1], $a[2], "\n";

The $# notation gets the highest index in the array, equivalent to the number of elements in the array minus 1. As in C, all arrays start indexing at zero.

You can also create hashes:

   %h = ('dog', 'bark', 'cat', 'meow', 'eel', 'zap');
   print "The dog says ", $h{'dog'};

Here, 'bark' is associated with the word 'dog', 'meow' with 'cat', and so on. A more expressive syntax for the same declaration is:

   %h = (
      dog => 'bark',
      cat => 'meow',
      eel => 'zap'
   );

The => operator quotes the left string and acts as a comma.

You can create a simple for loop like you do in C:

   for ($i = 0; $i < 10; $i++)
   {
      print $i, "\n";
   }

While statements are easy:

   $i = 0;
   while ( $i < 10 )
   {
      print $i, "\n";
      $i++;
   }

If statements are similarly easy:

   for ($i = 0; $i < 10; $i++)
   {
      if ($i != 5)
      {
         print $i, "\n";
      }
   }

The boolean operators work like they do in C:

For numbers:

Others:

If you have an array, you can loop through it easily with foreach:

   @a = ('dog', 'cat', 'eel');
   foreach $b (@a)
   {
      print $b, "\n";
   }

Foreach takes each element of the array @a and places it in $b until @a is exhausted.

PERL Note

You must use the "begin" and "end" braces -- { and } -- even for a single line.

You create a subroutine with the word sub. All variables passed to the subroutine arrive in an array called _. Therefore, the following code works:

   show ('cat', 'dog', 'eel');

   sub show
   {
      for ($i = 0; $i <= $#_; $i++)
      {
         print $_[$i], "\n";
      }
   }

Remember that $# returns the highest index in the array (the number of elements minus 1), so $#_ is the number of parameters minus 1. If you like that sort of obtuseness, then you will love PERL.

You can declare local variables in a subroutine with the word local, as in:

   sub xxx
   {
      local ($a, $b, $c)
      ...
   }

You can also call a function using &, as in:

&show ('a', 'b', 'c');

The & symbol is required only when there is ambiguity, but some programmers use it all the time.

To return a value from a subroutine, use the keyword return.

Reading from STDIN

To read in data from the stdin (standard in), use the STDIN handle. For example:

   print "Enter high number: ";
   $i = <STDIN>;
   for ($j = 0; $j <= $i; $j++)
   {
      print $j, "\n";
   }

As long as you enter an integer number, this program will work as expected. <STDIN> reads a line at a time. You can also use getc to read one character, as in:

$i = getc(STDIN);

Or use read:

read(STDIN, $i, 1);

The 1 in the third parameter to the read command is the length of the input to read.

Reading Environment Variables

PERL defines a global hash named ENV, and you can use it to retrieve the values of environment variables. For example:

print $ENV{'PATH'};

Reading Command Line Arguments

PERL defines a global array ARGV, which contains any command line arguments passed to the script. $#ARGV is the number of arguments passed minus 1, $ARGV[0] is the first argument passed, $ARGV[1] is the second, and so on.

You should now be able to read and write simple Perl scripts. You should also be able to wade into the full documentation to learn more. For additional information, see the links on the next page.