Rather than employing the usual tactic of showing you how to write the accursed ‘hello world’ program, so beloved of all computer programming introductions, let’s start with something entirely different:
print "This is most definitely not a hello world script.\n";
If you save this as the file
thing.pl, or similar, then you can run it from the command line (i.e. DOS prompt, bash shell, terminal window) thusly:
And it will probably do what you expect.
If you didn’t misspell anything in the script, when executed it should do something along the lines of printing out:
This is most definitely not a hello world script.
If you’re on a Unix or MacOSX, it’s better if you write this instead:
#!/usr/bin/perl print "This is most definitely not a hello world script.\n";
#!/bin/perl print "This is most definitely not a hello world script.\n";
at the top of your script, all on its own. Then you can just type:
at the command prompt. The
/usr/bin/perl (or whatever) after the shebang (
#!) is the path to the
perl interpreter, and it tells the Unix shell where to find it. Windows will also cope with a naked
thing.pl providing you have included
C:\Perl\bin in your
PATH environment variable and added
*.PL to the
PATHEXT environment variable. The ActiveState installer sorts this all out for you.
It’s generally considered polite to put a shebang at the top of scripts even if you’re not on Unix, in the name of portability:
perl works just about anywhere, and it’s a nice idea to try and make your Perl scripts and modules work anywhere too. Note the subtle difference between
perl (the executable program that interprets your program) and Perl (the language interpreted by the
perl interpreter). This distinction, and even more so the fact that neither is written ‘PERL’ is a shibboleth: using ‘PERL’ will mark you out as a n00b, so Don’t Do That.
So far so easy. The only thing I hope you may be wondering about is the
\n. This means ‘newline’, and will print whatever your operating system thinks is a newline at the end of the output. The
\n character is an escape sequence because whatever comes after a
\ has a special meaning to
\n is a newline,
\t is a tab, and
\\ is a literal \ character. We will come across many other sorts of escape sequence, especially when we get onto regexes later.
Something very important to notice about the script that we have just written is the
; semicolon at the end of the line: Perl statements must (almost) always end in a semicolon, unlike those in some other languages beginning with p. Forgetting trailing semicolons is a frequent source of bugs.
Strings and quoting
One of Perl’s strengths is the multitude of ways you can quote and manipulate a string like
"This is most definitely not a 'hello world' script". First, try this out:
#!/usr/bin/perl print 'This is most definitely not a hello world script.\n';
' single quote replacing the
" double quote we used in the first script. If you run this, it will print out:
This is most definitely not a hello world script.\n
In Perl, double quotes will interpolate escape characters like
\t, so an escape sequence like
\n gets converted into a real, literal newline between double quotes. Single quotes will translate only the escape
\' to a literal
' and the escape
\\ to a literal
\. Double quote are therefore what you need to do clever, intuitive things; and single quotes are what you need to do simple, literal things. Since single quotes only interpolate
\\, this is the way to insert single quotes into a single-quoted string:
#!/usr/bin/perl print 'This is most definitely not a \'hello world\' script.';
which prints out:
This is most definitely not a 'hello world' script.
with embedded single quotes around the ‘hello world’ bit. The reason we must escape the single quotes within our string is that if you don’t escape the two single quotes around the
'hello world' bit,
perl will think you mean the string to end just before the hello, and will throw a tizzy:
#!/usr/bin/perl print 'This is most definitely not a 'hello world' script.\n';
without suitable escaping of the quotes, looks to
#!/usr/bin/perl print 'This is most definitely not a ' # FIRST UNESCAPED SINGLE QUOTE ENDS THE STRING, # WHAT THE HELL IS THIS NEXT BIT SUPPOSED TO MEAN? script.\n';
The same applies to double quotes: if you want double quotes in a double-quoted string, you’ll need to escape them too
#!/usr/bin/perl print "I said \"hello world\". Are you deaf?\n";
In Perl, all strings (i.e. anything that isn’t a number) should be quoted somehow or other. If you don’t use quotes,
perl will quite likely barf, probably saying something about ‘bare-words’. You must put quotes around strings, so:
is fine, as is:
but the bare-word
Nubbins in this:
will not be interpreted as a string. Perl has several other methods of quoting strings like heredocs and
printf, which we’ll also cover later.
Next up…input and output.