Advanced Unix Tools CS214 Spring 2004

Lecture 7 Monday March 8, 2004

Perl, the Practical Extraction and Report Language, is a scripting language with powerful textmanipulation capabilities. Because of these, Perl has become a very popular language in the context of web servers. As you have seen doing homeworks 1 and 3, creating web pages involves hacking HTML, which is just a textual representation. This week, we will cover the basics of Perl programming. We will be using Perl 5.6, /usr/local/bin/perl on babbage, although I doubt that we will ever talk about features specic to 5.6. We will use Perl to write scripts, in a way similar to the way we have been writing shell scripts. We will store Perl scripts in text les with extension .pl, which is the conventional extension. There are two ways of executing a Perl script. The rst is simply to invoke the Perl interpreter perl with the script to execute as an argument, e.g., perl foo.pl. The second way is to use the magic number approach. Recall that when Unix is instructed to execute a le and that le starts with the two special characters #!, it uses the remainder of the rst line as an indication of what program to use to execute the le. In the case of shell scripts, we had scripts begin with #! /bin/bash to indicate that bash was to be invoked to execute the script. In our case, we can put #! /usr/local/bin/perl as the rst line of our scripts, and set execute permissions on the script, and Unix will automatically invoke perl to execute the script.

Basics of Perl
A Perl script is a sequence of statements. There are a few forms of statements, most of which we will cover in the remainder of this lecture. Let me rst note that Perl programming is not shell programming. Whereas shell programming involves putting together shell commands, Perl is a different language. Although they are syntactical similarities, you cannot directly use shell commands in Perl scripts. Comments can be put anywhere in a Perl script. A comment starts with the character #. Anything on a line after a # is ignored by the Perl interpreter. Perl recognizes two different types of values: integers and strings.1 Integer literals are written as one expects, 1, 2, -1, . . .. A string is a sequence of characters. String literals can be written in two distinct ways. A string this is a string written using single-quotes is taken to mean the sequence of characters appearing as is between the single-quotes. A string "this is also a string" written using double-quotes is taken to mean the sequence of characters obtained after
1

In fact, Perl 5.0 introduced a new type, objects, but we will not cover those in this course.

interpreting whats between the double-quotes. For most characters, interpretation doesnt affect anything. Hence, "this is also a string" represents the same string as this is also a string. We will see in a little while some examples on which single-quotes and double-quotes differ. Perl allows the use of variables. A (scalar) variable can contains strings or integers. All scalar variables in Perl are required to start with the $ characters. An assignment statement is used to assign value to variables. For example, $i = 10; $word = hello; Notice that statements are always terminated with a semicolon ;. Also, notice that $ is part of the variable name, and hence must be used even when assigning to a variable. (This is different than the use of variables in the shell, a frequent beginners problem.) One of the difference between single-quotes and double-quotes for dening strings is how to handle variables appearing in the string being dened. Compare the following denitions: $test1 = this is the value of variable i: $i; $test2 = "this is the value of variable i: $i"; Since $test1 is dened as a literal string, all the characters between the single-quotes are part of the string. However, $test2 is interpreted, which among other things means that variables appearing inside the string are replaced by their value. Hence, the $i in $test2 is replaced by 10, so that the string stored in $test2 is in fact the string this is the value of variable i: 10. Variable assignment is our rst example of statement. A statement that comes in handy is print that simply prints values to standard output. In fact, print is an example of a built-in function. Perl has many built-in functions, and we will describe some of them as we encounter them in examples. The basic form of print is: print expr ,expr ,...,expr ; where expr s can be integers or strings, or in general any expression yielding such. (We will dene useful operations to build expressions shortly.) The values printed are printed without spaces in between them. Also, print does not put newlines after printing. If you want a newline, you need to explicitely use "\n". (Notice the double-quotes; if you use single-quotes \n, then you simply output the characters \ followed by n; you want to interpret it as a newline, requiring you to use double-quotes.) Here are some examples:

$i = 42; print "Hello,\n"; print "This is the value "; print "of integer variable i:", $i, "\n"; Executing the above script outputs the following to stdout: Hello, This is the value of integer variable i: 42 Operations are useful for constructing expressions. Some operations assume that their arguments are integers, such as +, -, *, /, and others. If these operations are given arguments that are strings, Perl will automatically convert those values to integers before applying the operation. If the string cannot be converted to a meaningful integer, then 0 is used. Consider the following example: $a = "10"; $b = $a + 2; $c = $b + " 34 "; The variable $c gets nal value 46. Other operations assume that their arguments are strings. The basic example of such an operation is . representing string concatenation. Hence, "foo" . "bar" yields the string "foobar". As before, if an argument is not a string, it is converted to a string prior to application of the operation. This leads to some cute behavior: $a $b $c $d = = = = 5; $a + 10; $b . "200"; $c + 10;

You can check that the variables are assigned the following values: $a gets 5, $b gets 15, $c gets 15200 (!), $d gets 15210.