Shell Scripting 101, Lesson 3

In Lesson 2, I showed you how to use system-provided environment variables such as USERNAME and COMPUTERNAME to retrieve information about currently logged-on users and their computers. You can also use environment variables for two additional purposes:

Creating Environment Variables to Represent a Specified Value
To create environment variables to represent a specified value, you need to use the Set command. Using the Set command with environment variables is a great way to make your lines of code shorter, easier to read, and simpler to change.

For example, suppose you’ve updated a file called listofstuff.txt on your server (\H10serverstufflistofstuff.txt) and now you must update a copy of that file on the D drive (D:stuffmorestuffmystuffyourstufflotsofstufflistofstuff.txt). The script must delete the old file if it exists, then copy the updated file. Listing 1 shows how this script looks if you hard-code the file’s pathname each time. As Listing 1 illustrates, this code is long and difficult to read. In addition, this code is difficult to edit if you need to change the D drive’s pathname because you must change each occurrence.

A better approach is to create an environment variable for the commonly occurring pathname. As the script in Listing 2 shows, you can use the Set command to create the StuffFile environment variable, which contains the pathname "D:stuffmorestuffmystuffyourstufflotsofstufflistofstuff.txt". Enclosing the pathname in quotation marks handles any spaces in the path. Without quotation marks, a pathname with a space will cause the script to fail.

After you create the environment variable, you simply specify that variable’s name enclosed in percent (%) signs each time you want to include the pathname. The percent signs let you retrieve and use the environment variable’s value.

Creating an environment variable instead of hard-coding the pathname can save a lot of space, especially in scripts that use long pathnames often. Even better, if the pathname changes, you need to change only one occurrence: the pathname in the Set command.

Limiting the Environment Variable’s Scope
By default, an environment variable you create with the Set command affects your entire system—that is, the environment variable has a global scope. You can use the Setlocal and Endlocal commands to make an environment variable affect only that script or a portion of it. In other words, you make the environment variable have a local scope. I highly recommend that you use the Setlocal and Endlocal commands to keep your Set commands from accidentally affecting System variables or causing a script that’s run in the same command shell window from inheriting values from an earlier run of a similar script.

To make an environment variable apply to an entire script, you place the Setlocal command before the Set command and the Endlocal command at the end of your script, as Listing 3 shows. To make the environment variable apply to a certain portion of the script, you place the Setlocal command before the Set command and the Endlocal command at the point at which you want the environment variable’s scope to end. Take, for example, the code in Listing 4. The Echo %TestVariable% command is no longer under the environment variable’s scope. As a result, if you run the code in Listing 4, the %TestVariable% won’t have a value. As this example illustrates, if you’re limiting the environment variable’s scope to only a portion of a script, you need to make sure that the code you exclude doesn’t need that variable’s value. For this reason, I recommend that you use the Setlocal command before any Set commands and the Endlocal command at the end of your script.

Using an Environment Variable to Temporarily Hold a Value
A special group of environment variables are the argument-holding variables %0 through %9. You can use these variables to bring input into your script at runtime. For example, suppose you want to run the script in Listing 2, but you want to change the pathname each time the script runs. Instead of opening the file and editing the Set command each time, you can use an argument-holding variable. As the script Test.bat in Listing 5 shows, you don’t include the Set command and you use %1 instead of %StuffFile%. Then, when you run Test.bat, you type the launch command

Test.bat "D:stuffmorestuffmystuffyourstufflotsofstufflistofstuff.txt"

in the command-shell window. When the script runs, the %1 variable brings your external input (i.e., "D:stuffmorestuffmystuffyourstufflotsofstufflistofstuff.txt") into the script.

If you want to use additional arguments, you add the next consecutive argument-holding variable to the code and that variable’s value to the launch command. You’re limited to 10 arguments (i.e., %0 through %9). You might be wondering, "What is the value of %0?" To learn the answer and to see how to use two arguments, create the script Test2.bat that Listing 6 shows. Then use the launch command

Test2.bat "The First Argument" "Another Argument"

to run Test2.bat. As Figure 1 shows, if you launch the script from the directory in which you placed the script, the value of %0 is the script’s filename. If you launch the script from a different directory, the value of %0 is the script’s pathname.

Practice Exercises