The Bash Shell

Bash is short for Bourne Again Shell. This shell is available as free software under the GNU license. The name reflects the bash shell's origins as an enhanced version of the Bourne shell. It also is typical of the sly sense of humor—some might say silliness—sometimes used in naming Linux free software programs.

Bash is the default Red Hat Linux shell. It is generally an easy shell to work with. According to the Free Software Foundation, "bash is an sh-compatible shell that incorporates useful features from the Korn shell ksh and the C shell csh."

Unless you are already used to another shell, you probably should stick with the bash shell.

Working with environment variables

An environment variable is used to expose information to the shell and to programs invoked at the shell's prompt. In other words, the information contained in an environment variable is used to determine interactions among the user, the shell, the system, and programs.

By convention, environment variables are all uppercase. Also, you need to know that $ is used as a special character to mean the content of a variable rather than the literal string that is the variable's name.

It is typical for an environment variable to be used to contain a location. For example, the environment variable HOME is used to store a user's home directory.

Some environment variables are used to tell a particular program the location of files it needs. For example, to invoke the PostgreSQL's Postmaster service, the user postgres—described in Chapter 9—had to have the environment variable PGLIB set to the location of PostgreSQL's data files.

Many environment variables are set in the overall or default configuration files. (Configuration files are explained later in this chapter.) Until they are changed for a particular user, these environment variables are set the same for all users.

Other environment variables are specific to a particular user, like the postgres user I just mentioned. This occurs when the environment variable is set on the fly interactively, or added only to a user-specific configuration file.

The echo command displays a line of text.

To see the current value of an environment variable:

To list all the current environment variables:

Piping to more

When you displayed your environment variables using the set command, they may have filled more than one terminal or telnet screen, so you couldn't view the variables at the beginning.

The more Linux command displays file contents one screen at a time. When you've finished viewing each screen, you press the spacebar to move on to the next screen.

To pipe a command means to send the output of one Linux command to another. The symbol that indicates piping is |.

To view environment variables one screen at a time:

									set | more
								

"Piping to more" is very useful and can be used to view directory listings.

To view a directory listing one screen at a time:

									ls –a –l | more
								

The more command can also be used to read the contents of files.

To use more to view the contents of a file:

Setting environment variables

You can set environment variables interactively at the command line. This is true for environment variables that you make up on the spot, those that programs require, and those that come preconfigured—although you should probably be a bit circumspect about editing variables that the shell sets automatically.

Keep in mind that an environment variable set on the fly exists only as long as the session in which it was created. This is sometimes useful in debugging configuration issues, but if you want the environment variable to have its value the next time you log on, you will need to set it in a configuration file. I'll explain how to do this later in this chapter.

Setting an environment variable involves two steps:

• Assigning the variable

• Exporting it so that it is available to all programs and scripts running in the current session

To set an environment variable on the fly:

Setting the PATH environment variable

A very important environment variable, PATH is used to store the value of the path—the directories where the shell can look for files that have been invoked without a fully qualified path being required.

For example, if myprogram is in the /harold/ bin directory, I can run myprogram simply by typing myprogram if /harold/bin is on the path—that is, if /harold/bin is included in the PATH environment variable.

On the other hand, if /harold/bin is not on the path, to invoke myprogram I would have to include its location:

							/harold/bin/myprogram
						

Within the PATH variable, the various directories that are on the path—that is, that are available without full qualification—are separated by colons (:), like this:

							/bin:/usr/bin:/usr/X11R6/bin:/harold:/harold/bin
						

If you were adding a new directory to your path, one thing you would not want to do would be to delete the current contents of the PATH variable. If you did so, a number of things quite possibly wouldn't work right.

Fortunately, the $ operator can be used to reference the contents of the current PATH variable. New additions to the PATH variable can then be concatenated with the existing value.

To add a directory to the PATH variable:

• At the command prompt, type
  
  

  									export PATH=$PATH:
  									new directory
  								

  
  For example, type
  
  

  									export PATH=$PATH:/harold/bin
  								

  
  

  Tip

  If the PATH string contains any spaces, you should use quotation marks. For example,

  										export PATH= "$PATH:/harold/my wpdocs"
  									

  
  

Editing the configuration files

Editing or adding environment variables at the command prompt is all very well and good, but your changes do not persist, meaning that the next time you log on, they will be gone.

If you want to create environment variables (or change existing ones) and have your work persist between sessions, you need to edit a configuration file.

If you are using Red Hat Linux 6 and the bash shell, there are four relevant configuration files, listed in Table 10.2.

When you look at Table 10.2, you'll see a special character being used: the tilde (~). This character indicates the user's home directory.

To go to your home directory:

									cd ~
								

If you look at Table 10.2, you'll see that the files in the user's home directory start with a period: for example, .bash_profile. The period means that they are hidden files.

If you open the Gnome File Manager with its default settings, hidden files will not be displayed.

To view hidden files in Gnome File Manager:

From a system administrator's viewpoint, if you create a user with the Gnome/Linux Configuration tool (see Figure 10.13), and if the home directory you set is under the /home directory, then the configuration files stored in /etc/skel are automatically copied to the user's new home directory.

In this case, skel is shorthand for skeleton, meaning a template. As an administrator, if you want to change all of the configuration files for new users, you would simply edit the files in etc/skel.

Of course, it would be the administrator's option to manually tweak a specific user's configuration files—or to add a user with no configuration files at all, for that matter.

For more information on using the Gnome/ Linux Configuration tool to add users and groups, see Chapter 6. For information on administering users at the command line, see Chapter 11.

To add an environment variable forauser:

										MYENV= "Lions, and tigers, _
   and bears,  Oh my!"
export MYENV

To add to the path globally

										PATH="$PATH:/home/hdavis/ _
   Linuxvqs"

Changing the bash prompt

You can easily configure the command-line prompt to include various useful pieces of information. To change the system-wide prompt, edit /etc/bashrc. To change a single user prompt, edit the .bashrc file in that user's home directory.

In either case, you will need to edit the line beginning PS1, which is often the only line that is not a comment in these files. The various kinds of information that you can include go between the square brackets and are represented by code letters (see Table 10.3). You can also include literal text if you wish.

To add full directory, date, and time information to the prompt:

									PS1="[u working in w Date: _
   d Time:   	]\$"