Beginner’s Bash

 

This article was written by Ramon Casha

This tutorial presents the Linux terminal and the “bash” shell to people who have never used a command line to give commands to an operating system before, or who have never done so in Linux/Unix. People who have already used a Unix shell before might find it a bit simple.

Due to the popularity of the Microsoft Windows operating system, and the large number of ex-Windows-users who have discovered Linux, I have provided comparisons to equivalent or similar features and terminology in Windows’ MS-DOS prompt or Command prompt. These are provided as an additional help for Windows users, and are not necessary to follow this document.

Since this tutorial is intended as an introduction, it is purposely not comprehensive. Several commands, for example, are only explained in the depth necessary to gain an understanding of what they do and how to use them, not necessarily to use them to their full potential.

In the article I assume that the user is already familiar with concepts such as files and directories (folders), as well as filenames, etc.

Introduction

Nowadays, as soon as you get Linux installed, you get a nice graphical interface and rarely if ever need to make use of the so-called terminal mode (aka shell prompt).

However, in Linux the simple, modest terminal is not merely an afterthought, but an extremely powerful tool. While it may be true that you don’t need to use it, it’s not that difficult to learn, and very useful to know. Fortunately for yourself, with Linux’s users and security, you can create a new user for playing around, then you can experiment to your heart’s content without breaking anything.

In this document we will go step by step through many common tasks. Hopefully by the end you will feel quite familiar with the terminal. This will come in handy when, for instance, you encounter some document which instructs you to open a terminal and enter certain commands.

Preparation

As I mentioned before, it’s a good idea to create a new user for experimenting. In Linux, a normal user cannot break anything in the system, but can still delete his/her own files. By creating a new user you will have a new “playpen” to play around in, so if you make some mistake you won’t delete the files you normally work with.

One thing you must never do (until you know what you’re doing) is to work as the “root” user. This user has the necessary permissions to delete or alter all files on the computer, which is generally not considered a good move.

So the first thing to do is to use your favourite tool to create a new user. Use all the normal settings, but ensure that the “shell” setting is “/bin/bash”. I called mine “sandbox” but you can use any name you like. Since different Linux distributions provide different tools which can be used to perform this step you should use whichever tool you find best. Some examples include KDE’s User Manager or the LinuxConf tool.

This picture shows KDE’s User Manager being used to create the user “sandbox”. Any other user-management program may be used instead.

After you have created the new user, you can log out, then log in using that user to ensure that you can perform any command safely.

The “Bash” Shell

A “shell” is a program which interprets commands, either typed in directly by the user, or contained in a file called a “shell script”, which is a simple interpreted program. The equivalents in WindowsTM would be “command processor” for shell, “COMMAND.COM” or “CMD.EXE” instead of bash, and “.BAT files” instead of shell scripts.

Linux has a variety of different shells, but certainly the most popular is “bash”, so it is this one which will be described here (even though many of these instructions apply to all shells). Some of the others are retained simply because there are lots of people who got used to them and don’t wish to change, or because they are aimed at a specialised set of users.

Trivia:

Linux, like all unixes, uses a lot of acronyms for its program names, many of them somewhat humorous (eg, yacc=”yet another compiler compiler”). Most shells end in “sh”, and include ksh (korn shell), csh (C shell… seashell, get it?) and bash (Bourne again shell).

Accessing the Terminal

The easiest way to access the terminal, once you have logged in as the new user you created, is to click on the terminal icon from the icon panel at the bottom of the screen, or from the start menu.

The first picture shows KDE’s terminal icon, and the second one shows Gnome’s. The instructions in this document apply to both, since both will load the user’s default shell which is bash. With other GUI’s, look for “xterm” in the menus, or any other program name ending in “term” or “terminal”.

Another way of accessing the terminal is to load Linux in text mode. Many servers are configured not to load graphics by default, so the first thing you’d meet is a login prompt in text mode. “Home” or “desktop” computers on the other hand usually start up in graphics mode.

You can try the text-mode prompt easily by pressing Alt-Ctrl-F1to switch to text mode and Alt-Ctrl-F7 to switch back. Actually, from Alt-Ctrl-F1 to Alt-Ctrl-F6 are usually set up as 6 individual text-mode login prompts, while Alt-Ctrl-F7 is the graphics mode.

When you first load the terminal window (or log in from the text-mode login prompt), you will see a

prompt somewhat like the above. The actual text within the prompt can be user-defined, and varies between distributions. Generally however, the prompt includes the current username, the name of the computer, and the current directory. Thus in the above picture, the user is “sandbox”, the computer is “laptop”, and the directory is “~”, which is actually short for the home directory.

This picture shows a terminal window – in this case, KDE’s “Konsole”.

Entering commands

Throughout this document you will be entering many commands. Each time, type in the words and symbols as given and press Enter (or Return) at the end.

Keep in mind that, as with almost everything else in Linux, the commands are case-sensitive. So, if you’re supposed to type in “ls”, typing “LS” won’t work.

Exiting from the terminal

The first thing you might want to know is how to exit. Simply type in “exit” and press Enter, or else press Ctrl-D. The terminal window will quit. Although it is also possible to exit by clicking the “close” button or menu-option, this is the preferred method since it ensures that you are not in a text editor or other application where you could lose data.

A Simple Command

In the terminal window, enter “ls” and press Enter. You should see something like this (the actual contents may vary):

sandbox@laptop:~ > ls
KDesktop public_html  snapshot1.png
sandbox@laptop:~ >

The “ls” (list) command lists the contents of the current directory. When used from a terminal, it generally uses colours to differentiate between directories, images, executable files etc. As you can see, the prompt reappears at the end.

Adding options

Like practically all commands in Linux, you can add options to the “ls” command to alter its output or influence its behaviour. An option is preceded by a dash (eg, “ls -a“). Try out the following variations of the ls command, to see different forms of output:

ls -l
Produces a “long format” directory listing. For each file or directory, it also shows the owner, group, size, date modified and permissions
ls -a
Lists all the files in the directory, including hidden ones. In Linux, files that start with a period (.) are usually not shown.
ls -R
Lists the contents of each subdirectory, their subdirectories etc (recursive).

When you want to give more than one option, you can group them together with a single dash. For example, the command “ls -al
” is the same as “ls -a -l

Some options consist of a word (or words) instead of a letter, and have two dashes instead of one. For example, the command “ls -l --full-time” displays the date and time of modification in full.

Finally, some options may also have a value. For example, “ls -l --sort=size
” sorts the listing by size.

Adding parameters

Apart from options (which are preceded by one or two dashes), you can also specify
parameters, such as filenames, directory names and so on.

For example with the “ls” command, if you don’t specify any parameter, it will list the contents of the current directory. However, you could instead give it a parameter specifying what to list. For example if you type in “ls /usr“, it will list the contents of the “/usr” directory. You can specify more than one parameter, but we’ll see more about that later.

Obtaining Help

Contrary to popular belief, commands and programs in Linux tend to be very well documented – usually more so than in Windows, which tends to document commands only if they are very popular, and then only document up to a certain “level”.

This section outlines the main ways of getting such information. Note that many distributions, such as Mandrake, RedHat and SuSE, have provided a graphical user interface which can be used to access such information using something like a specialised browser.

The “man” command

Almost every command in Linux has online help available from the command line, through the “man” (manual) command.

Try it now – type in “man ls“. The resulting page will describe the command, then describe every option, then give further details about the program, the author, and so on. This information is shown using the “less” command (which we’ll describe later on). For now, it is sufficient to know that you can use the up and down arrow, PgUp and PgDn keys to move around, and the Q key to quit.

The “info” command

Another source of online help is the “info” command. Some Linux commands may supply both “man” and “info” documentation. As a general rule, “info” documentation is more verbose and descriptive, like a user guide, while “man” documentation is more like a reference manual, giving lists of options and parameters, and the meaning of each.

Try typing “info ls” now. The method for moving around in “info” is quite similar to “man” – you can also use the arrows and PgUp/PgDn to move, and Q to quit. The main difference is that info pages can contain “menus” of links which lead to other pages. To follow a link, move the text cursor to it with the arrow keys, and press Enter.

The “–help” option

Most (but not all) programs have a --help option which displays a very short description of its main options and parameters. Try typing “ls --help” to see. This will produce more than one screenful of information, so you’ll have to use the terminal’s scrollbar to see what was displayed.

The “–help” information rarely says anything that isn’t also found in the “man” documentation, so it’s rarely needed, except in a tiny number of programs which do not supply any other form of documentation.

The Linux Documentation Project

The L.D.P. is a project which collects not only all the man and info pages, but has a huge collection of longer guides, howtos and mini-howtos on a wide variety of topics. Unlike man and info pages, these howtos are not about a specific command, but rather how to accomplish a particular task. When you’re looking for information on the internet, this is probably the best place to start.

Linux Directories

Linux Directories

As you probably already know from working in graphics mode, in Linux the directories (aka “folders”) use the slash (/) as a separator (Windows uses backslash (\) ). In other words it works just like websites or ftp servers.

Any directory which starts with a slash, such as “/usr/bin”, means it is an “absolute” name – the name specifies the entire sequence of directories from the “root” directory (/) up to the specific directory being requested (bin). Thus, it doesn’t matter which directory is the “current” directory when you specify that name, it will always point to the /usr/bin directory.

On the other hand a directory which does not start with a slash is relative to the current directory. For example the directory “bin” will point to different directories depending on whether you are in the root directory (in which case it will point to “/bin”), in the “/usr” directory (in which case it will point to /usr/bin) or in the “/usr/local” directory (in which case it will point to /usr/local/bin).

The same applies to files – if you specify “file.txt” it is assumed to be in the current directory, while if you specify “/tmp/file.txt” it will always point to “file.txt” in the temporary directory.

Two special directory names are the current directory, represented by a single period (.) and the parent directory, represented by a double period (..). Thus, if you are in the /home/sandbox directory and type in ls .., it will list the contents of the parent directory, which is /home.

Some System Directories

Below is a list of some common directories that are found in Linux and Unix systems, and what they are used for.

/
This is the root directory, inside which all other directories reside
This is similar to the root directory of a drive in Windows (C:\),
except that in Linux even different hard disks reside within this root.
/bin
This stands for “binary”, and contains program (executable) files.
This (and other “bin” directories) is where commands such as “ls” can be
found.
In Windows, the c:\windows\commands holds some of command-line programs, but others are scattered in various other directories.
/dev
This stands for “devices”. It contains a number of special pseudo-files that are used to access the physical hardware that make up, or are connected to, your computer. For example the parallel-port would be a file called “lp0” in this directory, while the hard disk would be “hda”, and its first partition would be “hda0”.
Windows/DOS uses a similar method, however in Windows these are not in any particular directory. Devices have names likeLPT1COM1 or CON – any time you try to access a file with that name from any directory, you will get the parallel printer, serial port or console, respectively.
/etc
This is where (almost) all system-wide configuration information is stored. Almost all configuration information is stored in text files, so you can go into this directory and have a look around with a text viewer if you like. Some of the files are quite cryptic though.
There is no equivalent in Windows, where configuration data can be stored anywhere, including the registry, INI files and other data files in various directories.
/home
This is where users’ home directories are usually found. Thus, if you created a user called “sandbox”, there will be a directory with the same name in this directory, which will be that user’s home directory.
The nearest equivalent in Windows is c:\windows\profiles, where some user-specific data is held, together with c:\My Documents, where user-created documents go. However other data can be written in many other directories.
/lib
This is where the library files are found. Libraries are files containing reusable functions and routines for programs to use.
There is no equivalent in Windows/DOS.
/mnt
This is where storage devices other than the hard disks are mounted. This directory usually contains subdirectories called “cdrom”, “floppy”, etc., which – when these devices are mounted – show the contents of the CD-ROM or floppy disk respectively. Your Windows drives may also be automatically mounted in this directory.
There is no equivalent in Windows/DOS.
/opt
This is where optional components of the system are installed. Products such as KDE, Gnome and Oracle may be installed into this directory.
The nearest thing in Windows is the c:\Program Files directory.
/tmp
This is a temporary directory. All files placed in here will automatically be deleted eventually.
The equivalent in Windows/DOS is c:\windows\temp.
/usr
Contains a copy of most of the directories in the root. For example, there is a “bin” directory containing programs, a “lib” directory containing libraries, etc. Usually, “core” Linux files are contained in the root direcories, while “non-core” files are in the “/usr” subdirectories.
There is no equivalent in Windows/DOS.
/var
Stands for “various”. Among the files stored here are the system log files, spool files and other data files.
There is no equivalent in Windows/DOS.

Directory Commands

Here are the most common commands to work with directories.

mkdir new-directory-name
Creates a new directory, “new-directory-name”
cd directory-name
Goes to the specified directory, making it the “current directory”
cd
When you don’t give a directory name, goes to your “home” directory.
rmdir directory-name
Removes (deletes) the directory. As a safety measure, the directory must be
empty before it can be deleted.
pwd
Displays the current directory.
ls directory-name
Lists the contents of the directory.

The following sequence of commands (and results) demonstrates
the above commands. For clarity, the prompt has been coloured grey.
After displaying the directory contents, a new directory called “testing” is
created and the directory listed again. Then we go into the new directory,
display the current directory, go back to the “home” directory, and display
the current directory again. Finally the “testing” directory is removed.

sandbox@laptop:~ > ls
KDesktop public_html 
sandbox@laptop:~ > mkdir testing
sandbox@laptop:~ > ls
KDesktop public_html testing
sandbox@laptop:~ > cd testing
sandbox@laptop:~/testing > pwd
/home/sandbox/testing
sandbox@laptop:~/testing > cd
sandbox@laptop:~ > pwd
/home/sandbox
sandbox@laptop:~ > rmdir testing
sandbox@laptop:~ >

Linux Files

Here are some of the common commands to work with files.

cp filename1 filename2
cp filename1 filename2 filename2 (etc) directory
Copies a file, from filename1 to filename2 or (second form) copies one or more files into the specified directory. Warning: if the destination file already exists, it will be overwritten.
mv filename1 filename2
Renames a file, from filename1 to filename2. Warning: if the second file already exists, it will be overwritten.
mv filename1 filename2 filename2 (etc) directory
Moves one or more files into the specified directory. Warning: if the directory already contains files with the same names, they will be overwritten.
less filename
Displays the contents of the specified file onto the screen, allowing you to use the arrow keys, PgUp/PgDown etc to move around (like the “man” command).
file filename
Displays the file-type by examining its contents, with a very high degree of accuracy.
locate file-or-directory-name
searches for a file or directory in the entire hard disk and displays all the places it’s found. You can also specify a partial name or a section of the entire path.

Wildcards

Wherever you can specify a file or directory name in Linux, you can use wildcards.By using one or more special symbols, the shell will find those files which match a pattern, and place them on the command line instead of the pattern itself. The word “wild card” refers to the “Joker” in a pack of cards, since this card can stand for any other card in many card games. In the same way, the “wildcard” character can stand for other letters and characters in a filename.

Testing Wildcards

To get the hang of wildcards, the best thing to do is to go to a directory which is full of files and try using the “ls” command with the wildcards as arguments. As we saw before, the “ls” command can take a parameter which tells it what to display. Instead of giving it a directory, we’re going to pass it a list of all filenames to display. This list will come from the wildcard patterns which we will see below.

So, before you continue, in the terminal window type the command “cd /usr/bin“. This will switch to the main directory containing the operating system commands. It’s full of files, so it’s ideal for our experiments.

sandbox@laptop:~ > cd /usr/bin
sandbox@laptop:/usr/bin >

The * Wildcard

The first wildcard is the asterisk (*). The asterisk stands for zero or more other characters. By placing this wildcard at the beginning, middle or end of a pattern, you can build a pattern which has the rest of the pattern at one or either end. For example the pattern “*txt” means any sequence of letters which ends with “txt”.

Below is a table of patterns, and an example of which filenames would match such a pattern, and others that would NOT match.

Pattern Matching files NON-matching files Why not
*.txt File.txt
another-file.txt
txt
File.TXT
File.txt2
txtfile
Filetxt
TXT is uppercase
ends in “2”
“txt” is not at the end
no period (.)
*txt File.txt
another-file.txt
txt
Filetxt
File.TXT
File.txt2
txtfile
TXT is uppercase
ends in “2”
“txt” is not at the end
*txt* File.txt
another-file.txt
txt
Filetxt
File.txt2
txtfile
File.TXT uppercase

The ? wildcard

While the * wildcard could stand for zero or more letters or characters, the ? wildcard stands for exactly one.
Thus, a pattern of “???” stands for filenames which are exactly three characters long. The pattern “x??” matches
any three-letter filename which starts with “x”.

The [] wildcard

The square brackets are used to contain a set of characters to match. For example, the pattern “[ABC]*” matches any
filename which starts in one of the letters A B or C, followed by zero or more characters.

If the first character is an exclamation mark (!) or caret (^), then the pattern matches any character except those
given. Thus, the pattern “[^x]*” means any filename except those starting with “x”.

Instead of individual letters, the set can contain a range. For example, the pattern “[A-Z]*” means any filename which starts
with an uppercase letter between A and Z inclusive, followed by zero or more other characters, while “[A-Za-z123]” means
a single character which is an uppercase or lowercase letter, or the digits 1, 2 or 3.

How wildcards work

There is a very significant difference between the way that Windows handles wildcards and how this is done in Linux
or other Unixes. In Windows, the program or command being executed receives the wildcard expression intact. If a program
was not designed to cater for wildcards, it will try to open a file called (say) “*.txt”.

In Linux, on the other hand, it is the bash shell that does all the work. It takes the pattern containing wildcards, convert it into
a list of matching filenames, and pass that to the program in place of the pattern. The table below shows how certain commands
would be “translated” by bash. The actual filenames depend on the contents of the directory so they could vary.

Original command What actually gets executed
ls ls
ls y* ls yacc ybmtopbm yes ypcat ypchfn ypchsh ypmatch yppasswd ypwhich yuvsplittoppm yuvtoppm
ls ?a? ls cal man tac
ls blubble* ls blubble*

Note the last example – bash could find no file with that name so the pattern, including wildcards, is passed to the program “as is”.

This makes things easier for the programs themselves, because they only need to cater for lists of filenames on their command lines. However, there are a few techniques that one could do in MS-DOS that cannot be done in Linux. For example in MS-DOS, one could enter the command “copy *.doc *.bak” – which would copy all files ending with “doc” into an equivalent filename, but ending in “bak”. In Linux, that command would be translated to something like “copy file1.doc file2.doc file3.doc file2.bak file4.bak” – giving a totally different and
probably undesired result. Actually this technique no longer works well in Windows due to the introduction of long file names.

Wildcards with directories

Wildcards with Linux work on directories too. For example, the pattern “*/file.txt” means, all files called “file.txt” in any subdirectory.

Hidden files

Wildcards will not match hidden files unless the wildcard pattern itself starts with a period. Thus, the pattern “.*” matches all hidden files (hidden files are files which start with a period, such as .profile or .kde2)

Typing Tricks

When you’re in the bash prompt, you can use the up- and down-arrow keys to recall previously typed commands.

You can also press Ctrl-R and start typing part of another command to find the last command that contains the letters you are typing. Thus if you want to find the last change-directory, type “[Ctrl-R]cd“, and the command line will display the last “cd” command you typed.

If you start typing a filename or directory name, you can press [Tab] and bash will complete the file or directory name for you, assuming that such a file exists and is the only one that starts with the typed-in part. For example, if you type “ls br[Tab]“, bash will complete the filename to “brushtopbm“, if this file exists and is the only file starting with “br”.

Redirecting Output

Most programs, when executed, display lots of text on the screen. You can save this text into a file if you want to retain it, or process it further. To do this, you use the redirection operator, “>”.

For example, suppose that you want to save a copy of a directory listing into a file. You would type in the complete “ls” command, followed by “>” and the name of the new file to be created.

In the following example, we will list the contents of the /usr directory in long format, and write this listing into a file in our home directory (since we cannot create files in the /usr directory).

sandbox@laptop:~ > cd /usr
sandbox@laptop:/usr > ls -l > ~/usr-listing.txt
sandbox@laptop:/usr > cd
sandbox@laptop:~ > ls
KDesktop public_html  snapshot1.png  usr-listing.txt
sandbox@laptop:~ >

Note the second line. First we have the command to list the directory “ls -l“, then we have the redirection symbol “>” telling bash to dump all the results into a file, and finally we have the name of the new file to create (~/usr-listing.txt). As you may recall, the “~” symbol stands for the user’s home directory.

The symbol we gave so far (>creates a new file with the specified name. If that file already exists it is first emptied before it receives the program’s output. If you give a double-angle-bracket (>>), the output is appended to the end of the file if it already exists. This can be used to redirect the output of several commands into one file.

Piping output to a program

When we used redirection, above, the output of a command was sent into a new file. With pipes, this output of one program is instead sent to the input of another program. This second program processes the output of the first program, and may produce its own output based on it.

Let us try this out. The “cat” command displays the contents of a file. The “/etc/services” file contains a list of recognised TCP/IP services. Try displaying this file now – it should show you many screenfuls of text.

Note that in the following example only a small number of lines are shown: your system will have much more.

sandbox@laptop:~ >cat /etc/services
#                 0/tcp    Reserved
#                 0/udp    Reserved
tcpmux            1/tcp                 # TCP Port Service Multiplexer
tcpmux            1/udp                 # TCP Port Service Multiplexer
compressnet       2/tcp                 # Management Utility
compressnet       2/udp                 # Management Utility
......lots more lines.....
nimhub          48002/tcp               # Nimbus Hub
nimhub          48002/udp               # Nimbus Hub
nimgtw          48003/tcp               # Nimbus Gateway
nimgtw          48003/udp               # Nimbus Gateway

Now, the “sort” command (as its name implies) sorts its input in alphabetical order, and sends it to the output. So, in order to sort the output of the “cat” command, we must pipe it to the sort command. To do this, we use the pipe symbol (|).

sandbox@laptop:~ >cat /etc/services | sort
3Com-nsd        1742/tcp                # 3Com-nsd
3Com-nsd        1742/udp                # 3Com-nsd
3com-amp3       629/tcp                 # 3Com AMP3
3com-amp3       629/udp                 # 3Com AMP3
......lots more lines.....
zip             6/ddp                   # Zone Information Protocol
zserv           346/tcp                 # Zebra server
zserv           346/udp                 # Zebra server

One very popular use of the pipe is to send the output of such long commands to the “less” command. This command allows you to scroll up and down, or even sideways, to view the complete results.

ps -Hefw | less

The above command line displays the list of processes in the system and uses “less” to allow you to scroll around (using the arrow keys and PgUp/PgDown), and Q to quit.

It is possible to string together several commands separated by a pipe. For example we could use the “cat” command to display the contents of the services file above, pipe that to the “sort” command to sort it, then pipe the results of the sort command into the “tail” command to pick out only the last 50 lines, and finally pipe those 50 lines to the “less” command to scroll around in the results.

cat /etc/services | sort | tail -n 50 | less

Environment Variables

In Linux (as well as MS-DOS, Windows etc), each program has a number of “environment variables”. These are a series of “settings” which can be used to control certain aspects of the system, user preferences, etc. Whenever a program is executed it receives a copy of the environment variables from the process which executes it. So, if you run a program from the bash shell, that program will receive a copy of all the environment variables from the bash shell itself.

In the following example, I am using the “date” program to display the date. However, I am changing the language-code each time. The “export” command is used to set a new value in an environment variable, or to create it if it doesn’t exist.

sandbox@laptop:~ >date
Thu Aug  2 04:35:55 CEST 2001
sandbox@laptop:~ >export LANG=it_IT
sandbox@laptop:~ >date
gio ago  2 04:45:05 CEST 2001
sandbox@laptop:~ >export LANG=mt_MT
sandbox@laptop:~ >date
Ħam Awi  2 04:45:55 CEST 2001

You can use the “set” command to view all your current environment variables. Since there is usually more than one screenful, you might want to pipe that to the “less” command to view it one screen at a time.

You can include the value of any environment variable within a command by preceding it with a dollar sign. For example, the command “ls -l $SHELL” will produce a long-listing of the bash program itself, since the SHELL environment variable contains the location of the current shell.

Among the most important environment variables are HOME, PATH and PS1. The first obviously points to the user’s home directory. It can be used to construct filenames, such as “$HOME/file1.txt“.

The PATH

The second variable, PATH, is a list of directory names separated by colons. This is the list of places that bash will look in when searching for the command names you specify. Thus, when you type “ls”, bash will start searching in these directories one at a time until it finds an executable file called “ls”, and executes it.

You can use “$PATH” within another command to set the PATH to a different value, in order to add a new directory to the PATH. For example, suppose that you create a new directory for programs in your own home directory. You can add this new directory to the PATH so that bash will look inside it too. To do that (assuming your new directory is called “bin”), you could give the following command:

export PATH=$PATH:$HOME/bin

In that command, I am setting the PATH environment variable to the previous value of itself, followed by a colon, and a directory-name formed from the HOME variable and “/bin”.

The PS1 prompt

The last environment variables, PS1, tells bash what to display in the prompt. You can play around with the PS1 variable to produce a prompt that you like. While producing this document, for example, I set up the prompt to contain the required HTML code to colour itself grey.

sandbox@laptop:~ >export PS1="I am \u on \h> "
I am sandbox on laptop>

Note however than any changes you make to the environment variables will be temporary. They will revert to normal next time you open a terminal window, or log on. We will see how to make these settings permanent in a moment.

Environment variables

In Linux (as well as MS-DOS, Windows etc), each program has a number of “environment variables”. These are a series of “settings” which can be used to control certain aspects of the system, user preferences, etc. Whenever a program is executed it receives a copy of the environment variables from the process which executes it. So, if you run a program from the bash shell, that program will receive a copy of all the environment variables from the bash shell itself.

In the following example, I am using the “date” program to display the date. However, I am changing the language-code each time. The “export” command is used to set a new value in an environment variable, or to create it if it doesn’t exist.

sandbox@laptop:~ >date
Thu Aug  2 04:35:55 CEST 2001
sandbox@laptop:~ >export LANG=it_IT
sandbox@laptop:~ >date
gio ago  2 04:45:05 CEST 2001
sandbox@laptop:~ >export LANG=mt_MT
sandbox@laptop:~ >date
Ħam Awi  2 04:45:55 CEST 2001

You can use the “set” command to view all your current environment variables. Since there is usually more than one screenful, you might want to pipe that to the “less” command to view it one screen at a time.

You can include the value of any environment variable within a command by preceding it with a dollar sign. For example, the command “ls -l $SHELL” will produce a long-listing of the bash program itself, since the SHELL environment variable contains the location of the current shell.

Among the most important environment variables are HOME, PATH and PS1. The first obviously points to the user’s home directory. It can be used to construct filenames, such as “$HOME/file1.txt“.

The PATH

The second variable, PATH, is a list of directory names separated by colons. This is the list of places that bash will look in when searching for the command names you specify. Thus, when you type “ls”, bash will start searching in these directories one at a time until it finds an executable file called “ls”, and executes it.

You can use “$PATH” within another command to set the PATH to a different value, in order to add a new directory to the PATH. For example, suppose that you create a new directory for programs in your own home directory. You can add this new directory to the PATH so that bash will look inside it too. To do that (assuming your new directory is called “bin”), you could give the following command:

export PATH=$PATH:$HOME/bin

In that command, I am setting the PATH environment variable to the previous value of itself, followed by a colon, and a directory-name formed from the HOME variable and “/bin”.

The PS1 prompt

The last environment variables, PS1, tells bash what to display in the prompt. You can play around with the PS1 variable to produce a prompt that you like. While producing this document, for example, I set up the prompt to contain the required HTML code to colour itself grey.

sandbox@laptop:~ >export PS1="I am \u on \h> "
I am sandbox on laptop>

Note however than any changes you make to the environment variables will be temporary. They will revert to normal next time you open a terminal window, or log on. We will see how to make these settings permanent in a moment.

Scripts

A shell script is a series of shell commands, similar to the ones given above, contained in a file, which can be given to the bash program to be executed in sequence. Technically, it is a program with “bash” as the language.

You can use any text editor to create your shell script – simply choose one you are comfortable with. If you are using KDE you can use the Advanced Editor (kate or kwrite), or kedit. Do not use a word processing program – these are not text editors, and insert additional formatting information into the text. If, however, you know how to use your word processor to save plain-text file, go ahead and use it.

Start with a new file, and type the following lines.

#!/bin/bash
echo "I am about to list the home directory"
# here is the actual listing:
ls $HOME
echo "Done!"

The first line has a special meaning. It tells the current shell which program should be used to interpret this file. Here, we have given the filename of the bash program. This is so that, if this script is invoked from within a different shell, or from a file-management program such as Konqueror and Nautilus, it will still know that this script requires the bash shell to run it.

The second line is our first command. We use the “echo” command to display a simple line of information.

The third line is a comment. Bash simply ignores it, but it can be useful to a person reading the program in the future to understand what the code is trying to do. While comments are not very useful in such a short, simple script, they are vital in longer and more cryptic ones.

This is followed by two more commands – the “ls” command which takes the directory name as a parameter, and finally another “echo” to display our final declaration of triumph!

When you save this file, it shows up as an ordinary file which cannot be executed. In order to run it, you must make it executable. To do this, you use the “chmod” command. In the following example, I have used “myscript” as the script name – you should change that to whichever name you used.

chmod u+x myscript

That’s it! You can now execute the new script. To do this, enter the folloowing command, again replacing “myscript” with the name you used if different.

sandbox@laptop:~ >./myscript
I am about to list the home directory
KDesktop  myscript  public_html  snapshot1.png  usr-listing.txt
Done!

Note the “./” at the beginning of the file – this is actually a directory name (a single period means the current directory). You can avoid this by placing the new script, and any others you create, in a directory within the PATH environment variable. Then you can enter the command from wherever you are.

Bash’s startup scripts

Bash uses a special, hidden file in your home directory called “.bashrc“. This is essentially the same as any other script, except that it lacks the first line we used above, and it is executed automatically by bash as soon as it starts up. This is where you should place any modified environment variables or other settings for them to be “remembered” by bash every time you log in.

Aliases

Bash allows you to define new commands which translate to other, longer commands. For example, if you type the following…

alias lh="ls -l -a $HOME"

, then the command “lh” will become equivalent to the longer command “lh -l -a $HOME”.

You can even use this to re-define existing commands. For example if you type “alias ls="ls -a"“, then the ls command will always start showing hidden files. You can remove an existing alias by using the unalias command.

You can place these aliases into the .bashrc file as well, to be set up when you load bash.

You can even add “safety features” to certain commands. For example, if you create an alias which makes the copy (cp) command equal to “cp -i”, then the copy command will prompt you every time you are about to copy onto an existing file, instead of overwriting it without warning. However you have to keep in mind that these settings will not exist when you’re working in a different user, until you add them to the .bashrc file of that user as well.

Switching to Root

Occasionally you may encounter a situation where you have to switch to the root user to perform a certain command. Before doing so you should be aware that, while working as the root user you are able to do anything. You can install files to the system areas, or delete all the data on your hard disk. Be careful what you do while logged in as root.

The command to switch from one user to another is su – which stands for “set user”. When you enter this command without any parameters it will switch to the root user. Alternatively, you can follow it with the username to which you want to switch. In either case you will be prompted for the password for that user. When you enter the password you will be working as that user – you will be able to access the files that that user can access, perform tasks he is entitled to perform, etc. In the case of the root user, that means all files and all tasks.

Try it first with your own normal username. In this example, I am using ’ramon’ as the other username.

sandbox@laptop:~ > ls /home/ramon
Access denied.
sandbox@laptop:~ > ls 
file1.txt   file2.txt
sandbox@laptop:~ > su ramon
Password: (enter ramon’s password)
ramon@laptop:/home/sandbox > ls
Access denied.
ramon@laptop:/home/sandbox > ls /home/ramon
Desktop  Mail  Documents
ramon@laptop:/home/sandbox > exit
sandbox@laptop:~ >

Note that, while logged in as “sandbox” I could list the contents of sandbox’s home directory but was denied access to ramon’s home directory, and vice versa. Note also that I used the exit command to exit from the “ramon” user back to the sandbox user. This is because the su command creates a new shell process using the “ramon” user.

If we use the su command without parameters we are prompted for root’s password. Once logged in as root we can access all areas.

sandbox@laptop:~ > su 
Password: (enter root’s password)
root@laptop:/home/sandbox > ls
file1.txt   file2.txt
root@laptop:/home/sandbox > ls /home/ramon
Desktop  Mail  Documents
root@laptop:/home/sandbox > exit
sandbox@laptop:~ >

There is a difference between logging in as root (or any user) then loading the terminal, and starting off as another user, loading the terminal and then using su to switch to root. This difference lies in the startup-scripts that get loaded. When you use the “su” command as shown above, it inherits all the settings from the previous shell. Thus, for instance, if you have set up some environment variables they will be carried over and are still accessible as root. This is useful in some compilations. If you use your regular user to compile a program and then switch to root to perform the final installation step, it may be useful to carry over any environment variables you may have set up.

However this has the disadvantage that any commands in root’s startup scripts will not get executed. If you ever need to switch to root (or some other user) and ensure that the startup scripts DO get executed, add a hyphen (-) between the su command and the username parameter, if any. Thus, the commands used above would become:

su - ramon
su -

You will still be prompted for the password, but this time the startup scripts will be executed and any settings from your own shell session will be forgotten – it will be as if you had logged in as that user in the login screen.

Compiling from Source

One of the main reasons why some people may need to use a terminal window is to compile a program which they downloaded from the web. Compiling is a process that takes text source code and converts it into a binary program, which can only run on the specific processor where it was compiled.

All such programs include intructions on how to compile them – these instructions are found in text files called “INSTALL”, “README” or some similar name. Here I will describe one of the commonest routines used.

First of all, assume that the file we downloaded was called “MyProgram-1.2.3.tgz”. Since this file is an archive, like a WinZip file, we have to extract its contents. This is done using the tar command:

tar -xvzf MyProgram-1.2.3.tgz

The options given instruct tar to extract (x) and verbose (v) the zipped contents (z) of an archive file (f) with the specified name. Tar will extract the contents of the archive and display what it is doing.

Next, we have to cd to the newly created directory, which generally has the same name as the archive, minus the “.tgz”.

cd MyProgram-1.2.3

Usually, the next step is to run the “configure” script. This examines your system and produces a compilation and installation script that is ideally suited for your configuration. Another task done at this stage is to check whether you have all prerequisite software installed. The “configure” command produces many lines of output. Unless it ends with an error, these can be ignored.

./configure

When the “configure” script writes the recipe, the “make” command takes the ingredients and bakes the cake. This is often done in two steps. First, you type in “make” to generate the binary program, then you type “make install” to place these binary programs in their proper location on your computer. This last step must be performed using the “root” user, since it requires permission to place files in system areas. You can switch to the root user by using the “su” command.

make
su
make install
exit

If no errors were reported, you’re ready.

Conclusion

That concludes our basic overview of how to use a command shell. There is a lot more, which would be very useful especially when writing shell scripts or performing advanced tasks. However the above should be serve as a good beginning to understand what is going on in the system, as well as a starting point from which you can learn more details about Linux shell programs. Remember that the bash shell has its own man page, as well as longer tutorials and documentation available at the Linux Documentation Project website.

Comments

If you would like to send comments, suggestions etc. please send them to me at
ramon.casha@linux.org.mt, especially if you found
any section hard to understand.

Copyright © 2001 Ramon Casha

CVS: $Id: terminal.xml,v 1.4 2002/05/27 12:19:11 rac Exp $

{ 3 comments… add one }
  • FromTheInside 25/01/2014, 00:24

    Very helpful! Enjoyed reading. Learned some new features. Wrote my first bash-script. Yeah! 😉

    Reply
  • Ton 16/02/2014, 06:23

    i’ve just install cmake on puppy linux precise and use full NOT & (follow the procedure mentioned on http://www.cmake.org/cmake/help/install.html) similart to such above:
    cmake .
    make
    make install

    but, when i check cmake –ver, it still mention the previous version.

    Questions:
    1. do i have to copy some files somewhere else (e.g. to /bin , etc)?
    2. or do i have to extract the tgz to a certain location?

    thanks

    when i type cmake –ver, again display 2.8.7

    OR, perhaps , both version co-exist together??? or, why the latest one doesnt show??? why no error messages during isntallation???

    Reply
  • Shivaji 02/03/2014, 09:11

    Very Very Useful document for me… Thanks

    Reply

Leave a Comment


Time limit is exhausted. Please reload CAPTCHA.

Next Post: