Shell

Within these castle walls be forged Mavens of Computer Science ...

— Merlin, The Coder

Prerequisites

• W1001 The Wiki

The Client and the Server

Client-server-model

A client-server model is a means of distributing resources between the provider of a service and a consumer of that service, the client. You're likely very familiar with this concept though less familiar with the name. One common example of a service provided in this manner is email. You, a user, can open an email client (for example, an app on your iPhone or Outlook on a PC) which then connects to a remote server. The server is responsible for providing several services, including receiving mail on your behalf from other servers, transmitting email on your behalf to others, and storing your email for your later use.

Observe

Observe, Ponder, and Journal: : Section 1

1. Name at least two other services that you personally use which are implemented using a Client/Server Model

There are several advantages to this model:

• Users don't need to personally possess the hardware required to fulfill a particular service, only a means of connecting and communicating with a server which provides the service
• Users don't need to worry about installing and configuring the server hardware
• Users are able to access their data from anywhere in the world (as long as they have internet access)
• The state of a users data remains consistent, regardless of how and from where they access the server

Merlin is built upon this Client/Server Model, so you won't need any particular type of hardware or software, only a device capable of running a web-browser and a terminal emulator (available in Google Chrome).

First Connection

Coder Merlin is a cloud-based server where we’ll be doing most of our exercises. To access Merlin, you’ll need a secure shell client.

If you’re using a Mac or a Linux based device you’ll probably have a built-in client that you can access from the terminal.

If you’re using a Windows-based device, you can access a terminal by installing Google Chrome, and then within Google Chrome typing “Secure Shell Extension” and then installing the extension.

Either way, you’ll need a connection string to access the server. The connection string will begin with your first name and last name separated by a hyphen, for example: john-williams. The rest of the connection string will be @ssh.codermerlin.com

For John Williams, the entire connection string would be:

 john-williams@ssh.codermerlin.com

After you’re connected, you’ll be prompted to enter your password. After a successful logon, you'll land in the "shell".

The Shell

A shell is a user interface which provides access to an operating system’s services. There are two basic types of shells: a “command-line interface” (CLI) and a “graphical user interface” (GUI). You are likely familiar with common GUI’s from Microsoft Windows and perhaps Apple’s macOS. This document will focus on a CLI known as bash, a shell for the Unix operating system by Brian Fox. The shell is an acronym for Bourne-again shell, named for the Bourne shell that it replaces. Login to the Merlin Server as discussed previously. After some introductory text, you’ll see a prompt which will look something like this:

john-williams@codermerlin:~$ 

You’ll see your login name, an at (@) sign, the name of the server (“codermerlin”), a colon (:), a tilde (~), and a $. This will be followed by a flashing block called a cursor. The cursor indicates where new characters will appear as you type. The cursor will continue to flash patiently as it awaits input. We’ll get back to the meaning of the tilde soon.

We'll be using a server running Linux, a Unix-like operating system. Almost everything in Unix is either a file or a process. A file is either a destination or source for a stream of data. A process is a program that is being executed by the operating system. While printers and screens are considered to be “files” in Unix, a more general way of thinking of files is a collection of data that can be referred to by name. Files can be created by users directly (for example, via a text editor) or indirectly, such as the result of running a program like a compiler.

Directories

Directory Hierarchy

Linux File Hierarchy

All files are integrated into a single, hierarchical directory structure that appears as an inverted tree. A directory provides a logical grouping of related files (and potentially other directories). The top of this tree is called the root and is referred to by a slash (/). In the diagram, we see that the root contains five subdirectories (i.e., a directory below the parent directory): bin, etc, home, lib, and tmp. The home directory contains two subdirectories, jack-williams and john-williams. Note that this differs from a Windows file structure as Unix integrates all files into a single directory structure rather than listing different physical storage devices. Thus, the file structure is logical rather than physical. A path describes an ordered transition through one or more directories on the way to a destination (either a directory or a file). As such, note that while both jack-williams and john-williams each have a subdirectory named "projects" these are different directories. The path to the former is "/home/jack-williams/projects" while the path to the latter is "/home/john-williams/projects". Paths which begin at the root (symbolized by the initial slash) are called absolute paths. All other paths are relative paths.

Observe

Observe, Ponder, and Journal: : Section 2

1. It appears that both Jack Williams and John Williams can have a directory of the same name. Do you think the ability to have two different users create a directory of the same name is important? Why?

Display the Working Directory (Where am I?)

The shell uses a concept known as your working directory, which can be thought of as your current directory, i.e. where you are currently located in the file system. If you forget where you are, you can always use the command pwd to print the working directory. Try it now. Type “pwd” followed by <RETURN>. Note that you’ll need to press <RETURN> after any command before the shell will execute the command.

john-williams@codermerlin:~$  pwd

/home/john-williams 

The command will print something similar to “/home/john-williams” and then present you with a new prompt and cursor to indicate that it’s ready and awaiting your input. This directory is your home directory and it has a special symbol, the tilde (~).

Observe

Observe, Ponder, and Journal: : Section 3

1. Why do you think the home directory has a special symbol to represent it?

Create a New Directory

In order to keep our Merlin Missions organized, let’s place them all under a directory named “Merlin”. To make a new directory, we’ll use the command mkdir to which we’ll provide a single argument, the name of the directory that we want to create. Try it now. Type “mkdir Merlin” followed by <RETURN>.

john-williams@codermerlin:~$  mkdir Merlin

Display the Contents of a Directory

How do we find out what’s contained in our current directory? We can use the ls (that’s a lowercase ‘L’ followed by a lowercase ’S’) command to list the directory contents. Try it now. Type “ls” followed by return.

john-williams@codermerlin:~$  ls

The command will print the contents of the current directory (in this case our home directory) which will include the single directory, “Merlin”, that we created earlier.

Change to a New Directory

When we want to change our current directory we use the cd command followed by the name of the directory to which we want to change. Try it now. Type “cd Merlin”.

john-williams@codermerlin:~$  cd Merlin

john-williams@codermerlin:~/Merlin$  

Observe

Observe, Ponder, and Journal: : Section 4

1. Did the prompt change? If so, how? Why is this useful?

Note that our prompt has now changed. Before the command was executed, the prompt contained a single tilde indicating that our home directory was our current directory. After executing the command our prompt now includes a tilde, followed by a slash, followed by “Merlin”. This indicates that we are in the directory named “Merlin” below our home directory. The slash is a delimiter separating the levels of our hierarchy.

Let’s try the pwd command again.

john-williams@codermerlin:~/Merlin$  pwd

We can now see that from the root, we’re three levels down:
/ (root)
home
john-williams
Merlin

The cd command has a useful shortcut. If we type the command without an argument, it will take us back to our home directory. It’s the equivalent of "cd ~". Let’s try it now:

john-williams@codermerlin:~/Merlin$  cd

john-williams@codermerlin:~$  

Now, move back to the Merlin directory by issuing the appropriate command.

Change to a New Directory Temporarily

When we want to change our current directory temporarily (meaning we want to return later to where we are now), we use the pushd command followed by the name of the directory to which we want to change. Try it now. Type “pushd Merlin”.

john-williams@codermerlin:~$  pushd Merlin

john-williams@codermerlin:~/Merlin$  

We can now perform any actions we'd like in this directory and when we're done, we can easily return from whence we came without needing to type in the original directory name. The popd command will look up where we came (when we used pushd) and return us to there. Try it now. Type popd.

john-williams@codermerlin:~/Merlin$  popd

john-williams@codermerlin:~$  

Tab Completion

Very often, the shell can help us to complete file and directory names when we begin to type the name of a file (or directory) and then type TAB. Try it now. Type "cd M" and then press TAB. The shell will complete the rest of the name by appending "erlin" to what you’ve already typed. Then press ENTER. Because the completion is not ambiguous (i.e. there is only one solution) the shell will complete the name. If there were multiple possibilities, the shell would complete as much as possible and then beep. At this point, we could press TAB twice in succession to see the possible completions. Let’s try this. Within the Merlin directory, create three new directories: hello, hello2, and hello3:

john-williams@codermerlin:~/Merlin$  mkdir hello

john-williams@codermerlin:~/Merlin$  mkdir hello2

john-williams@codermerlin:~/Merlin$  mkdir hello3

Then, type “cd h”:

john-williams@codermerlin:~/Merlin$  cd h

Now, press TAB. The shell will complete as much of the name as possible and then beep. (Depending on your terminal you may not hear the beep or you may see a flash.)

john-williams@codermerlin:~/Merlin$  cd hello

Now press TAB twice in succession. The shell will display matching possibilities and then provide a new prompt followed by the previous entry:

hello/ hello2/ hello3/ 

john-williams@codermerlin:~/Merlin$  cd hello

You now have the option of pressing ENTER to accept "hello" or adding some more characters (in this case “2” to form “hello2” or “3” to form “hello3”) and then pressing ENTER to execute the command as-is or TAB to search for additional completions. In our case, let’s press ENTER to descend into the "hello" directory:

john-williams@codermerlin:~/Merlin$  cd hello

john-williams@codermerlin:~/Merlin/hello$  

Recall Previous Commands

Sometimes it's expedient to quickly recall a previous command without the need to re-type it. The most common means of accomplishing this is to use the up-arrow. Let's try it:

john-williams@codermerlin:~$  ls

Now, let's repeat the command, but rather than type ls again, try typing the up-arrow:

john-williams@codermerlin:~$  ↑

Try pressing the ↑ and ↓ several times.

Observe

Observe, Ponder, and Journal: : Section 5

1. What happens each time you press the ↑?
2. What happens each time you press the ↓?

Files

Echo to the Screen

The echo command echoes the argument to the output. Let’s try it. Type: echo "Hello, World!" The echo command dutifully echoes the argument to the screen. Try it now:

john-williams@codermerlin:~/Merlin/hello$  echo "Hello, World!"

Hello, World! 

Create an Empty File

The touch command creates a new, empty file, if such a file does not already exist. If such a file does already exist, the access and modification times will be set to the current time. Try it now:

john-williams@codermerlin:~/Merlin/hello$  touch newFile.txt

john-williams@codermerlin:~/Merlin/hello$  ls

newFile.txt 

john-williams@codermerlin:~/Merlin/hello$  cat newFile.txt

john-williams@codermerlin:~/Merlin/hello$  

Redirection

Remember that we said that almost everything in Unix is either a file or a process and that the screen could be considered as a file. We can redirect the output from echo to another file, one that we'll store in the file system. We do this with with the > (greater-than) symbol. Normally, commands will send their output to a device called stdout. This is what the echo command did when we used it earlier. In our case, stdout is printed to our terminal. Let’s try to redirect our output to a file. Type: echo "Hello, World!" > out.txt:

john-williams@codermerlin:~/Merlin/hello$  echo "Hello, World!" > out.txt

Observe

Observe, Ponder, and Journal: : Section 6

1. Was "Hello, World!" printed to the screen as before? If not, why not? What do you think happened to the text?

No output was displayed; we just received a new prompt. Let’s use the ls command to take a look:

john-williams@codermerlin:~/Merlin/hello$  ls

newFile.txt out.txt 

We can see that the directory contains a single file, "out.txt" which was created by the echo command. How can we see what's contained within this file?

Quickly View File Contents

To quickly view the contents of a file we can use the cat command. The cat command concatenates files together and then sends the output to stdout. (Remember that in our case this is the screen.) Let’s try it. Type "cat out.txt".

john-williams@codermerlin:~/Merlin/hello$  cat out.txt

Hello, World! 

Append a Small Amount of Text to an Existing File

If we already have an existing file with some text we can append (add text to the end) of this file by using echo and redirection as before but with two "greater-than" symbols adjacent to one another. Try it now. Type: echo "Hello, again!" >> out.txt

john-williams@codermerlin:~/Merlin/hello$  echo "Hello, again!" >> out.txt

john-williams@codermerlin:~/Merlin/hello$  

Let’s move back up to our home directory. Remember that we can do this by typing "cd" without any arguments.

john-williams@codermerlin:~/Merlin/hello$  cd

john-williams@codermerlin:~$  

Directories (Revisited)

Helpful Hint

There are useful shortcuts that you can use when specifying a path.

• "." (a single period or dot) refers to the current directory
• ".." (a double period, or two adjacent dots) refers to the parent directory, relative to wherever you happen to be in the path up to this point

View the Directory Hierarchy

Sometimes it’s helpful to obtain an overview of our directory hierarchy. We can do this with the tree command. Let's try:

john-williams@codermerlin:~$  tree

.

└─ Merlin

   ├─ hello

   │  ├─ newFile.txt

   │  └─ out.txt

   ├─ hello2

   └─ hello3

Observe

Observe, Ponder, and Journal: : Section 7

1. How does the ls command differ from the tree command?
2. How do you choose to use one over the other?

Remove a Directory

Let's remove the two directories that we won't be using anymore. We can do this with the rmdir command. Move to the Merlin directory and then remove the hello2 and hello3 directories. Then do a directory listing to see what’s left.

john-williams@codermerlin:~$  cd Merlin

john-williams@codermerlin:~/Merlin$  rmdir hello2

john-williams@codermerlin:~/Merlin$  rmdir hello3

john-williams@codermerlin:~/Merlin$  ls

hello 

Clear the Screen

In some cases it’s useful to clear the screen. (For example, this might be useful if you sense your guide creeping up behind you when you were doing something that you weren’t supposed to be doing. 😁) We can do this with the clear command. Type clear and then press RETURN.

john-williams@codermerlin:~$  clear

Jobs

A program is an executable file which contains a series of instructions which can be executed by the computer. A process is a program which is being executed on the computer. The shell manages processes using the concept of a job. You can obtain a list of jobs and suspend and resume jobs. This functionality enables you to maximize your use of time rather than wait for the prompt to return as the shell waits for the execution of a job to complete.

Identifying Jobs

The shell provides several different ways to identify a job:

• %n - The specific job number is preceded by a percentage sign
• %+ - A shortcut to specify the current job
• %- - A shortcut to specify the previous job

Listing Jobs

Jobs may be listed with the jobs command. Let's try an example by executing a few long-running jobs. The sleep command simply waits before returning. Let's try it:

john-williams@codermerlin:~$ sleep 3

Let's try it again:

john-williams@codermerlin:~$ sleep 5

Observe

Observe, Ponder, and Journal: : Section 8

1. What do you think the argument to sleep specifies?
2. Are you able to prove your hypothesis? How?

These same jobs can be executed in the background (more on this below) simply by specifying an ampersand (&) after the command.

Let's try it:

john-williams@codermerlin:~$ sleep 60 &

Observe

Observe, Ponder, and Journal: : Section 9

1. How is the behavior different from what you expected?
2. What do you think it means to run a job in the background?
3. What do you think the number displayed in brackets specifies?

Let's run several jobs in the background and then view a list of our jobs:

john-williams@codermerlin:~$ sleep 40 &

john-williams@codermerlin:~$ sleep 30 &

john-williams@codermerlin:~$ sleep 20 &

Now, let's use the jobs command:

john-williams@codermerlin:~$ jobs

Repeatedly execute the jobs command.

Observe

Observe, Ponder, and Journal: : Section 10

1. What do you observe as you execute the jobs command?
2. How do you explain your observations?

Terminating a Process

While a process is executing, we may decide that we want to terminate it. Perhaps it is taking too long or we began executing it with the wrong arguments. We can often terminate the program with a special key sequence, CONTROL-C.

Let's try an example. We'll execute the cat command without any arguments, which will cause it to wait for something to be entered on the console.

jane-williams@codermerlin:~/Merlin$ cat

The process is now waiting. If we realize that we made a mistake (perhaps we wanted to print the contents of a file and forgot to specify the file's name) we can terminate the process. Try it now by pressing CONTROL-C.

Foreground Processes

Generally, when we execute a program we run it in the foreground. This means that the our shell will execute the process and we won't be able to execute the next command until the currently running foreground process exits. This is ideal for short, interactive programs. Whenever we execute a program, running that program in the foreground is the default. For example, the find command can search for all files and directories within our home directory. Let's try it:

jane-williams@codermerlin:~/Merlin$ cd

jane-williams@codermerlin:~/Merlin$ find

We'll likely see a long list of files and directories quickly scroll by. But what if we want to generate a list of all of the files and directories on the entire server? This can take a while. Let's try it, just for fun:

jane-williams@codermerlin:~/Merlin$ find /

Eventually we'll encounter a permissions error and the process will halt. But if you don't have the patience to wait, remember that you can use the key sequence CONTROL-C to terminate the process immediately.

Background Processes

When we have a complex process that will take a long time to execute we don't want to be blocked from executing other commands while we wait for the long-running process to finish. In these cases, we can execute the long-running process in the background. Running the program in the background means that the process will continue to execute but we'll be able to enter new commands immediately.

Display a Manual

Finally, let's try out one more very helpful command. If we want to learn more about a command we can read its manual. We do this with the command man. As an example, look at the man page for the emacs command. Type man emacs.

john-williams@codermerlin:~/Merlin$  man emacs

The screen will be filled with information about the emacs command. While a man page is displayed we can use the following keys to navigate:

• f Typing the f key will move us forward one page in the manual
• b Typing the b key will move us backward one page in the manual
• q Typing the q key will quit (exit the manual)

Exit the manual now.

Exiting the Shell

In general, there's no need to exit the shell. It's almost always sufficient to close your terminal emulator. An advantage to this approach is that the next time you connect to the server you'll be in exactly the same place where you left off. However, on rare occasions the terminal emulator can become "confused" and display characters in odd places. In such a case, there's still no need to exit; instead, use the reset command.

References

• https://en.wikipedia.org/wiki/Operating_system
• https://en.wikipedia.org/wiki/System_software
• https://en.wikipedia.org/wiki/Computer_hardware
• https://en.wikipedia.org/wiki/Application_software
• https://en.wikipedia.org/wiki/Shell_(computing)
• https://en.wikibooks.org/wiki/A_Quick_Introduction_to_Unix/Files_and_Processes
• https://en.wikibooks.org/wiki/A_Quick_Introduction_to_Unix/Directory_Structure
• https://unix.stackexchange.com/questions/89114/what-does-the-jargon-filesystem-mean
• http://www.aboutlinux.info/2005/05/job-control-in-linux.html
• https://www.cis.gvsu.edu/linux-tutorial-job-control/
• https://unix.stackexchange.com/questions/4214/what-is-the-difference-between-a-job-and-a-process