Lab 01 -  Introduction to the command line 


0. Cut and paste in Linux... usually.


If you highlight text by clicking and dragging over it,
pressing the middle mouse button (scroll wheel, both buttons if none, or right click on WSL)
will paste the text at the cursor.




1. Getting started
Open a terminal window.
If using WSL, start your Xserver.
At the prompt, type


xeyes


and press ENTER.
A pair of eyes should appear that track your mouse pointer.  
Install xeyes if it is not installed.




2. Job control
Close xeyes by clicking the x in their window bar.


Back at the terminal, press "up arrow". 
The xeyes command should appear.
Press ENTER (eyes should appear again).
While holding down the CONTROL key, press z.
This key combination is denoted as Ctrl-z, or ^Z .


The eyes will no longer track your mouse.  
The xeyes program has been stopped, and the command prompt has returned.


fg


will bring xeyes alive again into the ForeGround.
Ctrl-c (^C) will kill them.


Bring up xeyes again, this time using "tab completion" by typing
xey and pressing TAB.  xeyes should appear on the line. Press ENTER.


press ^Z, and then enter


bg


The prompt returns, but your xeyes are alive.  
xeyes is running in the BackGround.


jobs


will show you what jobs are running. 
The job number associated with xeyes is likely [1].

You can kill xeyes, or bring them into the foreground via


kill %1
or
fg %1


You can set a process to run in the background from the start with the &


xeyes &


They will run in the background.  
A way to list the processes running is


ps


You'll get a small list that likely includes bash, xeyes and ps.
Every process has a process id (PID), usually not a small number like 1 or 2.


kill 51914


(Replace the above number with the PID assocaited with xeyes.)

Hard-to-kill jobs sometimes require the "-9" option.  
For the above, it would have been 


kill -9 51914.


Experiment with various options for ps, like
ps a, ps u, ps aux, ps -a


You can also try the command ‘top’ (press q to quit top).


3. Connecting to a remote machine


Log on to alfven, a machine in the physics dept.


ssh user_name@alfven.physics.mun.ca


where user_name is the prefix to your MUN email.
Answer "yes" if you get a question about being sure about connecting.

Ask about your password.


Once logged in to alfven, the first thing is changing your password with


passwd


You'll be prompted with a couple of questions.
I suggest changing your password to whatever your MUN password is.
The program will tell you if the change has been successful, after which


exit


This will log you out.  
Log back in again, allowing alfven graphics to appear on your machine.


ssh -X user_name@alfven.physics.mun.ca


And try old


xeyes


The eyes should appear on your screen, but they are being generated by alfven.

Kill them and log out (^C, exit).  
You should be back on your own machine.




4. Downloading from the Web


There is an archive for Project 1 on the website.  
Download the file with


wget http://www.physics.mun.ca/courses/p3800/PROJECTS/P1/Project1.tgz


This may complain.  If so, use, as is suggested,


wget --no-check-certificate http://www.physics.mun.ca/courses/p3800/PROJECTS/P1/Project1.tgz


A progress bar should show up. 
After not too long Project1.tgz should be saved to your machine.
The file is a compressed archive containing a directory with various file.
Linux will tell you that.


file Project1.tgz


Unpack it with


tar -zxvf Project1.tgz


The -z options is for compressing/decompressing with gzip
-x is for expand
-v is for verbose -- writing to the screen the files processed
-f is for given the file name (Project1.tgz in this case)






5.  Where are we and what have we done!


So we've downloaded a .tgz file and uncompressed it... 
but where did we put it, and what's inside?


You put it in your home directory, which is somewhere in the Linux directory structure.
To find out where we are currently in the directory "tree", enter


pwd


for “Print Working Directory.” 
The output is the absolute "path" starting from
the root directory /
The path refers to where a file is situated, and includes the sequence of directories in the tree.
The root directory is the starting level for all absolute paths.
A relative path starts at the current directory.
“root” also refers to a special user on a linux system with privileges over everything.


We can "list directory contents" with


ls


and you should at least see Project1.tgz, which we've downloaded, 
and PROJECT1, the directory that the .tgz file contained.


The "ls" command, as most in Linux, have several options.
Play around with the following ones
ls -a (shows "hidden" files and directories that being with a ".")
ls -l (detailed listing)
ls -rl (-r reversed the order)
ls -lrt (my favourite, most recently modified listings appear at bottom)
ls -F (directories are distinguished with a "/" at the end)
ls -d (treat directories like files - don't list their contents)


You can get a description of all the options through MANual pages


man ls


which exist for most commands. 
Press q to exit man and get back to the command line.


Try
ls Project1.tgz
ls PROJECT1


So if the argument is a directory, you get its listing.


Try
ls Pr*
ls PR*


The wildcard "*" stands for any number of any characters in a name.

We can Change Directory into PROJECT1 with


cd PROJECT1


Check where we are with


pwd


You are no longer in your home directory.


The argument to the cd command can be a relative path or absolute path.
In the above, we gave a relative path to PROJECT1, relative to where we were.


Play with these


cd   (go to our home directory)
cd - (go back to directory where you were previously)
cd .. (go up to parent directory of current directory)
cd .  (change to current directory - DOES NOTHING!)
cd / (change to root directory - try ls here!)
cd ~ (~ stand for our home directory)
cd $HOME ($HOME is the same as ~)
cd ~/PROJECT1 (absolute path to PROJECT1)
cd ../../ (relative path - try ls, see your friends' directories)
cd /usr/bin (absolute path - try ls here!)


An argument for cd, a path, can also serve as an argument for ls, e.g.


ls $HOME/..






6. File permissions


cd ~

ls -l


The long listing entry for PROJECT1 on my system is


drwxr-xr-x 2 saika users   4096 Jan  5 22:10 PROJECT1


The first "d" means that it is a directory.
It would be "-" for a regular file.
rwxr-xr-x is the permission string (more on this later)
There are 2 links (1 for a file, 2 + n for directory with n immediate subdirs).
The user who owns the directory is "saika".
The group associate with the directory is "users"
The directory (not its contents, just as a file) takes up 4096 bytes.
It was last modified on Jan  5 22:10.
The file/directory name is PROJECT1


The permissions string is grouped into three sets of three characters.


The first three, rwx, mean the user (u) has read, write and execute permissions.
The second three, r-x, mean the group (g) has only read and execute permissions.
The last three, r-x, mean that anyone else (o -- other) on the system also has read and execute permissions.


Executable for a directory means that it is accessible.
Executable for a file means that it can be run (more on this later).


Let's make the directory non-executable with


chmod -x PROJECT1

ls -l


now yields 


drw-r--r-- 2 saika users   4096 Jan  5 22:10 PROJECT1


and


cd PROJECT1


doesn't work!
Change permissions back.


chmod +x PROJECT1


cd PROJECT1


works! We are now in the PROJECT1 directory.  
Make sure with


pwd

ls -lrt


There are a bunch of files, including some Fortran 90 source files, 
the ones ending in .f90


Compile one using


gfortran integrate.f90 -o integrate


ls -lrt


There is a file called "integrate" and it is executable.

That means we can type


./integrate


And it will run!  Just press ^C to kill it.

There are other files, ending with .sh, that are script files 
and are meant to be executable.


Look at the contents of run_integrate2.sh using


cat run_integrate2.sh 
(dumps, or "concatenates" the file to the screen)


and 


less run_integrate2.sh
(can browse through a file, press q to quit)


The file contains bash scripting code, a list of commands, that the
shell can run.  We can make it executable by the owner (u) and the group (g) via


chmod u+x run_integrate2.sh
chmod g+x run_integrate2.sh


or, equivalently,


chmod u+x,g+x run_integrate2.sh 


Now we can run it with
./run_integrate2.sh


It should produce three columns of numbers.
Other users on the system (who are not part of the group) do not have execute privilege.
You can run scripts even if they are not executable with, e.g.,


bash ./run_integrate.sh


There is another way to specify permission strings, where
r = 4
w = 2
x = 1

To get the permission string

rw-r-x--x

You can either use, on say README.txt


chmod u+rw,u-x,g+rx,g-w,o-rw,o+x README.txt


Or add up the “values” of permissions for each of u (4+2+0), g (4+0+1), and o (0+0+1) and write


chmod 651 README.txt




7.  Creating, deleting, copying, renaming and moving  files and directories


Go to your home directory


cd 


Make a directory (use your name at the end of the name) 


mkdir FAKEP1_YOUR_NAME


Now execute the following commands.  
To observe their effect, use ls -lrt and pwd liberally in between.


cd FAKEP1_YOUR_NAME
cd ..
rmdir FAKEP1_YOUR_NAME    (deletes an empty directory)
mkdir FAKEP1_YOUR_NAME    (make the directory again)
touch funny_file          (creates empty file)
mv funny_file funny_file_renamed  (renames a file or directory)
mv funny_file_renamed FAKEP1_YOUR_NAME (moves a file to a directory)
cd FAKEP1_YOUR_NAME 
ls
cd ..
rmdir FAKEP1_YOUR_NAME  (shouldn't work)
nano funny_file2


This opens a text editor that is fairly intuitive.
Write some text into the file, and then look at the menu at the bottom.
Type Ctrl-X, then hit y for yes, then hit return to keep the default file name.
You should now be back on the command line having created a non-empty funny_file2.


cat funny_file2

cp PROJECT1/README.txt .
(copies the README.txt file that is in the PROJECT1 directory into
the current directory)

ls -l README.txt
(note the timestamp for the file)


touch README.txt  (updates timestamp)

ls -l README.txt

cp README.txt README_copy.txt (create a copy of a file)

cp README.txt FAKEP1_YOUR_NAME (create a copy in a directory)

ls FAKEP1_YOUR_NAME
(this directory should have two files in it)

cp -r FAKEP1_YOUR_NAME COPY_FAKE (this copies a whole direcotry)

ls COPY_FAKE   (all contents were copied as well)
rm -r COPY_FAKE  (delete a directory and all of its contents)



USE rm -r WITH CAUTION!


mv funny_file2 README_copy.txt FAKEP1_YOUR_NAME  (move multiple files to a directory)


8. Submitting a project directory


cd 


You should have a directory called FAKEP1_YOUR_NAME in your home directory.
Let's say this contains your workings for Project 1.  
You need to upload it to alfven, archive it, and submit it.
This is one way to do it.


scp -r FAKEP1_YOUR_NAME user_name@alfven.physics.mun.ca:


Don't forget the colon at the end.
Use your own username.
You will need to supply your password.


ssh user_name@alfven.physics.mun.ca
(supply password)


ls -lrt


You should see the directoy FAKEP1_YOUR_NAME now in your home directory on alfven.
If this were a real project submission, you would go into the directory and make sure everything runs properly on alfven.

Create an archive with 


tar -zcvf Fake_P1_YOUR_NAME.tgz FAKEP1_YOUR_NAME


Copy the archive to a special directory on alfven


cp Fake_P1_YOUR_NAME.tgz /home/Submissions/


ls -l /home


The Submissions directory has interesting permissions.  
You can write to it (copy files to it), but you can't read it.


exit
(logout from alfven)


9. Aliases


Typing out long commands can be tedious.  
You can create a short form of the command with


alias alf="ssh -X saika@alfven.physics.mun.ca"
(use your username at alfven)


Now instead of typing out


ssh -X saika@alfven.physics.mun.ca

you can just use

alf

Use nano to add the following lines to the .bashrc file
in your home directory (create the file if it is not there).
(On macOS, your default shell might be zsh, in which case you should edit the
.zshrc or .zprofile file.)


alias alf="ssh -X saika@alfven.physics.mun.ca"
alias ll="ls -lrt"

Now open a new terminal.  In that new terminal type


alias


The two aliases should appear.


If you update your .bashrc, you can engage the changes without
opening a new terminal by typing


source ~/.bashrc






10.  A couple of short cuts.


While on the command line


^A (Ctrl-A) brings you to the beginning of the line
^E brings you to the end of the line
^U delete the line up to the cursor position