Brief (to be removed)
touch
cat
mkdir
rm (-rf)
cp (-R)
mv
sudo (+ which directories you need to sudo to write to)
grep (-n) (alternatives: ag)
chmod and permission numbers
ssh (and how to ssh into labs)
man
You've already learnt about ls and cd. Now, we'll teach you some more useful (nay, essential) Linux commands.
Don't worry if the following list of commands seems very long at first. After lots of experience with using the Linux terminal you'll gradually start to memorise each command.
Remember, whenever a command needs a file/directory path you can give it as either an absolute path or relative to your working directory.
This is a really easy command. touch <file> simply creates an empty file at the specificed location! Here it is demonstrated:
Note that files don't have to end with an extension like .txt. They can be called whatever you want! Files only have extensions so that programs know what to do with that type of file.
cat <file> prints the contents of that file to your screen. In the following example we've already created a file called hello.txt in our home folder, which contains the text hello world.
mkdir <directory> creates an empty directory at the specified path.
Notice that files appear in white but directories appear in blue! The terminal colours may vary depending on what flavour of Linux you are running.
rm <file> removes that file permanently from your computer. Terminal commands don't use the Recycle Bin! Be very careful when using rm to remove files. In the example below we use touch to create an empty file and rm to remove it.
There's one catch: if you try to use rm on a directory, an error is printed! To remove a directory (and everything inside it) we must add the -r and -f flags.
-rtells the computer to remove the directory recursively (i.e. to remove all subdirectories, subsubdirectories, etc.)-ftells the computer to erase all files/directories without asking for confirmation, no matter what the file permissions are
We can stick these flags together into one string, so typing rm -rf <directory> will remove that directory from the computer. Remember this is a permanent action. In the example below we create a directory with mkdir and remove it with rm -rf.
An added bonus: you can attach any number of files/directories to the end of your rm command, and rm will remove all of them! e.g. rm file1 file2 file3 will remove the files file1, file2 and file3 from your working directory.
cp <file> <directory> copies the file into the specified directory. The original is left intact. Here it is in action:
Similarly to rm, cp doesn't immediately work on copying directories. We must add the -R flag (R = recursive) to copy entire directories. Therefore cp -R <directory1> <directory2> copies the whole of directory1 into directory2.
mv <file/directory> <directory> moves the file/directory given by the first argument into the directory given by the second argument. Unlike cp, mv doesn't need a special flag for dealing with directories; it can handle them in the same way it handles files.
sudo is probably one of the most important commands in the Linux terminal, however we need to give you some background information before explaining what it does.
All Linux computers have a secret user called root, which is allowed to read, write and execute every single file on the computer. However, there are many critical system files littered around which could cause serious damage if modified accidentally. Therefore, your personal user (and all other normal users) are only allowed to read and write to certain directories on the computer, and if you try to modify a sensitive directory (like the directories in /) you're presented with an error.
The way around this is to prepend the word sudo to your command. This executes the command as the root user, allowing you to do pretty much whatever you want, and Terminal will ask for your password before executing the command. However you MUST make sure what you're doing is safe, otherwise you risk permanently breaking your Linux system. Be warned!
In the following example, we write an empty file to the root directory.
In the terminal, whenever you are asked to input your password, the characters you type aren't printed to the screen (although they are registered by the computer). That's why the bit after [sudo] password for docsoc: appears blank, even though we did actually enter our password.
A small note: students actually aren't allowed to use sudo on Huxley lab computers! Presumably the Department has had enough of students messing around with critical system files...if there's something which you think you need sudo for, contact the Computing Support Group (CSG).
ssh is a very useful Linux tool, since it allows you to access any other public Linux computer in the world. The only prerequisites are you need to know the IP address (or DNS hostname) of the computer, and also the username and password for an account on that computer. The format for the ssh command is ssh <username>@<computer>, where <computer> can be either the hostname or the IP address of that computer. After executing this command, you will be prompted for the user's password before being able to remotely access the computer's terminal.
The Department of Computing provides a way for you to ssh into your personal account on the Huxley lab computers. There are 5 Linux servers which can be accessed via ssh, named as follows:
shell1.doc.ic.ac.ukshell2.doc.ic.ac.ukshell3.doc.ic.ac.ukshell4.doc.ic.ac.ukshell5.doc.ic.ac.uk
Logging in to any one of them with your Imperial College credentials will allow you to access your personal account. (When the terminal asks Are you sure you want to continue connecting?, answer yes.)
After you enter your password a very long introductory message is printed, after which you can access your personal account from the terminal!
grep allows you to search for phrases within files. grep -n "string" <file> will print out every instance where "string" occurs in file (the -n flag makes grep print the line number at each instance of "string").
For example, say we have a file called catsdogs which contains the string "cats" on some lines and "dogs" on others.
Running grep -n "cats" catsdogs will print every instance of "cats" in catsdogs, along with the line number. The same can be done for "dogs".
grep can also be used to search for a phrase within a directory, if you add the -r flag. The format is grep -rn "string" <directory>.
chmod allows you to change the permissions for a file or directory. The format is as follows: chmod <mode> <file/directory>.
The mode argument must be given as a 3-digit number, with each digit in the range 0-7. The first digit sets permissions for your own user, the second digit sets permissions for your user group and the third digit sets permissions for everyone.
Each digit sets permissions as follows:
| Digit | Shortcode | Permissions |
|---|---|---|
| 7 | rwx | read, write and execute |
| 6 | rw- | read and write |
| 5 | r-x | read and execute |
| 4 | r-- | only read |
| 3 | -wx | write and execute |
| 2 | -w- | only write |
| 1 | --x | only execute |
| 0 | --- | none |
We've saved the most useful command until last. The man command gives you help on how to use any other Linux command! Type man <commandname> to see the built-in Linux manual entry for that command. All the information on commands we have described in this booklet can also be found using man. This can be especially useful for first year Lexis tests, since you aren't allowed Internet access during the test, so this is your point of reference for any Linux commands.
For example, here's the beginning of the manual entry for ssh:
You can scroll through the manual entry using the arrow keys.