UP | HOME

Navigation
Using the Command-Line
COP-3402

Table of Contents

Getting started

Using ssh

See OS-specific guides in the syllabus

Logging in to eustis

ssh NID@eustis.eecs.ucf.edu

Then enter your password

You can create an ssh key and add it to eustis's .ssh/authorized_keys to login without a password. 

# if you haven't already generated a key, run ssh-keygen
ssh-keygen
# accept the default location (as long as you haven't already generated


# now let's tell eustis about your public key (be sure to use .pub!)

# linux/mac
cat ~/.ssh/id_rsa.pub | ssh NID@eustis.cecs.ucf.edu 'cat >> .ssh/authorized_keys'

# windows (https://superuser.com/questions/1747549/alternative-to-ssh-copy-id-on-windows)
type %USERPROFILE%\.ssh\id_rsa.pub | ssh NID@eustis.cecs.ucf.edu "cat >> .ssh/authorized_keys"

# windows alternative (rsa_id.pub instad of id_rsa.pub)
type %USERPROFILE%\.ssh\rsa_id.pub | ssh NID@eustis.cecs.ucf.edu "cat >> .ssh/authorized_keys"

Ctrl-L to clear your screen

Navigating the File System

Running commands

Type its name and hit <enter>

Where am i?

Print Working Directory: pwd

What's here?

LiSt the directory: ls

Running commands with arguments

Type its name, a space, then the argument.

Arguments are separate by spaces.

What if the argument has a space in it?

  • escape the space with a backslash ~\ ~
  • Use double-quotes (evaluates the contents, e.g., for bash variables and escape sequences)
  • Use single-quotes (does not evaluate contents or escape sequences)

Go to a new directory

Change (working) directory: cd path 

cd /usr/include

What kind of path is this, relative or absolute?

Use eustis's /usr/include as an example file hierarchy

These are where system-wide C header files are installed

Go the parent directory

cd ..

What is the .. directory?

Where are we now? How do we find out?

Get used to running pwd and ls to orient yourself

Go to a subdirectory

cd include

What kind of path is this, relative or absolute?

What if we had done cd ./include instead? What's the difference? Why do we have the . folder?

Making the command-line fast

fast.jpeg

At this point, you are typing a lot, and retyping some of the same paths over and over again.

Tab-completion

One of these most useful features for speed on the command-line

Works for files and directories, command-line arguments, and commands themselves.

You can even create completions for your own software.

Impatience

Don't spend time typing when you can use tab completion

Laziness

Don't try to remember yourself file names, let the shell do it

Example

  • Be sure we are in /usr/include

  • Type the following, but don't hit <enter> 

    ls gnum
  • Hit <tab>

  • The shell completes the file name for you! 

    ls gnumake.h

Multiple completions

  • Type the following, but don't hit <enter> 

    ls gnu
  • Hit <tab>. Nothing seems to happen!

  • Hit <tab> again to see all completions. 

    /usr/include$ ls gnu
gnu/            gnumake.h       gnu-versions.h  
/usr/include$ ls gnu

  • Add the hyphen and hit <tab> to complete. 

    ls gnu-

Completion happens only when there is a unique option. But hitting <tab> again will show all possibilities with the prefix written so far. Continue typing until you have a unique prefix, then hit <tab> again to complete the unique possibility.

Make it a habit to hit tab

  • Confirming existence of a file
  • Avoiding long typing
  • Narrowing down results (tab twice)

Editing commands

Keyboard Description Mnemonic
Ctrl-a Beginning of command (home) a is beginning of the alphabet
Ctrl-e End of command e for end
<backspace> Delete a character behind you it's the backspace key
Alt-<backspace> Delete a whole word behind you alt for alternative version
Alt-b Go back a word b for back, alt for alternative version
Alt-f Go forward a word f for forward, alt for alternative version

Additional commands

Keyboard Description Mnemonic
Ctrl-d Delete a character in front d for delete
Ctrl-b Move back b for back
Ctrl-f Go forward f for forward
Alt-d Delete a word in front d for delete, alt for alternative version

Kill (cut) and yank (paste)

Keyboard Description Mnemonic
Ctrl-k Kill (cut) until the end k for kill
Alt-<backspace> Kill (cut) a word to the back alt for alternative version
Alt-d Kill (cut) a word in front d for delete, alt for alternative version
Ctrl-y Yank (paste) y for yank

Rerunning commands

  • History saves all commands your run
  • history will print them all
  • Last command added to bottom of the list

Navigating the command history

Keyboard Mnemonic Description
Ctrl-p or <up> p for previous Previous command
Ctrl-n or <down> n for next Next command

Searching for a previous command

  • Ctrl-r (r for reverse lookup)
  • CAREFUL: hitting <enter> will immediately run
    • Use ctrl-a or ctrl-e to start editing the command
  • Ctrl-g to cancel search

"Save" commands by commenting them out

  • Ctrl-a, #, <enter>
  • Adds "saved" command to last one in history

Viewing files

Command Mnemonic Description
cat concatenate write file(s) to terminal (stdout)
more ask user to show more write file out, pausing after each screenful
less "less is more" interactive command-line file viewer

cat command

cat /usr/include/stdio.h

less command

less /usr/include/stdio.h

q to quit

Navigating less

  • q to quit (try esc then q if not working, you might have hit other keys)
  • pg-up, pg-down, up, down, left, right
  • ctrl-v, alt-v, j, k, h, l
  • / type pattern, then enter
    • n for next
    • N for previous
  • type a number then G to go to line
  • ctrl-g to show current command-line

The less command has vim-like commands.

Downloading and untarring files

Command Description
wget '<url>' Download a file
tar -xvf file.tar Unpack a tarball (x for eXpand, v for Verbose, f for Filename) 
cd ~/
wget 'https://www.cs.ucf.edu/~gazzillo/teaching/cop3402fall24/files/hw3.tar'
tar -xvf hw3.tar
cd hw3

Editing the file tree

Command Description
touch create file
rm ReMove
mv MoVe
cp CoPy
mkdir MaKe DIRectory
rmdir ReMove (an empty) DIRectory

Removing non-empty directories?

  • Remove contents of the directory first, then use rmdir
  • What's an algorithm to remove a whole directory tree?

You can ask rm to remove non-empty directories and their contents recursively, but it's a bit dangerous.

Review

Orienting yourself

pwd # to show your working directory
ls # to list current directory contents

Moving around

cd /usr/include # to change working directory
cd ../ # can use relative paths to, this puts you in /usr
cd ~ # to go to home directory
cd - # to go to last directory
cd # also to go to the home directory

Tab completion

  • Hit tab to auto-complete file and directory names
  • Hit tab again to show all files with a matching prefix
  • Use tab to quickly find paths

History

  • Hit up (or Ctrl-p) to repeat the prior command
  • Go up and down (or Ctrl-n) to step through the command history
  • Hit enter on a command to rerun it

Modifying the file tree

touch filename # to create an empty regular file
rm filename # to remove a (non-directory) file
mv filename newname # to rename a file
mkdir subdir # to make a new directory
mv filename subdir/ # to move a file
cp subdir/filename copiedname # to copy a file
rm subdir/filename # to remove a file
rmdir subdir # to remove an empty directory

Author: Paul Gazzillo

Created: 2024-10-10 Thu 11:34

Validate