CL readings

[CL] 1–4 except Links, 6

1. le shell

The “shell” is what takes keyboard commands and passes them into the OS to run. Almost all linux distros have bash

Terminal Emulators are programs like konsole or gnome-terminal that let you run the terminal when using a GUI.

Shell prompts may look differnt depending on the distro, but it usually has your username@machinename, then the working directory, and then a dollar sign $

• A # instead of $ indicates superuser privileges. (happens when logged in as the root user or using a terminal emulation that gives superuser privileges)

There are also a bunch of “virtual terminals / consoles” running in the background accessed by pressing CTRL + ALT + [F1 to F6] and when a session is accessed you can switch between them with ALT + [F1 to F6]. ALT+F7 will send you back to the graphical desktop.

• This lets you switch out from a buggy GUI session into a CLI and do all sorts of bug-fixing shenanigans All of our terminals are basically virtual now. Physical terminals looked like this
• These guys simply displayed things and returned what the keyboard inputted. NOTHING else.

2. navigation

pwd, ls and cd

files r in a hierarchy windows is diff cuz it has seperaete file system tree for each storage device. in linux, these storage devices are “mounted” at points in the big chonky tree that unix uses.

The path ~ represents the current user’s home directory. So ~ = home

When refering to the curernt directory with ./, in most scenerios it’s sufficient to just omit it:

• cd ./bin vs cd bin are identical.

Also:

• cd will simply go to the user’s home directory
• cd - will go to the previous working directory
  • cd ~user_name changes the directory to the home directory of the user supplied

Files that start with . are “hidden” (dotfiles).

• Won’t be listed with ls unless you add -a

Linux doesn’t use “file extensions”, but application programs do.

3. exploring le system

ls lists directory contents file determines the file type less views file contents

ls things

ls can take in other pathnames to list another directory:

• ls /usr OR you can list several directories
• ls ~ /usr ..

• Ask someone about thistodo. WHY IS THIS BEHAVING LIKE THIS!?!?!?
  • Wouldn’t ls .. print the contents of my home directory? Then why is it printing all this garbage?!

The option -l for ls adds more details to the output

• The first character is the “type” of file
  • - = regular
  • d = directory
• The next 9 characters detail the permissions.
  • Refer to this

file filename fetches info about the file filename

• in Linux, “everything is a file”

less filename lets you read text files

The directories within linux systems

Directory	Purpose
/	The root directory
/bin	Contains binaries (programs) for the system to boot. Newer linux distros instead use /usr/bin
/boot	Has the linux kernel, initial “RAM disk image” required by drivers at boot time, and the boot loader. - The boot loader runs before the OS starts and initializes hardware things + the kernel.
/dev	Contains “device nodes”. Since devices are also files, this folder is where the kernel maintains a list of all devices it understands
/etc	Contains system-wide config files. Also contains shell scripts that start system services at boot time. ”Everything in this directory SHOULD be readable text” One interesting file includes /etc/passwd which has a list of user accounts
/home	Each user has a folder here, that folder being their ~
/lib	Stores shared library files for system programs. Similar to windows DLLs. Deprecated, so use/usr/lib
/lost+found	Stays empty less something really bad happens with our system
/media	For removable media like USB drives
/mnt	”mount”. Mount points for devices mounted manually (todo what does this mean)
/opt	”optional” software
/proc	Reveals things abt the computers hardware. Managed by le kernel
/root	Home directory for the root acount
/run	Replacement for /tmp. “mounted using tempfs which stores its contents in memory and not on the disk
/sbin	System binaries (programs vital for the system). Deprecated for /usr/sbin
/sys	Contains info about devices by le kernel. It’s similar to /dev but more detailed
/tmp	deprecated. Temp files can be dumped here for the lols. Some distros empty it on reboot
/usr	”likely the largest directory” Contains programs and support files used by regular users
/usr/bin	Contains program executables
/usr/lib	contains shared libraries
/usr/local	For programs not included with the distro but still used system-wide These programs are usually installed in /usr/local/bin
/usr/sbin	System admin programs
/usr/share	Contains shared data by programs of /usr/bin
/usr/share/doc	Most “packages” (todo what is a package) installed on the system will include some kind of documentation. In /usr/share/doc, we will find documentation files organized by package.
/var	Where changing data is likely to be stored
/var/log	Records system activity.
~/.config and ~/.local	Stores user specific details and whatnot.

Symbolic Links

• Aka soft link or sym-link

• These are “a special kind of file”

• In Unix-like systems, you can have files referenced by multiple names

  • You can access the same underlying file using two distinct names within the file system
  • Think of it like shortcuts that go to a common file

• Pretend you have a file that several programs use, call it foo. This file constantly changes version and we want to update the file name to reflect that (i.e foo-v1.8.9). However if we update foo to v1.9.0, we would have to update the references that the several programs have on foo (how could we expect them to know what the new name will be!?)

  • To fix this, you can have a symbolic link, call it go_to_foo (semantically it should be called foo from what I understand) that points to foo-v1.8.9. All the programs will reference go_to_foo and NOT foo-v1.8.9. That way we only need to update what go_to_foo is pointing to to update foo to v1.9.0

• Writing into a symbolic link will update the referenced file. However, deleting a symbolic link can never delete the file itself. Also, if the file gets deleted, a symbolic link can be dangling / broken

Hard links

• This is in chapter 4 but they split it up lol
• the OG way to create links in Unix
• By default EVERY file has a single hard link that gives the file its name. Creating another hard link essentially creates another directory entry for a file. However
  1. They cannot reference a file outside its own file system
     1. A link cannot reference a file not in the same disk partition as the link itself.
  2. Hard links cannot reference directories
• They are “indistinguishable” from the file itself.
  • You can see a symbolic link is a symbolic link when running ls -l
    • todo try this out.
• You can delete a hard link but the file will persist; only when you delete ALL hard links does the file get deleted for real.

ln creates hard + symbolic links

4. Manipulating Files + Directories

Wildcards

• A shell feature!
• Helps specify groups of files
• 

Here are some examples:

• * = all files
• g* = all files beginning with g
• b*.txt = all files starting with b and ending with .txt
• Data??? = Any file beginning with Data and then exactly 3 characters
• [abc]* = any file beginning with a, b, or c
• [BACKUP.[0-9][0-9][0-9]] = starts with BACKUP. then 3 numbers
• [[:upper:]]* = anything beginning with an upper case letter
• [![:digit:]]* = anything not starting with a number
• *[[:lower:]123] a file ending with a lower case letter, 1, 2, or 3.

These also work in some places of GUIs!!

mkdir

• So i was right1
  • You can do mkdir dir1 dir2 dir3
• The ... means that argument can be repeatteddd

cp

• You can either
  1. copy a file / directory into another file / directory
     1. cp item1 item2
  2. Copy multiple items into a directory
     1. cp item... directory
     2. todo how does cp know when the directory starts? What if i want to copy a file into a directory?

#todo card these

mv

mv item1 item2 moves or renames the item item1 to item2 mv item... directory moves any number of items to another directory

-i will ask for permission -u will only move files that don’t exist or can update already-existing files -v for verbose logs

#todo again, card these

rm

AGAINtodo Card these!

ln to create links

• wait he said to ignore this… okay then
• Well anyway it’s like this: ln file link(position) for hard links and ln -s item link for symbolic links.

the hard link and soft link stuff is above in chapter 3 notes cuz i like combining things

[CL] 5.3, 6, 20 for week 2 lab

Chapta 5 (getting a command’s documentation)

help works for all shell built-ins and intends to… help you

[me@linuxbox ~]$ help cd
cd: cd [-L|[-P [-e]] [-@]] [dir]
	Change the shell working directory.
	
	Change the current directory to DIR. The default DIR is the
	value of the HOME shell variable...

	...etc. etc.

The [] square brackets indicate optional items and a | bar indicates mutual exclusivity (you can have one or the other)

In the above example, you can either use -L or -P; if you’re using -P then you can optionally also use [-e]

--help displays usage information for a lot of executable programs

man gives the manual for programs that provide one (aka man page)

• It even takes in a number to get a specific section
• man 5 passwd gives the file format of the /etc/passwd file

Other commands that… might be useful?

• apropos displays appropriate commands given a keyword
  • apropos partition would look for commands who, in their man pages, includes partition
• whatis in 1 line displays the name and description of the man page of a command
• info is an alternative to man pages. Info manuals are shown when running info command.
  • They are “hyperlinked” like websites, where there is a tree and you can go up and down it so as to organize the manual into “sections”
  • e.g. most of the GNU Project’s command line programs are under coreutils so running info coreutils will show all the commands under it

Creating commands with alias

alias name='commands ; go ; here'

• Will create a command named name that runs the commands outlined by it
• idk where the commands for alias r stored though

6. I/O Redirection

cat concatenates files but also spits out the entire file’s contents into stdout

• cat file.txt otherfile.txt
  • it would print both of these concatenated

sort, uniq, rep, are all filter commands. head and tail are similar wc = word count

tee = T = pipes. It basically takes in things from stdin and dumps it to stdout and other places

• You can route the input into 3 files as well as stdout

stdout and stderr are all files (as per the nature of unix’s “everything is a file” nature) however stdout and stderr by default are linked to the terminal and not saved on a disk file

ls -l /bin/usr > ls-output.txt

• Consider when ls-output.txt does not exist
• Then the output will be an error and that is thrown into stderr (i.e. NOT the output file; it will print it to the console)

Running > previous_output.txt will create a new, empty file / clears an existing file.

• When opening for writing, the file will ALWAYS be cleared before any programs do anything
• Think of this: nonexistent > myfile
  • nonexistent is a non-existent command. It WILL throw an error, but it will be too late: myfile will already have been cleared!
  • todo what happens if i use >> ?

>> does redirect, but it does NOT overwrite the file; instead it will append to the output

Group Commands

{ command1; command2; command3; } > logfile.txt

• You can chain commands and this is synonymous to writing:

command1 > logfile.txt
command2 >> logfile.txt
command3 >> logfile.txt

Redirecting Standard Error

• You must refer to the file descriptor 2 to redirect to stderr
• There are far more than just 3 (stderr, stdout, stdin) file descriptors.
• ls -l /bin/user 2> ls-error.txt
• The same is true for standard out, but by default > is equivalent to 1>

Redirecting Standard Output / Error to one file

ls -l /bin/usr > ls-output.txt 2>&1 outputs to the file. It also redirects the standard error to wherever stdout is pointing to (which is ls-output.txt)

Think of > redirections as occurring after the “meat” that are the commands have finished. Then, whatever is left over is finally what’s redirected

Consider 2>&1 > ls-output.txt

• This will route standard error to what stdout is pointing to… which is just the terminal! So nothing changes.
• Then we also have that “nothing” is redirected into ls-output.txt so that’ll be empty.

/dev/null

• You can redirect the output to this “null” file which simply does “nothing” with whatever you dump into it. ls -l /bin/usr 2> /dev/null
• This will completely shun away any errors; it will not show in the console or anything like that.

movie.mpeg.0* > movie.mpeg

• Merges all of these “subfiles” into 1 chonky file
• todo Play around with these and other wildcards

Also running cat will simply wait for you to type something since by default stdin is pointing to the keyboard.

• waittodo is every instance where you can do < replaceable with using the keyboard?
• uniq > test.txt
  • This will actually make you get input from the keyboard.
  • If you wanted to instead use a file you would have to do uniq < test.txt > other
  • todo I NEED TO PRACTICE THIS
  • > = stdout to file
  • | stdout to another command
    • Basically it throws whatever stdout already has into the command, instead of asking the user for it.
  • echo test > file1 > file2 would output only to file2 and NOT file1; file1 would be emptied though, by nature of redirection (i.e. opening for writing)

tee

ls /usr/bin | tee ls.txt | grep zip

• This throws the output of ls /usr/bin into stdout, which is then piped into tee meaning tee is taking the list of files, and outputting it into ls.txt. Due to the nature of tee (or rather, it’s sole purpose), it also “returns” or i guess stdout is set to what tee would return. That is then piped again into grep zip which prints all files with zip in it. It does NOT print the output into ls.txt since stdout has NOT been redirected!!!
• It really is like a T pipe.

[CL] 16.3 for sshs