Linux commands ¶

Usage 🐚: Alternative to man.

Example 🔥:

$ sudo apt install info
$ info ls

Usage 🐚:

• cat is commonly used display a file.
• cat is intended to be used to merge multiple files in one, and display the output.

👉 tac is printing the file from the end to the start. And concatenating files after reversing their lines.

Example 🔥:

$ cat file
the content here

# read for stdin
$ cat -

➡️ Concatenating files

$ cat f0 f1
f0
f1 Line 1
f1 Line 2

$ tac f0 f1
f0
f1 Line 2
f1 Line 1

Create a file f2 with input from f0, f1, and stdin.

$ cat f0 f1 - > f2

Usage 🐚: both are used to browse files/content from the CLI.

👉 more is rarely used as there is cat for short files, and less is more appropriate for large files.

Example 🔥:

$ more file
$ less file

There are QUITE a lot of options. Simply press h to see all of them.

• arrow up: to move up,
• arrow down: to move down
• q: to quit
• /something: search "something"
  • Enter or n: go to the next match
  • N: go back to the previous match

Usage 🐚: tee can be used to redirect one input to multiple sources. For instance, you can redirect the output to a file and stdout.

Example 🔥:

Output to both files and the standard output.

$ tee output1 output2 < file

Use -a/--append to append content.

Usage 🐚: clear the terminal

Example 🔥:

$ clear

👉 You may use the shortcut CTRL+L.

Usage 🐚: change the terminal behavior

Example 🔥:

Enable an option (ex: nullglob)

$ shopt -ps nullglob

Disable an option (ex: nullglob)

$ shopt -pu nullglob

The nullglob option is replacing patterns with the null string if there are no matches.

Usage 🐚: create a command that is an alias of another commands with usually some arguments.

Example 🔥:

Create a command la which is ls -la

$ alias la='ls -la'
$ la

See every defined alias

$ alias

⚠️ Aliases are temporary. You must load them every time you open a terminal, such as by adding them in a .bashrc in Bash.

Usage 🐚: username of the current user

Example 🔥:

$ whoami
username

Usage 🐚: info about the machine

Example 🔥:

$ uname # show OS name
$ uname -s # same
$ uname -rv # kernel
$ uname -m # architecture (x64, x86...)
$ uname -a # all

Usage 🐚: info about the machine

Example 🔥:

$ date
Mon 12 Oct 2020 08:32:11 PM EDT

Usage 🐚: stop or reboot an operating system.

Example 🔥:

To shutdown an operating system, use:

$ shutdown -h now

To reboot an operating system, use:

$ shutdown -r now
$ reboot

Usage 🐚: list all files in a directory. Both are the same, but dir is rarely used. tree display files in a tree-like format.

Example for ls/dir 🔥:

$ ls # current directory
$ ls . # same as "ls"
$ ls folder
$ ls *.txt # pattern

-l: print more information about each file

$ ls -l file # info on file

-a: print hidden files

$ ls -a folder
$ ls -la folder # info + hidden

Other options

• -R: recursive
• -p: add a trailing "/" to repositories
• -A: hide ".", and ".."
• -s: show the size
• --format="format": use a custom format
• --hide="pattern": hide files matching "pattern"
• -S: sort by size
• -t: sort by last modified date
• -u: sort by last access date

Example for tree 🔥:

List recursively given a folder.

$ tree .

Usage 🐚: move to another directory

Example 🔥:

$ cd folder

Move to the HOME folder

$ cd
$ cd ~

Move to the previous folder

$ cd -

The pushd and popd commands are quite useful when we are often moving between repositories:

$ cd /path/to/folderA
$ pushd .    # save
$ cd /path/to/folderB
$ pushd .    # save
$ cd /path/to/folderC
$ popd       # go back to folderB 
$ popd       # go back to folderA 

Usage 🐚: create one or more folders

Example 🔥:

$ mkdir folder

Create every non-existing folder in a path

$ mkdir -p folder0/folder1/folder2

Usage 🐚: there is no specific command to create a file, but there are multiple ways to achieve it: touch, cp, truncate, >, :>...

Example 🔥:

👉 touch: ensure a file exists, but it may not be empty 🔥. The command create a file or update the last modified date if it exists.

$ touch afile

👉 A trick is to copy /dev/null with cp

$ cp /dev/null file

👉 A popular way is to use a redirection of the output. Note that all of these syntax may not be supported everywhere.

$ echo -n "" > file
$ echo -n > file
$ > file
$ :> file

👉 You can use truncate to create an empty file or truncate an existing file.

$ truncate -s 0 file

Usage 🐚: create a symbolic/hard link

Example 🔥:

Create a symbolic link at output linking to target:

$ ln -s target output

Usage 🐚: takes a list of folders/files to move, and a destination. Move all files/folders to the destination.

Example 🔥:

$ mv toto ./all_toto/
$ mv toto -t ./all_toto/ # same
$ mv toto1 toto_2 ./all_toto/

Rename

$ mv toto toto0

Usage 🐚: takes a list of folders/files to copy, and a destination. Copy all files/folders to the destination.

Example 🔥:

$ cp toto ./all_toto/
$ cp toto1 toto_2 ./all_toto/

To create a copy

$ cp toto toto0 # ⚠️ toto still exists

-r ⚠️: required to copy a folder and its content

$ cp -r all_toto/ all_toto_copy

Usage 🐚: use rm to remove files/folders. rmdir is usually not used.

Example 🔥:

$ rm file
$ rm file0 file1

👉 rm by default is asking for confirmation.

$ rm -f file # do not ask (f=force)
$ rm -i file # ask

👉 Delete a folder

$ rm -d folder # if the folder is empty
$ rm -r folder # -R is doing the same

👉 Delete a folder and to not ask for confirmation

$ rm -rf folder # usual f + r

Usage 🐚: you can examine the difference between two files using diff. You can use patch to fix some differences.

Example 🔥:

Diff only.

$ diff f0 f1 -q
Files f0 and f1 differ
$ diff f0 f1
# one below the other diff
$ diff f0 f1 -y
# side-by-side diff

Generate a patch

$ diff f0 f1 -u
# ...
$ diff f0 f1 -u > f0.patch 

Then, you can use patch to apply the differences with f1 to f0.

$ patch < f0.patch
$ patch -d path/to/xxx [...]  # folder to apply patch on
$ patch -R [...]              # reverse patch
$ patch -p0 [...]             # n slashes skipped

Usage 🐚: get the filename from a path.

Example 🔥:

$ basename /etc/passwd
passwd

Usage 🐚: get the parent folder of a file.

Example 🔥:

$ dirname /etc/passwd
/etc
$ dirname /etc
/

Usage 🐚: get the absolute path to the given argument. Patterns are evaluated and symoblic links are resolved.

Example 🔥:

$ realpath ~
/home/folder

Usage 🐚: zip/unzip a .zip.

Example 🔥:

$ zip toto.zip f0 f1 f2
$ unzip toto.zip
$ gunzip -S .zip toto.zip

View the contents of a zip

$ zipinfo toto.zip
$ unzip -l toto.zip

Usage 🐚: add users to a group... Don't forget to log back in for the changes to be applied.

Example 🔥:

Add user yyy to group xxx

$ sudo usermod -a -G xxx yyy

Add "username" to sudoers

$ sudo usermod -a -G sudo username # Debian...
$ sudo usermod -a -G wheel username # Fedora...

Change the login username

$ sudo usermod -l newname oldname

Change user UID. You can't change the current user UID.

$ sudo usermod -u newUID oldUID

Lock Account

$ sudo usermod --lock [...]

Usage 🐚: delete a user

Example 🔥:

$ deluser username

Usage 🐚: change a user's password

Example 🔥:

Change the current user's password:

$ passwd

Delete the current user's password:

$ passwd -d # automatically logged/no password prompt

Change another user's password:

$ passwd username

Usage 🐚: create a group

Example 🔥:

$ groupadd group_name

Usage 🐚: edit a group information.

Example 🔥:

Change group GID.

$ sudo groupmod -g newGID oldGID

Change the name of the group.

$ sudo groupmod -n newname oldname

Usage 🐚: delete a group

Example 🔥:

$ delgroup group_name

Usage 🐚: manage a file/folder permissions

Example 🔥:

Use + to grant perms..

$ chmod u+x target
$ chmod g+rw target
$ chmod ug+r target
$ chmod g+x,o+rx target

Grant to a (all, alias of ugo)

$ chmod +x target
$ chmod a+x target # same
$ chmod ugo+x target # same

Use - instead of + to revoke permissions.

$ chmod -x target
$ chmod u-x target
$ chmod ug-rw target

Grant "perms" using the shortcut number.

# u=rwx, g=rx, o=x
$ chmod 751 target
# u=rwx, g=, o=
$ chmod 700 target

Usage 🐚: find a file/folder. There are a ton of options.

Example 🔥:

The search is done recursively starting from a folder (ex: /). We can indicate a pattern (ex: toto*).

$ find /
$ find / -name "toto*" # toto toto0...
$ find / -name "toto*" -type f # files
$ find / -name "toto*" -type d # folders

You can apply an operation on every file found. For instance, you can copy them somewhere.

# {} is the path to the match
$ find [...] -exec cp '{}' /tmp/found \;

When using find, we may get many errors. We usually redirect them to the "trash".

$ find [...] 2> /dev/null

Other options

• -empty: empty folders/files only
• -execdir: works the same as -exec, but in the directory of the match
• -quit: exit when one result is found
• -printf: change the output format
• -mindepth value: minimum depth (default=0)
• -maxdepth value: maximum depth

These are rarely used:

• -mtime n/-atime n: changed/accessed in the last n days
• -cmin n/-amin n: same, but in the last n minutes
• -newer file: modified after file
• -size n: a specific size such as 4M
• -perms 0744/-perms a=x/-writable/.../-perm -o w/-222: files matching the given perms
• -user xxx: files belonging to the user xxx
• -group xxx: files belonging to users in xxx

👉 You can add a + or - such as -size +4M.

👉 You can use !/-not to negate an option. For example, you can use -not -name.

👉 You can group conditions too: \( -name xxx -o -name yyy \). Use -o for OR and -a for AND.

Usage 🐚: search a file in a local database of files. The database must be updated manually sudo updatedb.

Example 🔥:

$ locate find

Usage 🐚: search files based on their content.

Example 🔥:

List files having 'toto' in them

$ grep "toto" *
$ grep --color "toto" * # highlight match

Best options

• -r recursive (-R to follow links too)
• -i Ignore case
• -v Inverse pattern
• -c: number of matches per file
• -n: add line number before each match
• -o show only the matched part
• -H show the file before every match
• -w words-only ("XxtotoxX" won't match "toto")

regex mode 🧪

By default, if you are using ?, (, or any character used by regexes, it won't be interpreted (eg. ? won't be considered as "optional").

$ grep "opt?" * # ❌ match "opt?"
$ grep "opt\?" * # ✅ "t" is optional
$ grep -E "opt?" * # ✅ same as grep -E
$ egrep "opt?" * # ✅ same as grep -E

Less frequently used options

• -L: stop when match found, show files without matches
• -l: stop when match found, show files with matches
• -q: no output, use the exit code to indicate if a match was found (0), or not (1).

Usage 🐚: sed has many uses. You can apply simple modification on the file, or search and replace something.

sed does not modify the file, but output on stdout the modified content, which you can redirect.

Example - syntax 1 🔥:
sed '<range><letter><args>' <file>.

• i: add a line with "XXX" before (i) every line
• a: add a line with "XXX" after (a) every line
• p: duplicate every line
• d: delete every line
• c: replace every line with "XXX"
• e: execute the command before every line (you may add parenthesis to make things cleaner)

$ sed 'iXXX' file
$ sed 'cXXX' file
$ sed 'aXXX' file
$ sed 'p' file
$ sed 'e(echo hello)' file

Instead of applying a command to every line, you can pick some lines

$ sed '1iXXX' file # line 1
$ sed '1,3iXXX' file # line 1 to 3
$ sed '$iXXX' file # last line
$ sed '1p' file # duplicate first
$ sed '1d' file # delete first
# ...

Example - syntax 2 🔥:
sed 's/<pattern>/<replacement>/<flag>' <file>.

• pattern: a regex
• flag:
  • none: first match of each line
  • g: every match is replaced
  • n: replace the nth match of each line
• replacement: what to replace the section with

Replace every "e" with "E"

$ sed "s/e/E/g" file

Replace the first "e" of each line with "E"

$ sed "s/e/E/" file
$ sed "s/e/E/1" file

Comment out every line starting with "S". We use a capture group 🚩 which is referenced using \1.

$ sed "s/^\(S.*\)/# \1./" file

Usage 🐚: most of the time, there are other solutions that are more appropriate such as sed. awk is a mix of sed, cut, tr...

👉 See also: The_AWK_Programming_Language & To awk or not

awk is considering space (-F to change) as a separator for columns. The first column is $1, the nth column is $n. $0 means every column.

The syntax is awk 'target {action}' file.

• target: filter where the action is applied
• action: print...

Example 🔥:

By default, awk prints every column.

$ awk '{print}' file
$ awk -F' ' '{print $0}' file # same

Only print the first column of lines having AT LEAST 6 columns (NF)

$ awk 'NF > 6 {print $1}' file

Print the first, and the third column, if the file has at least 3 lines (NR)

$ awk 'NR > 3 {print $1 $3}' file

Usage 🐚: extract "fields" from a table For instance, given the following file:

$ cat /etc/passwd
root:x:0:0:root:/root:/bin/bash

For every line, we got values separated by ":". These are what we call "fields". For cut, the row look like this: f1:f2:f3:f4:f5:f6:f7.

Example 🔥:

$ cat /etc/passwd | cut -d: -f1
root
$ cat /etc/passwd | cut -d: -f1-3
root:x:0
$ cat /etc/passwd | cut -d: -f1,3
root:0

Usage 🐚: replace a character with another, or delete a character.

Example 🔥:

Replace whitespaces with the newline character

$ cat file | tr ' ' '\n'

Delete every character W

$ cat file | tr -d 'W'

Usage 🐚: number lines of files

Example 🔥:

Number non-empty lines

$ nl file
$ nl -bt file

Number all lines

$ nl -ba file

Usage 🐚: count words/lines/characters

Example 🔥:

Use -l for lines, -w for words, and -c for characters.

$ cat file.txt | wc -l
7560 # lines

You can read something from stdin. Use CTRL+D to indicate the end of the input.

$ wc -w
Hello
World # <CTRL+D>
2

Usage 🐚: output the sorted file

Example 🔥:

$ sort
$ sort -d

• -b: ignore leading blanks
• -r: reverse (z to a)

You can define a separator (:), and only sort a column.

$ sort s -t: -k2 # sort by second column
3:a
1:b
2:c
$ sort s -t: -k2.0 # column.nth character
$ sort s -t: -k2.0,2.1 # range

Usage 🐚: remove lines or indicate how many duplicates there are for each unique line.

Example 🔥:

Remove duplicates

$ uniq < file

Add the number of duplicates for each unique line

$ uniq -c < file
  1: XXX
  1: YYY
  5: ZZZ

Usage 🐚: do GET/POST/... requests.

Example 🔥:

GET request.

$ curl https://example.com    # GET an URL
$ curl -u username:pass [...] # HTTP Basic Auth
$ curl -i [...]               # request headers (any request)
$ curl -k [...]               # ignore SSL errors
$ curl -O [...]               # save ; use remote filename
$ curl -o xxx [...]           # save ; use custom filename
$ curl --silent [...]         # -s | only print the output

Craft an HTTP request:

$ curl -X GET [...]           # same as "--request"
$ curl -H "Name: Value" [...] # same as "--header"
$ curl -d "key=value" [...]   # same as "--data"
$ curl -d '{"key":"value"}' [...] # ex: json
$ curl -A "xxx" [...]         # set user-agent
$ curl -b "NAME=VALUE" [...]  # set cookie
$ curl -T file [...]          # upload file using PUT

POST request.

$ curl -X POST 'URL' -d 'key=value' -H 'Content-Type: application/x-www-form-urlencoded'

Usage 🐚: track the time a command takes.

Example 🔥:

$ time sleep 5
real    0m5.002s
user    0m0.001s
sys     0m0.000s

Usage 🐚: set the size, orientation, brightness of the screen...

Example 🔥:

Set the brightness to "0.5"

$ xrandr --output DP-2 --brightness 0.5