Core utilities

From ArchWiki
Revision as of 07:18, 6 June 2018 by Larivact (talk | contribs) (→‎awk: mention that gawk is part of the base group)
Jump to navigation Jump to search

This article deals with so-called core utilities on a GNU/Linux system, such as less, ls, and grep. The scope of this article includes, but is not limited to, those utilities included with the GNU coreutils package. What follows are various tips and tricks and other helpful information related to these utilities.

Basic commands

The following table lists basic shell commands every Linux user should be familiar with. See the below sections and Related articles for details.

Command Description Manual page Example
man Show manual page for a command man(7) man ed
cd Change directory (shell built-in command) cd(1p) cd /etc/pacman.d
mkdir Create a directory mkdir(1) mkdir ~/newfolder
rmdir Remove empty directory rmdir(1) rmdir ~/emptyfolder
rm Remove a file rm(1) rm ~/file.txt
rm -r Remove directory and contents rm -r ~/.cache
ls List files ls(1) ls *.mkv
ls -a List hidden files ls -a /home/archie
ls -al List hidden files and file properties
mv Move a file mv(1) mv ~/ ~/archive/
cp Copy a file cp(1) cp ~/.bashrc ~/.bashrc.bak
chmod +x Make a file executable chmod(1) chmod +x ~/.local/bin/
cat Show file contents cat(1) cat /etc/hostname
strings Show printable characters in binary files strings(1) strings /usr/bin/free
find Search for a file find(1) find ~ -name myfile
mount Mount a partition mount(8) mount /dev/sdc1 /media/usb
df -h Show remaining space on all partitions df(1)
ps -A Show all running processes ps(1)
killall Kill all running instances of a process killall(1)
ss -at Display a list of open TCP sockets ss(8)


AWK is a pattern scanning and processing language. There are multiple implementations:

  • gawk — GNU version of awk || gawk (part of base)
  • nawk — The one, true implementation of AWK || nawk
  • mawk — A very fast AWK implementation || mawkAUR
  • BusyBox also includes an AWK implementation.


cat is a standard Unix utility that concatenates files to standard output.

  • Because cat is not built into the shell, on many occasions you may find it more convenient to use a redirection, for example in scripts, or if you care a lot about performance. In fact < file does the same as cat file.
  • cat can work with multiple lines:
$ cat << EOF >> path/file
first line
last line

Alternatively, using printf:

$ printf '%s\n' 'first line' ... 'last line'
  • If you need to list file lines in reverse order, there is a coreutil called tac (cat reversed).


See File permissions and attributes#Changing permissions.


See File permissions and attributes#Changing ownership.


dd is a utility for Unix and Unix-like operating systems whose primary purpose is to convert and copy a file.

Similarly to cp, by default dd makes a bit-to-bit copy of the file, but with lower-level I/O flow control features.

Some notable applications of dd are:

  • Binary file patching: let say one wants to replace offset 0x123AB of a file with the FF C0 14 hexadecimal sequence, this can be done with the command line:
    # printf '\xff\xc0\x14' | dd seek=$((0x123AB)) conv=notrunc bs=1 of=/path/to/file

For more information see dd(1) or the full documentation.

Tip: By default, dd outputs nothing until the task has finished. To monitor the progress of the operation, add the status=progress option to the command.
Warning: One should be extremely cautious using dd, as with any command of this kind it can destroy data irreversibly.


find is part of the findutils package, which belongs to the base package group.

Tip: fd is a simple, fast and user-friendly alternative to find that provides more sensible defaults (e.g. ignores hidden files, directories and .gitignore'd files, fd PATTERN instead of find -iname '*PATTERN*'). It features colorized output (similar to ls), Unicode awareness, regular expressions and more.

One would probably expect a find command to take as argument a file name and search the filesystem for files matching that name. For a program that does exactly that see #locate below.

Instead, find takes a set of directories and matches each file under them against a set of expressions. This design allows for some very powerful "one-liners" that would not be possible using the "intuitive" design described above. See GregsWiki:UsingFind for usage details.


grep (from ed's g/re/p, global/regular expression/print) is a command line text search utility originally written for Unix. The grep command searches files or standard input for lines matching a given regular expression, and prints these lines to standard output.

  • Remember that grep handles files, so a construct like cat file | grep pattern is replaceable with grep pattern file
  • There are grep alternatives optimized for VCS source code, such as ripgrep, the_silver_searcher, and ack.
  • To include file line numbers in the output, use the -n option.
  • grep can also be used for hexadecimal search in a binary file, to look for let say the A1 F2 sequence in a file, the command line is:
    $ LANG=C grep --text --perl-regexp "\xA1\xF2" /path/to/file
Note: Some commands send their output to stderr(3), and grep has no apparent effect. In this case, redirect stderr to stdout with command 2>&1 | grep args or (for Bash 4) command |& grep args. See also I/O Redirection.

For color support, see Color output in console#grep.

See grep(1) for more details.


iconv converts the encoding of characters from one codeset to another.

The following command will convert the file foo from ISO-8859-15 to UTF-8, saving it to foo.utf:

$ iconv -f ISO-8859-15 -t UTF-8 foo > foo.utf

See iconv(1) for more details.

Convert a file in place

Tip: You can use recode instead of iconv if you do not want to touch the mtime.

Unlike sed, iconv does not provide an option to convert a file in place. However, sponge from the moreutils package can help:

$ iconv -f WINDOWS-1251 -t UTF-8 foobar.txt | sponge foobar.txt

See sponge(1) for details.


ip allows you to show information about network devices, IP addresses, routing tables, and other objects in the Linux IP software stack. By appending various commands, you can also manipulate or configure most of these objects.

Note: The ip utility is provided by the iproute2 package, which is included in the base group.
Object Purpose Manual page
ip addr protocol address management ip-address(8)
ip addrlabel protocol address label management ip-addrlabel(8)
ip l2tp tunnel Ethernet over IP (L2TPv3) ip-l2tp(8)
ip link network device configuration ip-link(8)
ip maddr multicast addresses management ip-maddress(8)
ip monitor watch for netlink messages ip-monitor(8)
ip mroute multicast routing cache management ip-mroute(8)
ip mrule rule in multicast routing policy db
ip neigh neighbour/ARP tables management ip-neighbour(8)
ip netns process network namespace management ip-netns(8)
ip ntable neighbour table configuration ip-ntable(8)
ip route routing table management ip-route(8)
ip rule routing policy database management ip-rule(8)
ip tcp_metrics management for TCP Metrics ip-tcp_metrics(8)
ip tunnel tunnel configuration ip-tunnel(8)
ip tuntap manage TUN/TAP devices
ip xfrm manage IPsec policies ip-xfrm(8)

The help command is available for all objects. For example, typing ip addr help will show you the command syntax available for the address object. For advanced usage see the iproute2 documentation.

The Network configuration article shows how the ip command is used in practice for various common tasks.

Note: You might be familiar with the ifconfig command, which was used in older versions of Linux for interface configuration. It is now deprecated in Arch Linux; you should use ip instead.


Tango-view-fullscreen.pngThis article or section needs expansion.Tango-view-fullscreen.png

Reason: less is a complex beast, and this section should explain some of the basic less commands (Discuss in Talk:Core utilities#)

less is a terminal pager program used to view the contents of a text file one screen at a time. Whilst similar to other pagers such as more and pg, less offers a more advanced interface and complete feature-set.

See List of applications#Terminal pagers for alternatives.


Install the mlocate package. The package contains an updatedb.timer unit, which invokes a database update each day. The timer is enabled right after installation, start it manually if you want to use it before reboot. You can also manually run updatedb as root at any time. By default, paths such as /media and /mnt are ignored, so locate may not discover files on external devices. See updatedb(8) for details.

The locate command is a common Unix tool for quickly finding files by name. It offers speed improvements over the find tool by searching a pre-constructed database file, rather than the filesystem directly. The downside of this approach is that changes made since the construction of the database file cannot be detected by locate.

Before locate can be used, the database will need to be created. To do this, execute updatedb as root.

See also How locate works and rewrite it in one minute.

Vim as alternative pager

Vim includes a script to view the content of text files, compressed files, binaries and directories. Add the following line to your shell configuration file to use it as a pager:

alias less='/usr/share/vim/vim80/macros/'

There is also an alternative to the macro, which may work as the PAGER environment variable. Install vimpager and add the following to your shell configuration file:

export PAGER='vimpager'
alias less=$PAGER

Now programs that use the PAGER environment variable, like git, will use vim as pager.


ls lists directory contents.

See info ls or the online manual for more information.

exa is a modern, and more user friendly alternative to ls and tree, that has more features, such as displaying Git modifications along with filenames, colouring differently each columnn in --long mode, or displaying --long mode metadata along with a tree view. exa

Long format

The -l option displays some metadata, for example:

$ ls -l /path/to/directory
total 128
drwxr-xr-x 2 archie users  4096 Jul  5 21:03 Desktop
drwxr-xr-x 6 archie users  4096 Jul  5 17:37 Documents
drwxr-xr-x 2 archie users  4096 Jul  5 13:45 Downloads
-rw-rw-r-- 1 archie users  5120 Jun 27 08:28 customers.ods
-rw-r--r-- 1 archie users  3339 Jun 27 08:28 todo
-rwxr-xr-x 1 archie users  2048 Jul  6 12:56

The total value represents the total disk allocation for the files in the directory, by default in number of blocks.

Below, each file and subdirectory is represented by a line divided into 7 metadata fields, in the following order:

  • type and permissions:
    • the first character is the entry type, see info ls -n "What information is listed" for an explanation of all the possible types; for example:
      • - denotes a normal file;
      • d denotes a directory, i.e. a folder containing other files or folders;
      • p denotes a named pipe (aka FIFO);
      • l denotes a symbolic link;
    • the remaining characters are the entry's permissions;
  • number of hard links for the entity; files will have at least 1, i.e. the showed reference itself; folders will have at least 2: the showed reference, the self-referencing . entry, and then a .. entry in each of its subfolders;
  • owner user name;
  • group name;
  • size;
  • last modification timestamp;
  • entity name.

File names containing spaces enclosed in quotes

By default, file and directory names that contain spaces are displayed surrounded by single quotes. To change this behavior use the -N or --quoting-style=literal options. Alternatively, set the QUOTING_STYLE environment variable to literal. [1]


lsblk(8) will show all available block devices along with their partitioning schemes, for example:

$ lsblk -f
NAME   FSTYPE   LABEL       UUID                                 MOUNTPOINT
├─sda1 vfat                 C4DA-2C4D                            /boot
├─sda2 swap                 5b1564b2-2e2c-452c-bcfa-d1f572ae99f2 [SWAP]
└─sda3 ext4                 56adc99b-a61e-46af-aab7-a6d07e504652 /

The beginning of the device name specifies the type of block device. Most modern storage devices (e.g. hard disks, SSDs and USB flash drives) are recognised as SCSI disks (sd). The type is followed by a lower-case letter starting from a for the first device (sda), b for the second device (sdb), and so on. Existing partitions on each device will be listed with a number starting from 1 for the first partition (sda1), 2 for the second (sda2), and so on. In the example above, only one device is available (sda), and that device has three partitions (sda1 to sda3), each with a different file system.

Other common block device types include for example mmcblk for memory cards and nvme for NVMe devices. Unknown types can be searched in the kernel documentation[dead link 2017-11-11].


mkdir makes directories.

To create a directory and its whole hierarchy, the -p switch is used, otherwise an error is printed. As users are supposed to know what they want, -p switch may be used as a default:

alias mkdir='mkdir -p -v'

The -v switch make it verbose.

Changing mode of a just created directory using chmod is not necessary as the -m option lets you define the access permissions.

Tip: If you just want a temporary directory, a better alternative may be mktemp: mktemp -d


mv moves and renames files and directories.

To limit potential damage caused by the command, use an alias:

alias mv='timeout 8 mv -iv'

This alias suspends mv after eight seconds, asks for confirmation before overwriting any existing files, lists the operations in progress and does not store itself in the shell history file if the shell is configured to ignore space starting commands.


The od (octal dump) command is useful for visualizing data that is not in a human-readable format, like the executable code of a program, or the contents of an unformatted device. See the manual for more information.


You can use pv (pipe viewer) to monitor the progress of data through a pipeline, for example:

# dd if=/source/filestream | pv -monitor_options -s size_of_file | dd of=/destination/filestream

In most cases pv functions as a drop-in replacement for cat.


rm removes files or directories.

To limit potential damage caused by the command, use an alias:

alias rm='timeout 3 rm -Iv --one-file-system'

This alias suspends rm after three seconds, asks confirmation to delete three or more files, lists the operations in progress, does not involve more than one file systems and does not store itself in the shell history file if the shell is configured to ignore space starting commands. Substitute -I with -i if you prefer to confirm even for one file.

Zsh users may want to put noglob before timeout to avoid implicit expansions.

To remove directories believed to be empty, use rmdir as it fails if there are files inside the target.


sed is stream editor for filtering and transforming text.

Here is a handy list of sed one-liners examples.

Tip: More powerful alternatives are #awk and the Perl language.


seq prints a sequence of numbers. Shell built-in alternatives are available, so it is good practice to use them as explained on Wikipedia.


ss is a utility to investigate network ports and is part of the iproute2 package in the base group. It has a similar functionality to the deprecated netstat utility.

Common usage includes:

Display all TCP Sockets with service names:

$ ss -at

Display all TCP Sockets with port numbers:

$ ss -atn

Display all UDP Sockets:

$ ss -au

For more information see ss(8) or ss.html from the iproute2 package.


As an early Unix archiving format, .tar files—known as "tarballs"—are widely used for packaging in Unix-like operating systems. Both pacman and AUR packages are compressed tarballs, and Arch uses GNU's tar program by default.

For .tar archives, tar by default will extract the file according to its extension:

$ tar xvf file.EXTENSION

Forcing a given format:

File Type Extraction Command
file.tar tar xvf file.tar
file.tgz tar xvzf file.tgz
file.tar.gz tar xvzf file.tar.gz bzip -cd | tar xvf -
file.tar.bz2 tar xvjf file.tar.bz2
bzip2 -cd file.bz2 | tar xvf -
file.tar.xz tar xvJf file.tar.xz
xz -cd file.xz | tar xvf -
file.tar.zst tar -I zstd xvf file.tar.zst

The construction of some of these tar arguments may be considered legacy, but they are still useful when performing specific operations. See tar(1) for details.


which shows the full path of shell commands. In the following example the full path of ssh is used as an argument for journalctl:

# journalctl $(which sshd)


wipefs can list or erase file system, RAID or partition-table signatures (magic strings) from the specified device. It does not erase the file systems themselves nor any other data from the device.

See wipefs(8) for more information.

For example, to erase all signatures from the device /dev/sdb and create a signature backup ~/wipefs-sdb-offset.bak file for each signature:

# wipefs --all --backup /dev/sdb

See also