What's this?

commandlinefu.com is the place to record those command-line gems that you return to again and again.

Delete that bloated snippets file you've been using and share your personal repository with the world. That way others can gain from your CLI wisdom and you from theirs too. All commands can be commented on, discussed and voted up or down.

Get involved!

You can sign-in using OpenID credentials, or register a traditional username and password.

First-time OpenID users will be automatically assigned a username which can be changed after signing in.

Universal configuration monitoring and system of record for IT.

Stay in the loop…

Follow the Tweets.

Every new command is wrapped in a tweet and posted to Twitter. Following the stream is a great way of staying abreast of the latest commands. For the more discerning, there are Twitter accounts for commands that get a minimum of 3 and 10 votes - that way only the great commands get tweeted.

» http://twitter.com/commandlinefu
» http://twitter.com/commandlinefu3
» http://twitter.com/commandlinefu10

Subscribe to the feeds.

Use your favourite RSS aggregator to stay in touch with the latest commands. There are feeds mirroring the 3 Twitter streams as well as for virtually every other subset (users, tags, functions,…):

Subscribe to the feed for:



May 19, 2015 - A Look At The New Commandlinefu
I've put together a short writeup on what kind of newness you can expect from the next iteration of clfu. Check it out here.
March 2, 2015 - New Management
I'm Jon, I'll be maintaining and improving clfu. Thanks to David for building such a great resource!

Top Tags



Psst. Open beta.

Wow, didn't really expect you to read this far down. The latest iteration of the site is in open beta. It's a gentle open beta-- not in prime-time just yet. It's being hosted over at UpGuard (link) and you are more than welcome to give it a shot. Couple things:

  • » The open beta is running a copy of the database that will not carry over to the final version. Don't post anything you don't mind losing.
  • » If you wish to use your user account, you will probably need to reset your password.
Your feedback is appreciated via the form on the beta page. Thanks! -Jon & CLFU Team

All commands from sorted by
Terminal - All commands - 12,393 results
ip link
tr -dc 'a-zA-Z0-9' < /dev/urandom | fold -w 10 | sed 1q
kill `ps aux | grep xmms | grep -v grep | awk '{ print $2 }'`
2010-11-11 17:26:26
User: corp769
Functions: awk grep kill

I use this command, within a cron job, to kill XMMS after a certain amount of time. This command returns the PID used by XMMS, and gets passed to the kill command. Another alternative would be ps aux | grep xmms | grep -v grep | awk '{ print $2 }' | xargs kill

2009-03-24 16:19:12
User: haivu
Functions: bc

bc is a wonderful calculator. Just type bc at the command line and have at it. Ctrl+D (or type quit) will get you out. This usage is just scratching the surface: bc can handle a mini scripting language, complete with variable, statements, loop, conditional statements and more. Do a man page on it to find out.

svn status |grep '\?' |awk '{print $2}'| parallel -Xj1 svn add
2010-01-28 08:47:54
Functions: awk grep
Tags: xargs parallel

xargs deals badly with special characters (such as space, ' and "). To see the problem try this:

touch important_file

touch 'not important_file'

ls not* | xargs rm

Parallel https://savannah.nongnu.org/projects/parallel/ does not have this problem.

leapyear() { if [ $[$1 % 4] -eq 0 ] && [ $[$1 % 100] -ne 0 ] || [ $[$1 % 400] -eq 0 ]; then echo $1' is a leap year!'; else echo $1' is not a leap year.'; fi; }
2010-03-30 17:19:20
User: kaedenn
Functions: echo
Tags: echo test

Tested on bash, and follows all the rules about leap years.

ls -t1 | head -n1
head -1000 < lines.txt | tail -1
killall xmms
/opt/psa/bin/pleskbackup server -v --output-file=plesk_server.bak
wget -O - -q http://checkip.dyndns.org/ | cut -d':' -f2 | cut -d'<' -f1| cut -c2-
2011-09-17 13:42:01
User: ztank1013
Functions: cut wget

This is just a "cut" addicted variant of the previous unixmonkey24730 command...

svn st | grep /main/java | awk '{print $2}' | xargs echo | xargs svn ci -m "my comment here"
BACKUP_FILE_SIZE=`eval ls -l ${BACKUP_FILE} | awk {'print $5'}`; if [ $BACKUP_FILE_SIZE -le 20 ]; then echo "its empty"; else echo "its not empty"; fi
2009-12-29 08:34:37
User: Redrocket
Functions: awk echo ls

If you gzip an empty file it becomes 20 bytes. Some backup checks i do check to see if the file is greater than zero size (-s flag) but this is no good here. Im sure someone has a better check than me for this? No check to see if file exists before checking it's size.

java -jar compiler.jar --js file.js
svn st | cut -c 9- | parallel -X tar -czvf ../backup.tgz
2010-01-28 11:43:16
Functions: cut tar

xargs deals badly with special characters (such as space, ' and "). In this case if you have a file called '12" record'.

Parallel https://savannah.nongnu.org/projects/parallel/ does not have this problem.

Both solutions work bad if the number of files is more than the allowed line length of the shell.

2011-04-14 06:59:06
User: atgf0127
Functions: top

usage: top -hv | -bcisSHM -d delay -n iterations [-u user | -U user] -p pid [,pid ...]

ldapsearch -H ldap://localhost:389 -D cn=username,ou=users,dc=domain -x -W -b ou=groups,dc=domain '(member=cn=username,ou=users,dc=domain)' | grep ^dn | sed "s/dn\: cn=\([^,]*\),ou=\([^,]*\),.*/\2 \1/"
2009-06-11 14:50:11
User: nitehawk
Functions: grep sed

This expression looks for groups inside of a GroupOfNames class element, that is itself inside one (or many) Organizational Unit (ou) nodes in the ldap tree. Give you a quick dump of all the groups the user belongs to. Handy for displaying on a webpage.

watch -n 1 "awk 'NR==3 {print \"Signal strength = \" \$3 \"00 %\"}''' /proc/net/wireless"
nc -kl 5432 -c 'echo -e "HTTP/1.1 200 OK\r\n$(date)\r\n\r\n";echo "<p>How are you today?</p>"'
2013-11-12 14:00:11
User: gvitalie
Functions: echo

-k, --keep-open will keep connection alive, and we could exclude using 'while true'

nc is such a powerful command, it could be used instead of any OS! :p

git rev-list --reverse --topo-order master... | while read rev; do git checkout preview; git cherry-pick $rev || break; done
2015-04-23 14:28:06
User: shadyvb
Functions: read
Tags: git preview

Creating feature-branches off master, and trying to merge them in an integration branch (preview), sometimes causes conflicts because the feature-branch might hold changes from 'master' that aren't on preview yet. So this ensures only the commits added to the feature-branch are moved to integration (preview).

Note: This assumes you're currently on the feature-branch. Adjust 'master/preview' branch names to suit your environment.

curl -s http://whatthecommit.com | sed -n '/<p>/,/<\/p>/p' | sed '$d' | sed 's/<p>//'
wget -O - -q http://www.chisono.it/ip.asp && echo
2011-09-18 15:38:02
User: scanepa
Functions: wget

The echo at the end is for pretty printing as the output is just the IP address without any html

du -s * | sort -nr | head | cut -f2 | parallel -k du -sh
2010-01-28 12:59:14
Functions: cut du head sort
Tags: du xargs parallel

If a directory name contains space xargs will do the wrong thing. Parallel https://savannah.nongnu.org/projects/parallel/ deals better with that.

2011-08-19 14:50:28
User: anhstar

Basic search and replaceEdit

The :substitute command searches for a text pattern, and replaces it with a text string. There are many options, but these are what you probably want:


Find each occurrence of 'foo', and replace it with 'bar'.


Change each 'foo' to 'bar', but ask for confirmation first.


Change only whole words exactly matching 'foo' to 'bar'; ask for confirmation.


Change each 'foo' (case insensitive) to 'bar'; ask for confirmation.


Change each 'foo' (case sensitive) to 'bar'; ask for confirmation.

The g flag means global ? each occurrence in the line is changed, rather than just the first.


Search range:

:s/foo/bar/g Change each 'foo' to 'bar' in the current line.

:%s/foo/bar/g Change each 'foo' to 'bar' in all lines.

:5,12s/foo/bar/g Change each 'foo' to 'bar' for all lines from line 5 to line 12 inclusive.

:'a,'bs/foo/bar/g Change each 'foo' to 'bar' for all lines from mark a to mark b inclusive.

:.,$s/foo/bar/g Change each 'foo' to 'bar' for all lines from the current line (.) to the last line ($) inclusive.

:.,+2s/foo/bar/g Change each 'foo' to 'bar' for the current line (.) and the two next lines (+2).

:%s/foo/bar/g Equivalent to :1,$s/foo/bar/g (change all lines).

:g/^baz/s/foo/bar/g Change each 'foo' to 'bar' in each line starting with 'baz'.

When searching:

., *, \, [, ], ^, and $ are metacharacters.

+, ?, |, {, }, (, and ) must be escaped to use their special function.

\/ is / (use backslash + forward slash to search for forward slash)

\t is tab, \s is whitespace

\n is newline, \r is CR (carriage return = Ctrl-M = ^M)

\{#\} is used for repetition. /foo.\{2\} will match foo and the two following characters. The \ is not required on the closing } so /foo.\{2} will do the same thing.

\(foo\) makes a backreference to foo. Parenthesis without escapes are literally matched. Here the \ is required for the closing \).

When replacing:

\r is newline, \n is a null byte (0x00).

\& is ampersand (& is the text that matches the search pattern).

\1 inserts the text of the first backreference. \2 inserts the second backreference, and so on.

You can use other delimiters with substitute:


Save typing by using \zs and \ze to set the start and end of a pattern. For example, instead of:

:s/Copyright 2007 All Rights Reserved/Copyright 2008 All Rights Reserved/


:s/Copyright \zs2007\ze All Rights Reserved/2008/

Using the current word or registersEdit


Replace each match of the last search pattern with 'bar'.

For example, you might first place the cursor on the word foo then press * to search for that word.

The above substitute would then change all words exactly matching 'foo' to 'bar'.


Replace each occurrence of 'foo' with the word under the cursor.

means that you press Ctrl-R then Ctrl-W.

The word under the cursor will be inserted as though you typed it.


Replace each occurrence of 'foo' with the WORD under the cursor (delimited by whitespace).

means that you press Ctrl-R then Ctrl-A.

The WORD under the cursor will be inserted as though you typed it.


Replace each occurrence of 'foo' with the contents of register 'a'.

a means that you press Ctrl-R then a.

The contents of register 'a' will be inserted as though you typed it.

:%s/foo/\[email protected]/g

Replace each occurrence of 'foo' with the contents of register 'a'.

\[email protected] is a reference to register 'a'.

The contents of register 'a' is not shown in the command. This is useful if the register contains many lines of text.


Replace each match of the last search pattern with the / register (the last search pattern).

After pressing Ctrl-R then / to insert the last search pattern (and before pressing Enter to perform the command), you could edit the text to make any required change.


Replace all occurrences of the text in the system clipboard (in the * register) with 'bar' (see next example if multiline).

On some systems, selecting text (in Vim or another application) is all that is required to place that text in the * register.


Replace all occurrences of the text in register 'a' with 'bar'.

a means that you press Ctrl-R then a. The contents of register 'a' will be inserted as though you typed it.

Any newlines in register 'a' are inserted as ^M and are not found.

The search works if each ^M is manually replaced with '\n' (two characters: backslash, 'n').

This replacement can be performed while you type the command:


The "\n" (double quotes) represents the single character newline; the '\\n' (single quotes) represents two backslashes followed by 'n'.

The substitute() function is evaluated by the = (Ctrl-R =) expression register; it replaces each newline with a single backslash followed by 'n'.

The indicates that you press Enter to finish the = expression.

See Paste registers in search or colon commands instead of using the clipboard.

Additional examplesEdit


On each line, replace the first occurrence of "foo" with "bar".


On each line, replace the last occurrence of "foo" with "bar".


On each line, delete the whole word "foo" and all following text (to end of line).


On each line, delete the whole word "foo" and the following five characters.


On each line, delete all text following the whole word "foo" (to end of line).


On each line, delete the whole word "foo" and all preceding text (from beginning of line).


On each line, delete all the text preceding the whole word "foo" (from beginning of line).


On each line, delete all the text preceding and following the whole word "foo".

Special casesEdit

For substituting patterns with a corresponding case-sensitive text, Michael Geddes's keepcase plugin can be used, e.g.:


Substitute 'Hello hello helLo HELLO' by 'Goodbye goodbye goodBye GOODBYE'

For changing the offsets in a patch file (line number of a block), this little snippet can be used:

s/^@@ -\(\d\+\),\(\d\+\) +\(\d\+\),\(\d\+\) @@$/\="@@ -".eval(submatch(1)+offsetdiff).",".submatch(2)." +".eval(submatch(3)+offsetdiff).",".submatch(4)." @@"/g

Useful when we want to strip some blocks from a patch, without patch having to complain about offset differences.

Note Should try to make the expression more compact, but don't know how without having the possibility of modifying unwanted lines.

yes|for x in one two three; do echo result - $x; done
2010-06-01 14:49:29
User: Panikos
Functions: echo yes

simple bash one liner to pass multiple arguments to command one by one. optional yes/no pipe at beginning of command