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.

If you have a new feature suggestion or find a bug, please get in touch via http://commandlinefu.uservoice.com/

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.


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:



2011-03-12 - Confoo 2011 presentation
Slides are available from the commandlinefu presentation at Confoo 2011: http://presentations.codeinthehole.com/confoo2011/
2011-01-04 - Moderation now required for new commands
To try and put and end to the spamming, new commands require moderation before they will appear on the site.
2010-12-27 - Apologies for not banning the trolls sooner
Have been away from the interwebs over Christmas. Will be more vigilant henceforth.
2010-09-24 - OAuth and pagination problems fixed
Apologies for the delay in getting Twitter's OAuth supported. Annoying pagination gremlin also fixed.




Commands using read from sorted by
Terminal - Commands using read - 288 results
git ls-files | while read i; do git blame $i | sed -e 's/^[^(]*(//' -e 's/^\([^[:digit:]]*\)[[:space:]]\+[[:digit:]].*/\1/'; done | sort | uniq -ic | sort -nr
2009-10-25 09:40:01
User: pipping
Functions: read sed sort uniq
Tags: statistics git

You'll run into trouble if you have files w/ missing newlines at the end. I tried to use

PAGER='sed \$q' git blame

and even

PAGER='sed \$q' git -p blame

to force a newline at the end, but as soon as the output is redirected, git seems to ignore the pager.

seg() { echo -e "$1" | while read LINE; do for b in $(seq 10); do echo $LINE.$b; done; done; }
ord () { seq 1 127 | while read i; do echo `chr $i` $i; done | grep "^$1 " | cut -c '3-' }
2009-10-16 21:54:01
User: infinull
Functions: cut echo grep read seq

uses the previous "chr" function and uses it to create the inverse function "ord" by brute force.

It's slow, It's inelegant, but it works.

I thought I needed ord/chr to do a cartesian cipher in shell script a whie ago, but eventualy I realized I could get fancy with tr and do the same thing...

dpkg --get-selections | cut -f1 | while read pkg; do dpkg -L $pkg | xargs -I'{}' bash -c 'if [ ! -d "{}" ]; then echo "{}"; fi' | tr '\n' '\000' | du -c --files0-from - | tail -1 | sed "s/total/$pkg/"; done
2009-10-12 14:57:54
User: pykler
Functions: bash cut du echo read sed tail tr xargs
Tags: Debian wajig

Calculates the size on disk for each package installed on the filesystem (or removed but not purged). This is missing the

| sort -rn

which would put the biggest packges on top. That was purposely left out as the command is slightly on the slow side

Also you may need to run this as root as some files can only be checked by du if you can read them ;)

cat file.txt | while read line; do printf "%7.2f -> %7.2f\n" $line; done
beepwhenup () { echo 'Enter host you want to ping:'; read PHOST; if [[ "$PHOST" == "" ]]; then exit; fi; while true; do ping -c1 -W2 $PHOST 2>&1 >/dev/null; if [[ "$?" == "0" ]]; then for j in $(seq 1 4); do beep; done; ping -c1 $PHOST; break; fi; done; }
2009-09-24 18:11:10
Functions: echo host ping read seq
Tags: ping beep

After this, just type:


You need to install "beep" before this would make the beep sound.

Save it in your .profile if you want to use it later

WARNING: this command won't exit until it is successful. You won't be able to CONTROL+C out of it.

tar -cvf /dev/null . | while read i; do ls -l $i; done
2009-09-16 16:59:15
User: lbonanomi
Functions: ls read tar

I find the ouput of ls -lR to be un-satisfying (why is the path data up there?) and find syntax to be awkward. Running 'du -a' means you will have likely to trim-off filesize data before feeding filenames to the next step in the pipe.

find repMainPath -maxdepth 1 -mindepth 1 -type d | while read dir; do echo processing $dir; sudo svnadmin dump --deltas $dir >dumpPath/`basename $dir`; done
2009-09-15 20:14:51
User: Marco
Functions: dump echo find read sudo
Tags: bash svn

This command dumps all SVN repositories inside of folder "repMainPath" (not recursively) to the folder "dumpPath", where one dump file will be created for each SVN repository.

find -type f |while read a;do [ "`head -c3 -- "${a}"`" == $'\xef\xbb\xbf' ] && echo "Match: ${a}";done
function sepath { echo $PATH |tr ":" "\n" |sort -u |while read L ; do cd "$L" 2>/dev/null && find . \( ! -name . -prune \) \( -type f -o -type l \) 2>/dev/null |sed "s@^\./@@" |egrep -i "${*}" |sed "s@^@$L/@" ; done ; }
2009-09-11 15:03:22
User: mobidyc
Functions: cd echo egrep find read sed sort tr
Tags: bash ksh PATH

search argument in PATH

accept grep expressions

without args, list all binaries found in PATH

yes "$(seq 1 255)" | while read i; do printf "\x1b[48;5;${i}m\n"; sleep .01; done
du -a --max-depth=1 | sort -n | cut -d/ -f2 | sed '$d' | while read i; do if [ -f $i ]; then du -h "$i"; else echo "$(du -h --max-depth=0 "$i")/"; fi; done
2009-09-03 20:43:43
User: nickwe
Functions: cut du echo read sed sort

Based on the MrMerry one, just add some visuals to differentiate files and directories

find . -maxdepth 1 -type d|xargs du -a --max-depth=0|sort -rn|cut -d/ -f2|sed '1d'|while read i;do echo "$(du -h --max-depth=0 "$i")/";done;find . -maxdepth 1 -type f|xargs du -a|sort -rn|cut -d/ -f2|sed '$d'|while read i;do du -h "$i";done
2009-09-03 20:33:21
User: nickwe
Functions: cut du echo find read sed sort xargs

Based on the MrMerry one, just add some visuals and sort directory and files

while read server; do ssh -n user@$server "command"; done < servers.txt
2009-08-29 06:52:34
User: sharfah
Functions: read ssh

The important thing to note in this command, is the "-n" flag.

ls *.c | while read F; do gcc -Wall -o `echo $F | cut -d . -f 1 - ` $F; done
2009-08-28 13:01:56
User: pichinep
Functions: cut gcc ls read

Compile *.c files with "gcc -Wall" in actual directory, using as output file the file name without extension.

ls | while read f; do mv "$f" "${f// /_}";done
find /dir | awk '{print length, $0}' | sort -nr | sed 's/^[[:digit:]]* //' | while read dirfile; do outfile="$(echo "$(basename "$dirfile")" | unaccent UTF-8)"; mv "$dirfile" "$(dirname "$dirfile")/$outfile"; done
2009-08-24 21:24:18
User: Patola
Functions: awk basename find mv read sed sort

This command changes all filename and directories within a directory tree to unaccented ones. I had to do this to 'sanitize' some samba-exported trees. The reason it works might seem a little difficult to see at first - it first reverses-sort by pathname length, then it renames only the basename of the path. This way it'll always go in the right order to rename everything.

Some notes:

1. You'll have to have the 'unaccent' command. On Ubuntu, just aptitude install unaccent.

2. In this case, the encoding of the tree was UTF-8 - but you might be using another one, just adjust the command to your encoding.

3. The program might spit a few harmless errors saying the files are the same - not to fear.

pgrep -u `id -u` firefox-bin || find ~/.mozilla/firefox -name '*.sqlite'|(while read -e f; do echo 'vacuum;'|sqlite3 "$f" ; done)
2009-08-22 10:36:05
User: kamathln
Functions: echo find read

Sqlite database keeps collecting cruft as time passes, which can be cleaned by the 'vacuum;' command. This command cleans up the cruft in all sqlite files relating to the user you have logged in as. This command has to be run when firefox is not running, or it will exit displaying the pid of the firefox running.

ls -1 *.jpg | while read fn; do export pa=`exiv2 "$fn" | grep timestamp | awk '{ print $4 " " $5 ".jpg"}' | tr ":" "-"`; mv "$fn" "$pa"; done
2009-08-10 00:52:22
User: axanc
Functions: awk export grep ls mv read tr

Renames all the jpg files as their timestamps with ".jpg" extension.

md5sum * | sed 's/^\(\w*\)\s*\(.*\)/\2 \1/' | while read LINE; do mv $LINE; done
find . -iname '*filename*.doc' | { while read line; do antiword "$line"; done; } | grep -C4 search_term;
2009-07-28 15:49:58
User: Ben
Functions: find grep read

Find Word docs by filename in the current directory, convert each of them to plain text using antiword (taking care of spaces in filenames), then grep for a search term in the particular file.

(Of course, it's better to save your data as plain text to make for easier grepping, but that's not always possible.)

Requires antiword. Or you can modify it to use catdoc instead.

while true ; do IFS="" read i; echo "$i"; sleep .01; done < <(tr -c "[:digit:]" " " < /dev/urandom | dd cbs=$COLUMNS conv=unblock | GREP_COLOR="1;32" grep --color "[^ ]")
2009-07-22 03:59:07
Functions: dd echo grep read sleep tr true
Tags: color

Same as above but slooooow it down

dpkg -S `which nm` | cut -d':' -f1 | (read PACKAGE; echo "[${PACKAGE}]"; dpkg -s "${PACKAGE}"; dpkg -L "${PACKAGE}") | less
2009-07-14 20:21:48
User: mohan43u
Functions: cut echo read
Tags: which dpkg

In Debian based distros, this command will list 'binutils' package details which contains 'nm' command. You can replace 'nm' to any other command.

ssh root@server 'tail --max-unchanged-stats=10 -n0 -F /var/log/auth.log ' | grep Accepted | while read l ; do kdialog --title "SSH monitor" --passivepopup "$l" 3; done
yes "$(seq 232 255;seq 254 -1 233)" | while read i; do printf "\x1b[48;5;${i}m\n"; sleep .01; done