Hide

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.

Hide

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:

Hide

News

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.
Hide

Tags

Hide

Functions

Commands using dirname from sorted by
Terminal - Commands using dirname - 20 results
IFS=$(echo -en "\n\b"); input="/my/input/dir/*.flac"; mkdir -p $(dirname $f)/mp3; for f in $input; do ffmpeg -i $f -ab 196k -ac 2 -ar 48000 $(dirname $f)/mp3/$(basename "${f:0:${#f}-4}mp3"); done
function cdf () { [ -f $1 ] && { cd $(dirname $1); } || { cd $1 ; }; pwd; };
2012-09-08 10:50:58
User: Josso
Functions: cd dirname
Tags: bash ksh
0

Had trouble with the other function, because of missing semicolons. (According to my bash on OS X)

find Files/ -type d | parallel 'mkdir -p /BKP/{}' && find Files/ -type f | parallel 'rsync -a {} MKD/$(dirname {})'
2012-08-08 21:01:37
User: phribbr
Functions: dirname find
0

Copy files and dir in parallel. It is Faster.

Using the gnu-parallel.

Has 2 commands:

- First - Create dir structure in /BKP

find Files/ -type d | parallel 'mkdir -p /BKP/{}'

- Second - Copy for structure created

find Files/ -type f | parallel 'rsync -a {} /BKP/$(dirname {})'

- Great for backups!

- Can use "rsync" or "cp".

- Compare with a simple "rsync" or "cp"!

ls -ltr --directory $(find . -regex "./.*[^/]*\'" -type f | xargs -n 1 dirname | sort | uniq)
2012-03-02 03:48:47
User: pdkl95
Functions: dirname find ls sort xargs
0

This let me find some a set of modifications that were made to a rather large tree of files, where the file-names themselves were not unique (actually: insanely redundant and useless. "1.dat 2.dat ..."). Pruning down to last-branch brough things back to the "project-name" scope, and it's then easy to see which branches of the tree have recently changed, or any other similar search.

Ideally, it should sort the directories by the mtime of the most recent *file* *inside* the directory, but that's probably outside the scope of a (sane...) command line.

dirname $(readlink -f ${BASH_SOURCE[0]})
BASEDIR=$(dirname $(readlink -f $0))
STARTING_DIR=$(cd $(dirname $0) && pwd)
2011-11-30 17:35:15
User: bbbco
Functions: cd dirname
0

Sometimes you need the full path to your script, regardless of how it was executed (which starting directory) in order to maintain other relative paths in the script.

If you attempt to just use something simple like:

STARTING_DIR="${0%/*}"

you will only get the relative path depending on where you first executed the script from.

You can get the relative path to the script (from your starting point) by using dirname, but you actually have to change directories and print the working directory to get the absolute full path.

myreadlink() { [ ! -h "$1" ] && echo "$1" || (local link="$(expr "$(command ls -ld -- "$1")" : '.*-> \(.*\)$')"; cd $(dirname $1); myreadlink "$link"; }
2011-09-13 11:02:27
User: keymon
Functions: cd command dirname echo ls
0

This is a equivalent to the GNU ' readlink' tool, but it supports following all the links, even in different directories.

An interesting alternative is this one, that gets the path of the destination file

myreadlink() { [ ! -h "$1" ] && echo "$1" || (local link="$(expr "$(command ls -ld -- "$1")" : '.*-> \(.*\)$')"; cd $(dirname $1); myreadlink "$link" | sed "s|^\([^/].*\)\$|$(dirname $1)/\1|"); }
dirname `pwd`
cd /etc/network/if-up.d && iptables-save > firewall.conf && echo -e '#!/bin/sh -e\niptables-restore < $(dirname $0)/firewall.conf' > iptables && chmod a+x iptables
mydir=$(cd $(dirname ${BASH_SOURCE:-$0});pwd)
2011-04-27 16:33:38
User: xeor
Functions: cd dirname
Tags: cd script pwd
0

I submitted a command like this without $0 if $BASH_SOURCE is unset. Therefor, it did only work when using ./script, not using 'sh script'. This version handles both, and will set $mydir in a script to the current working directory. It also works on linux, osx and probably bsd.

current_dir=$(cd $(dirname $BASH_SOURCE);pwd)
2011-04-18 09:24:14
User: xeor
Functions: cd dirname
-2

I think this is the cleanest way of getting the current working directory of a script. It also works on osx, Linux, and probably bsd as well..

find ~ -maxdepth 2 -name .git -print | while read repo; do cd $(dirname $repo); git pull; done
cd() { if [ -z "$1" ]; then command cd; else if [ -f "$1" ]; then command cd $(dirname "$1"); else command cd "$1"; fi; fi; }
2010-04-23 19:17:43
User: xeor
Functions: cd command dirname
Tags: cd command
9

This little function will smarten 'cd'. If you try to cd into a file (which I guess we all have done), it cd's into the directory of that file instead.

I had to use nesten if's, to get cd to still work with 'cd' (to get to $HOME), 'cd -' (to get to last directory), and 'cd foo\ bar'.

script_path=$(cd $(dirname $0);pwd)
2009-10-14 16:04:03
User: jgc
Functions: cd dirname
Tags: cd pwd PATH
7

Another way of doing it that's a bit clearer. I'm a fan of readable code.

find . -iname ".project"| xargs -I {} dirname {} | LC_ALL=C xargs -I {} svn info {} | grep "Last Changed Rev\|Path" | sed "s/Last Changed Rev: /;/" | sed "s/Path: //" | sed '$!N;s/\n//'
2009-10-07 16:13:27
User: hurz
Functions: dirname find grep info sed xargs
0

Searches for all .project files in current folder and below and uses "svn info" to get the last changed revision. The last sed joins every two lines.

function fcd () { [ -f $1 ] && { cd $(dirname $1); } || { cd $1 ; } pwd }
2009-09-03 18:58:13
User: relay
Functions: cd dirname pwd
Tags: bash ksh
1

fcd : file change directory

A bash function that takes a fully qualified file path and cd's into the directory where it lives. Useful on the commadline when you have a file name in a variable and you'd like to cd to the directory to RCS check it in or look at other files associated with it.

Will run on any ksh, bash, likely sh, maybe zsh.

mv $1 $2 && ln -s $2/$(basename $1) $(dirname $1)
cd $(dirname $(find ~ -name emails.txt))
2009-05-01 21:26:58
User: haivu
Functions: cd dirname find
Tags: bash dirname
9

This command looks for a single file named emails.txt which is located somewhere in my home directory and cd to that directory. This command is especially helpful when the file is burried deep in the directory structure. I tested it against the bash shells in Xubuntu 8.10 and Mac OS X Leopard 10.5.6

for i in $(locate your_search_phrase); do dirname $i; done | sort | uniq
2009-02-05 14:03:20
User: realbrewer
Functions: dirname locate sort
1

Ever use 'locate' to find a common phrase in a filename or directory name? Often you'll get a huge list of matches, many of which are redundant, and typically the results are not sorted. This command will 'locate' your search phrase, then show you a sorted list of just the relevant directories, with no duplications. So, for example, maybe you have installed several versions of the java jre and you want to track down every directory where files matching "java" might exist. Well, a 'locate java' is likely to return a huge list with many repeated directories since many files in one directory could contain the phrase "java". This command will whittle down the results to a minimal list of unique directory names where your search phrase finds a match.