Useful Linux commands

Useful Linux commands

ack --sass "([\d\.]*)r?em" --output="\$1" -h | sort -u
	Find all unique em and rem values in all Sass files below the current directory

cat / less
	Output contents of a text file. ("less" is paginated but disappears after end of file, "cat" outputs in one go but output remains visible in console).

chown user:group /something
	Changes owner of /something to 'user', group to 'group'.
	
cp
	This copies the directory /home/jane, with all its files and subdirectories, and creates a subdirectory named jane in the current directory (.) :
	
	cp -r /home/jane .

/dev/null
	a black hole which unwanted output can be written to to silence a command.
	eg. eval `ssh-agent -s` &> /dev/null
	
df -h
	Show how much disk space is free
	
du -hs [directory]
	Display summary of a directory's disk usage in human-readable form  	

du -ch -d 1 | sort -hr
	Show a summary of all sub-directories' file sizes, sorted from largest to smallest

find
	Recursively search directory and subdirectories for files that match a pattern, eg:
	
	find . -name "*.ext"

git --git-dir=../somerepo/.git format-patch -k -1 --stdout 03ace01 | git am -3 -k
	Cherry-pick a single commit (with hash 03ace01) from somerepo

grep 'something' *.php
	Find all occurrences of 'something' in all the php files in the current directory
	
grep -c "." [filename]
	Echo number of lines in file (if replacing "." with a regex, echo number of matches in file)	

lame
	Batch encode all wavs in current directory:
	for f in *.wav; do lame -b320 $f; done
	
	Rename all files from .wav.mp3 to .mp3 and move them to ../mp3/
	for f in *.mp3 ; do AUFILE=`echo "$f" | sed 's/.wav//g'` ; mv -f "$f" "../mp3/$AUFILE"; done &
	
	(nb. mv -f forces overwrite of identically-named files without prompting)

less a.txt
	view contents of a.txt
	
less a.txt | grep http*
	display all hyperlinks in a.txt

ln -s a b
	Create a symbolic link "b" that links to file "a" 
	
ls -alFhv --color
	Display colour-coded verbose directory listing with human-readable file sizes (show all, and append / to directories)
	and ordered in proper numeric rather than alphabetical order (ie. listing 1000 after 999)

ls -l | wc -l
	Display number of files in a directory (ie. return line count from listing directory)

mkdir -p path/of/some/depth
	Create directory and all its parents, if they also don't yet exist (without the -p flag, mkdir complains if /path/of/some doesn't exist)

mkdir -p {dir1,dir2,dir3/subdir}
	Create multiple directories simultaneously. The above would create dir1/ dir2/ and dir3/subdir/
	The curly bracket thing is called "shell expansion". You could also mkdir topdir/{dir1,dir2}/{subdir1,subdir2}, which would create topdir/dir1 and topdir/dir2, each of which would have subdir1 and subdir2 inside them.
	
rm -r dir
	Delete directory "dir", all its contents, and all its sub-directories

scp username@remote.host:/path/to/file localfile
	Copy file from remote host to "localfile"
	
sudo /etc/init.d/apache2 restart
sudo apachectl graceful

	Restarts Apache (two alternatives)

tail -n
	list last n lines of file (default = 10)

type
	alternative to `which` that also works for aliases
	eg. `type gs` --> "gs is aliased to `git status'"

&
	putting & at the end of a command runs it in the background and returns you to the command line to get on with other stuff.

>
	send output of previous command to a file, eg:
	
	ls -R > dirlist.txt