collection of bash commands which may be useful



Assign date to variable, useful for scripts

DATE_OF_RUN=$(date +%F_%H_%M)

Date X days ago

date +%Y-%m-%d -d '1 day ago'

Find files using newer modification time than X days ago.

find /directory -newermt $(date +%Y-%m-%d -d 'X day ago') -type f

Sort a file (which has content per line) and then output as single line

sort file | tr '\n' ' '

Words seperated by spaces in a file - convert to per line and sorted

cat /tmp/ubuntu_pkgs | tr " " "\n" | sort

Copy ACL

getfacl source-folder | setfacl --set-file=- target-folder

Replace with sed recursively (from current directory)

find . -type f -exec sed -i 's/old/new/g' {} +

Find and count files in directories... using a CWD, which ONLY contains directories (note the ls -1 used to feed from CWD).

ls -1 | while read -r dir; do printf "%s:\t" "$dir"; find "$dir" -type f | wc -l; done

Add first column of numbers in file without using bc, but awk

awk '{s+=$1} END {print s}' some_filename

...an alternative to the last command, numbers printed for each line and final total.

awk '{ sum+=$1;print $1} END {print "Sum";print sum}' some_filename

Remove newlines from file1 to file2

tr -d '\n' < file.txt > file2.txt

ps - Show top10 CPU utilisation, all processes, apply formatting

ps -Ao user:50,uid,comm,pid,pcpu,tty --sort=-pcpu | head

iostat

iostat -kxh

clear page cache (sudo/root)

sync; echo 1 > /proc/sys/vm/drop_caches

replace comment at the start of line (remove) identified by word baseurl

sed -e '/baseurl/ s/^#*//' file

replace content identified by word baseurl

sed -e '/baseurl/ s/old_words/new_words/' file

comment start of line

sed -e '/^.*word_to_match*/ s/^/#/' file