This is an old revision of the document!
Table of Contents
Bash
: This isn't really bash, it's random Linux stuff that I'm tired of looking up.
HowTo
Find the subnet given IP and mask or CIDR
ipcalc -n 172.16.16.113 255.255.255.252 | grep Network | awk '{print $2}'
172.16.16.112/30
This assumes you have ipcalc installed (Github and download links at bottom of that page.) : This shouldn't be under Linux.
Pad an IP to make it sortable
echo 172.16.16.112/30 | awk -F '[./]' '{printf "%03d.%03d.%03d.%03d/%03d\n", $1,$2,$3,$4,$5}'
172.016.016.112/030
Date
My preferred date/time format is ISO 8601.
date '+%Y-%m-%d_%H:%M:%S'
Can also be accomplished with…
date '+%F_%T'
Specific timezone and 3 hours in the future…
TZ="America/New_York" date -d "+3 hours" '+%Y-%m-%d_%H:%M:%S'
Command Substitution
Backticks or the other thing I never remember.
`ping -c10 -W2 1.1.1.1`
$(ping -c10 -W2 1.1.1.1)
Loops
for x in 0{1..9} {10..12} ; do mkdir 2016/2016-$x
while read LINE; do echo $LINE; done < FILE
Ping Sweep with For Loop
From: https://www.rubyguides.com/2012/02/cli-ninja-ping-sweep/
Linux Ping Sweep
for i in {1..254} ;do (ping -c 1 192.168.1.$i | grep "bytes from" &) ;done
What this does is a for loop from 1 to 254, $i takes the value of the current iteration so in the first one it will be 1 then 2, 3… and so on, then we tell it to call the ping command with the -c option which means only ping once otherwise it would ping forever after that we pipe the output to grep so we only see the hosts that actually responded and the & at the end send it to the background so it will launch all the pings in parallel. If we only want the ip address and not the whole line we can further filter this using cut.
.inputrc
To reload after editing…
bind -f ~/.inputrc
My usual stuff
# Make auto-complete cycle through options instead of listing them all TAB: menu-complete # Make shift tab reverse the above "\e[Z": "\e-1\C-i" # Use up and down arrow to search command history. Invaluable! "\e[A": history-search-backward "\e[B": history-search-forward # Use Control-g to keep a command in history without executing "\C-g": "\C-a history -s \C-j"
find
I never remember how to find old files… either of these will work. The second one uses minutes (I don't know why you'd want to) and shows how to exclude files.
find DIR -name "*_FILENAME-*.gz" -mtime +45 find . ! -name '*NOT_THIS*.gz' -mmin +$((45*24*60))
Add this to the end and anything that matches will be deleted.
-exec rm -f {} \;
Limit depth to current directory.
find . -maxdepth 1 -name "v*.tar.gz" -mtime +30
Sort These
# to delete empty dirs
find -type d -empty -delete
# to delete files over some age & empty dirs
find /some/path -depth \( \( -type f -daystart -mtime +100 \) -o -type d -empty \) -ls
=======================
# Check if script is cron or shell
https://stackoverflow.com/questions/3214935/can-a-bash-script-tell-if-its-being-run-via-cron
if [ -t 1 ] ; then
echo "interacive mode";
else
#send mail
fi
-----------------------
CRON=$(pstree -s $$ | grep -q cron && echo true || echo false)
then test with
if $CRON
then
echo "Being run by cron"
else
echo "Not being run by cron"
fi
=======================
