Stick up your best shell scripts

Pygmalion

A nethack wrapper, if your nethack is built with the dumplog patch, it'll check to see whether the dumplog has been updated, and then copy it to ~/dumps/ with the date and time as filename, then make it world-readable.

#!/bin/bash

NAME="bunbun"
DUMP="/var/games/nethack/dumps/$NAME.lastgame.txt"

#Check whether the dump file exists, get last modified time if so, 0 otherwise
if [ ! -e $DUMP ]
then
    LDATE=0
else
    LDATE=$(date -r $DUMP +%s)
fi

#Execute nethack with the options given
nethack $*

#If dump exists and is newer than the one we saw earlier copy the dump
if [ -e $DUMP ]
then
    if [ $(date -r $DUMP +%s) -gt $LDATE ]
    then
        #Only eval `date ...` once, in case the second changes between these lines
        OUTFILE="$HOME/dumps/`date +%Y-%m-%d-%H:%M:%S`.txt"
        cp $DUMP $OUTFILE
        chmod 644 $OUTFILE
    fi
fi

Just change the username and location of dumplog file at the top, and ensure the dumps directory exists.

Khannie

Nothing super mega about this, but I use it all the time.

Set up key exchange with a remote server:

#!/bin/bash

ssh "$1" "echo `cat ~/.ssh/id_?sa.pub` >> ~/.ssh/authorized_keys2"
ssh "$1" "chmod 700 ~/.ssh/authorized_keys2"

usage:
./sendKeyTo <user@remoteserver>

Tillotson

Helps in my never-ending war against badly named mp3's. Converts files to Start Case, eg: "some file.mp3" -> "Some File.mp3".

#!/bin/bash

set -e

function usage {
    echo -e "USAGE: $0 [OPTION] [FILE]"
    echo -e "Converts filename to Start Case\n"
    printf "%4s%20s\n" "-d" "dry run"
    printf "%4s%20s\n" "-s" "verbose"
    echo -e "Example: $0 *.mp3"

}

function err_exit {
    echo -e "$1\n"
    usage
    exit 1
}

declare dry verbose

while getopts ":dv" opt; do
    case $opt in
        d) dry=y ;;
        v) verbose=y ;;
        \?) usage; exit 1 ;;
    esac
done

shift $((OPTIND-1))

for f0; do
    [[ -f $f0 ]] || err_exit "$f0: is not a file"

    # Keep extension unchanged
    f1="$(echo $f0 | cut -d'.' -f 1)"
    ext="$(echo $f0 | cut -d'.' -f 2)"
    f1="$(echo $f1 | sed 's/\b\(.\)/\u\1/g').$ext"

    if [[ ! -z $dry  ]] || [[ ! -z $verbose ]]
    then
        printf "%s%40s\n" "$f0" "-> $f1"
    fi
    if [[ $f0 != $f1 ]] && [[ -z $dry ]]    # Test filenames not the same
    then
        mv "$f0" "$f1"
    fi
done

KAGY

Khannie wrote: »

Set up key exchange with a remote server:

<snip>
usage:
./sendKeyTo <user@remoteserver>

ssh-copy-id does the same thing, part of openssh i think,

ssh-copy-id [-i [identity_file]] [user@]machine

, though your's is easier to use as the rsa file location is hard coded

Khannie

I use this all the time for using up all my cores on apps that aren't multithreaded (mostly we're talking about oggenc and convert (part of imagemagick)).

#!/bin/bash

NUMCPU="$(grep ^processor /proc/cpuinfo | wc -l)"
     
until [ -z "$1" ]
do
	while [ `jobs -p | wc -l` -ge $NUMCPU ] ; do
		sleep 0.1
    done
  	oggenc -q8 "$1" &
	shift
done

Usage would be:
oggenc_cpus.sh *.flac

edit: be somewhat careful with hyperthreaded CPU's (mine is not).

PrzemoF

A simple script to monitor modification date of a file and trigger an action.
Example use:
0. Save the script to trig.sh file and make executable

chmod a+x trig.sh

1. start script with a tex file name as parameter i.e.

./trig.sh my_tex_file.tex

2. Open the tex file in an editor:

vim my_tex_file.tex

3. Make any modifications to the file and save it - it starts pdflatex that produces my_tex_file.pdf
4. Open evince to see the content of the pdf file

evince my_tex_file.pdf

From now on every saved modification of my_tex_file.tex will be caught by trig.sh. trig.sh will run pdflatex and evince will show updated version of the pdf file.

trig.sh:

#!/bin/bash
# Delay between checks
DELAY=1
DATE_ORIG=$(stat -c %Y $1)
DATE_MOD=$DATE_ORIG

while [ 1 ];
do
        if [ $DATE_ORIG -ne $DATE_MOD ]; then
                #File changed - add your action here
                pdflatex $1
                DATE_ORIG=$DATE_MOD
        else
                sleep $DELAY
                DATE_MOD=$(stat -c %Y $1)
        fi
done

Fenster

I blog a lot of photographs.

1. Take in a folder name followed by a list of images as arguments.
2. Create said, copy images to it as JPG, and rename them in sequential order.
3. Create a smaller thumbnail size of the images to fit my blog.
4. Move this folder to my Dropbox Public folder.
5. Insert HTML code for the images into my clipboard.

#!/bin/bash

dropbox="$HOME/Dropbox/Public/content"
uuid="4144919"
thumbs="m"
large_size=1500
thumb_size=600
temp="/tmp/prep"

prune_spaces() {
    cp "$1" $(echo "$1" | sed -e 's/ /_/g')
}

add_line() {
    echo "<a title=\"CHANGE ME\" href=\"http://dl.dropboxusercontent.com/u/$1/content/$2/$3\"><img src=\"http://dl.dropboxusercontent.com/u/$1/content/$2/m/$3\" alt=\"CHANGE ME\" /></a>" >> $4
}

resize_image() {
    mogrify -quality 100 -format jpg -resize "$1"x\> $2
}

if [ ! -d $1 ]; then
    mkdir -p $1/$thumbs
fi

if [ -d $dropbox/$1 ]; then
    rm -r $dropbox/$1
fi

src=$1
shift

count=1

for n in "$@"; do
    if [ -e "$1" ]; then
        cp "$1" $src/$count.jpg
        cp "$1" $src/$thumbs/$count.jpg
        let count++
        shift
    fi
done

cd $src

if [ $(ls -1 *.jpg 2> /dev/null | wc -l) == 0 ]; then
    echo "No images selected!"
    rm -r $(pwd)
    exit 1
fi

for n in *.jpg; do
    resize_image $large_size $n
    resize_image $thumb_size $thumbs/$n
    add_line $uuid $src $n $temp
done

cat $temp | xclip -sel clip
rm $temp
mv $(pwd) $dropbox

exit 0

Keplar240B

My First Arch Linux script , a backup utility.

#!/bin/bash

#GWB 22-06-16
#Backup script backs up files to remote directory
#1: Make sure destination  is mounted
#2: Run as root



echo -e  "\e[42m******************************\e[0m"
echo -e  "\e[42m*Written By Gwb     *\e[0m"
echo -e  "\e[42m*Verision 3 21-06-16         *\e[0m"
echo -e  "\e[42m*Backup script Backup_3.sh   *\e[0m"
echo -e  "\e[42m******************************\e[0m"
echo " "

#Check that user ran as sudo
if (( $EUID != 0 )); then
    echo -e "\e[41mPlease run as root\e[0m"
    exit
fi

#get user input for which path for backup
Dest1="/run/media/gwb/Linux_backup"
Dest2="/run/media/gwb/iomeaga_320"
echo -e  "\e[44mPick destination directory for backup :--  \e[0m"
echo "(1)   "  "$Dest1"
echo "(2)   "  "$Dest2"
echo "(3)    Specify a path" 
echo "(4)    exit"
echo ""
echo "Press option followed by [ENTER]"
echo ""
read -r choice
#check that paths exist and change path to dest
case "$choice" in

1)  echo $Dest1
    if  [ -d $Dest1 ] 
    then
      cd $Dest1
    else
	  echo  -e "\e[91mPath not found to destination directory\e[0m"
	  echo  -e "\e[91mNOTE : The Hard drives  internal and external must be mounted\e[0m"
          echo ""
          exit
    fi
    ;;

2)  echo  $Dest2
    if  [ -d $Dest2 ] 
    then
      cd $Dest2
    else
	  echo  -e "\e[91mPath not found to destination directory\e[0m"
	  echo  -e "\e[91mNOTE : The Hard drives  internal and external must be mounted\e[0m"
          echo ""
          exit
    fi
    
    ;;
3)  echo -e "\e[44mType a custom destination path\e[0m"
    read -r Path1
    if  [ -d $Path1 ] 
    then
      echo "$Path1"
      cd $Path1
    else
	  echo  -e "\e[91mPath not found to destination directory\e[0m"
          echo ""
          exit
    fi
    ;;

*) echo "exit"
    exit
   ;;
esac

#make the backup directory
TODAYSBACKUPDATE=`date +%d-%b-%Y`
mkdir $TODAYSBACKUPDATE
cd $TODAYSBACKUPDATE
echo "Backup Directory made at " "$TODAYSBACKUPDATE" at "backup path"
echo ""

#begin the backup
echo -e "\e[30;48;5;10m1 Make copy of first 512 bytes MBR with dd\e[0m"
dd if=/dev/sdb1 of=hda-mbr.bin bs=512 count=1
echo -e "\e[1;10mDone.\e[0m"
echo ""

echo -e "\e[30;48;5;10m2 Make a copy of etc dir\e[0m"
cp -a  -u /etc .
echo -e "\e[1;10mDone.\e[0m"
echo ""

echo -e "\e[30;48;5;10m3 Make a copy of home dir\e[0m"
cp -a  -u /home .
echo -e "\e[1;10mDone.\e[0m"
echo ""
sync


echo -e "\e[30;48;5;10m4 Make tarball of all except tmp dev proc sys run\e[0m"
tar --one-file-system --exclude=/tmp/* --exclude=/dev/* --exclude=/proc/* --exclude=/sys/* --exclude=/run/* -pzcf RootFS_backup.tar.gz /
echo -e "\e[1;10mDone.\e[0m"
echo ""
sync

echo -e "\e[30;48;5;10m5 Make copy of package lists\e[0m"
pacman -Qqen > pkglist.txt
pacman -Qm > pkglistAUR.txt
echo -e "\e[1;10mDone.\e[0m"
echo ""

Keplar240B

My Second Arch Linux script , a maintenace and update utility. uses cower as a AUR helper an clamav as antivirus

#!/bin/bash

#Gwb 23-06-16
#Arch Linux distro update script
#update system official and AUR repo using pacman and cower
#check for broken symlinks, log errors and failed systemd services
#and run anti virus check clam and update the database.
echo -e "\e[40;38;5;82m*********************************\e[0m"
echo -e "\e[40;38;5;82m*gwb 23-06-16 updatman.sh *\e[0m"
echo -e "\e[40;38;5;82m*Arch Linux distro update script*\e[0m"
echo -e "\e[40;38;5;82m*update system official and      *\e[0m"
echo -e "\e[40;38;5;82m*AUR repo using cower and pacman*\e[0m"
echo -e "\e[40;38;5;82m*check for broken symlinks,     *\e[0m"
echo -e "\e[40;38;5;82m*log errors and failed systemd  *\e[0m"
echo -e "\e[40;38;5;82m*services and anti virus check  *\e[0m"
echo -e "\e[40;38;5;82m*********************************\e[0m"
echo ""
#cower AUR
cd /home/gwb/Downloads
TODAYSBACKUPDATE=`date +%d-%b-%Y`
mkdir $TODAYSBACKUPDATE
cd $TODAYSBACKUPDATE
echo -e "\e[30;48;5;10m***** UPDATE COWER PACKAGES *****\e[0m"
echo "Directory $TODAYSBACKUPDATE made at /home/gwb/Downloads for cower updates"
echo ""
cower -vdu
echo ""
# look for empty dir (if no updates) 
if [ "$(ls -A .)" ] 
then
     	echo -e  "\e[44mCower updates available for package build\e[0m"
     	echo -e  "\e[44mDo you wish to build and install  them now?\e[0m"
      	echo ""
	echo "Pick Option"
	echo "1) Yes"
	echo "2) No"
	echo ""
	echo "Press 1 or 2 followed by [ENTER]"
	echo ""
	read choice
        if [ $choice = "1" ]
	then
		echo "Building and installing cower package updates"	
		echo ""	
		#build and install packages
               find . -name PKGBUILD -execdir makepkg -si \;
			
	fi	

else
    	echo ""
	echo " No updates of AUR packages by Cower..."
fi	
echo ""
echo -e "\e[1;10mDone.\e[0m"
echo ""

# -systemd --failed:
echo -e "\e[30;48;5;10m***** ALL FAILED SYSTEMD SERVICES *****\e[0m"
systemctl --failed --all
echo ""
echo -e "\e[30;48;5;10m***** FAILED ACTIVE SYSTEMD SERVICES *****\e[0m"
systemctl --failed
echo ""
# -Logfiles:
echo -e "\e[30;48;5;10m***** LOGFILES *****\e[0m"
cd /home/gwb/Documents/Tech/Linux/MyLinux
journalctl -p 3 -xb >errorfile
echo -e "\e[1;10mDone.\e[0m"
echo ""
# Checking for broken symlinks:
echo -e "\e[30;48;5;10m***** CHECKING FOR BROKEN SYMLINKS *****\e[0m"
cd ~
find . -type l -! -exec test -e {} \; -print > /home/gwb/Documents/Tech/Linux/MyLinux/log.txt
echo -e "\e[1;10mDone.\e[0m"
echo ""
#pacman
cd ~
echo -e "\e[30;48;5;10m***** UPDATE PACMAN *****\e[0m"
echo ""
sudo pacman -Syu
echo ""
echo "Delete orphans"
sudo pacman -Rns $(pacman -Qtdq)
echo ""
echo "Prune older packages from cache"
sudo paccache -r
echo "Writing installed package lists to file "
sudo pacman -Qqen > /home/gwb/Documents/Tech/Linux/MyLinux/packagesQe.txt
sudo pacman -Qm > /home/gwb/Documents/Tech/Linux/MyLinux/package.txt
echo -e "\e[1;10mDone.\e[0m"    
echo ""
#Anti-virus clam Av
# update clamscan virus definitions:
echo -e "\e[30;48;5;10m***** UPDATING CLAMAVSCAN DATABASE *****\e[0m"
sudo freshclam
echo""
echo -e "\e[30;48;5;10m***** SCANNING / RECURSIVELY WITH CLAMAV*****\e[0m"
echo "Do you wish to run anti-virus check with clamAv at this point?"
echo ""
echo "Pick Option"
echo "1) Yes"
echo "2) No"
echo ""
echo "Press 1 or 2 followed by [ENTER]"
echo ""
read choice1
echo "You picked option " $choice1 
echo "" 
if [ $choice1 = "1" ]
	then
		# scan entire system
		
		sudo clamscan --recursive=yes --infected --exclude-dir='^/sys|^/proc|^/dev|^/lib|^/bin|^/sbin' /
		echo "" 
		echo -e "\e[1;10mDone.\e[0m"
			
	else
		echo -e "\e[1;10mDone.\e[0m"
		exit
fi

sKeith

Here is one I call rsync.no.quit as it will keep on going until job is finished. I needed it some time ago when I was migrating a huge amount of data off a server where rsync would keep failing with some pipe issue. (most likely flakey network or something similar)
Set this script up with source and destination and fire and forget.

*Warning* With the current rsync commands used Destination will mirror source. Any files or folders in destination will be destroyed /*Warning*

#!/bin/bash

source="192.168.1.1:/home/main/"
destination="/mnt/backup/"

if [ -z $source ]; then
        echo "Source unset"
        exit 1
fi
if [ -z $destination ]; then
        echo "Destination unset"
        exit 1
fi

echo Warning: Everything in $destination will we destroyed.
read -n1 -t5 -r -p "CRLT+C to cancel" key
echo
while [ 1 ];    do
rsync --progress --timeout=30 -a --delete-during $source $destination
if [ "$?" = "0" ] ; then
        echo "rsync completed normally"
        exit
else
        echo "Rsync failure. Backing off and retrying..."
        sleep 10
fi
done