Post and share your one liner

Haven't found a thread here where people are sharing there favorite one liners (except "Share your greatest one-liner!!") which ended pretty fast in 05'

I know allot of one liners will have to be built as needed and most will not likely be used everyday. Some maybe useless and some useful. But I find it fun to share and I hope people will benefit from this and maybe even some new aliases will be born!

[Please use code tags]

# Find last occurance of pattern and append string:
num=`awk '/<pattern>/ { print NR }' filename | tail -1`; sed "$num a <string>" filename

# Remove comments and blank lines
cat file | grep -v ^# | grep .

# Get memory total (lol)
a1=`ls -la /proc/kcore | awk '{print $5}'`;echo $(($a1/1024/1024));

# grep top 20 404's
grep "File does not exist" error_log | awk '{print $13}' | sort | uniq -c | sort -rn | head -20

Just to start

Originally Posted by micxz # Find last occurance of pattern and append string: num=`awk '/<pattern>/ { print NR }' filename | tail -1`; sed "$num a <string>" filename # Remove comments and blank lines grep file | grep -v ^# | grep . # Get memory total (lol) a1=`ls -la /proc/kcore | awk '{print $5}' | tr -d ''`;echo $(($a1/1024/1024)); # grep top 20 404's grep "File does not exist" error_log | awk '{print $13}' | sort | uniq -c | sort -rn | head -20

In the second example you should use cat instead of the first grep.

In the third example you can omit pipe tr -d ''.

In the fourth example you should use /var/log/httpd/ directory.

Originally Posted by micxz Haven't found a thread here where people are sharing there favorite one liners (except "Share your greatest one-liner!!") which ended pretty fast in 05'

It's more of an old "tips 'n tricks" thread but I'm sure there's oneliners too: http://www.linuxquestions.org/questi...d-hints-27957/


Most of my oneliners are functions. Examples:
secp: copy directory contents and keep SELinux contexts:

secp() { cd "$1" && tar --preserve --selinux -vcf - . | ( cd "$2" && tar --selinux -xBf - ); }

syscallName2Num: convert system call name to number (for Auditd logs mostly)
* both depend on HDRLOC="$(readlink -f /lib/modules/$(uname -r)/build)/include/asm-i386/unistd.h"

syscallName2Num() { grep "^#define __NR_${1}[[:blank:]]" $HDRLOC | awk '{print $3}'; }

syscallNum2Name: do the reverse

syscallNum2Name() { grep "^#define __NR_.*[[:blank:]]$1$" $HDRLOC | awk -F'_' '{print $4}'; }

shapertop: for watching the Wondershaper do it's work (or not)

shapertop() { /sbin/iptables -n -t mangle -L SHAPER -v -x | sort -grk1 | column -t; }

a variant of your "remove comments and blank lines":

stripcomments() { grep -v ^# | grep -v ^$; }

and a convoluted way to cut say lines 21 to 50 from some plaintext file:

showlines() { arg=($@); lines=$(wc -l ${arg[0]} | awk {print $1}); arg[1]=$[${lines}-${arg[1]}]; arg[2]=$[${lines}-${arg[2]}]; echo $-${arg[1]},$-${arg[2]} p | ed -s ${arg[0]}; }

Originally Posted by w1k0 In the second example you should use cat instead of the first grep. In the third example you can omit pipe tr -d ''. In the fourth example you should use /var/log/httpd/ directory.

Yer right' I fixed the first two but I'm hoping people will know where they're error_log is.

&Thanks_for_that(unSpawn);

The first command determines current date and ``highlights'' it in the output of cal command using < and >:

set date `date`; /usr/bin/cal | sed -e 's/^/ /' -e "s/ $4$/<$4>/" -e "s/ $4 /<$4>/"

      August 2009    
 Su Mo Tu We Th Fr Sa
                    1
  2  3  4  5  6  7  8
  9 10 11 12 13 14 15
 16 17 18 19 20 21<22>
 23 24 25 26 27 28 29
 30 31

The second command allows to perform different calculation in command line interface or in scripts:

alias c='perl -e '\''$_="@ARGV"; print eval $_, "\n"'\'''

I described it in that thread: Using Perl to perform calculations in Bash.

These aren't one liners but short scripts. I use them to find SUID or SGID files and directories, huge old files, huge files, huge directories, and files owned by the users or groups non-existing in the system.

1.1.suid

#!/bin/sh

# 1.1.suid seeks the system for the files having set SUID bit.
> 1.1.suid.`hostname`
echo "Still SUID or SGID files at `hostname`:" >> 1.1.suid.`hostname`
find / -mount -type f \( -perm -4000 -o -perm -2000 \) -exec ls -ld '{}' \; | \
    cut -d " " -f -4,8- >> 1.1.suid.`hostname`

1.2.suid

#!/bin/sh

# 1.2.suid seeks the system for the directories having set SUID bit.
> 1.2.suid.`hostname`
echo "Still SUID or SGID directories at `hostname`:" >> 1.2.suid.`hostname`
find / -mount -type d \( -perm -4000 -o -perm -2000 \) -exec ls -ld '{}' \; | \
    cut -d " " -f -4,8- >> 1.2.suid.`hostname`

2.huge+old

#!/bin/sh

# 2.huge+old seeks the system for the huge and old files (the files over
# 1 MB, modified or used for the last time three months ago or before).
> 2.huge+old.`hostname`
echo "Huge and old files at `hostname`:" >> 2.huge+old.`hostname`
find / -mount -size +10000k \( -mtime +120 -a -atime +120 \) -exec ls -l {} \; | \
    cut -d " " -f 5- | sort -nr >> 2.huge+old.`hostname`

3.huge-fil

#!/bin/sh

# 3.huge-fil seeks the system for the huge files exclusively.
> 3.huge-fil.`hostname`
echo "Huge files at `hostname`:" >> 3.huge-fil.`hostname`
find / -mount -size +10000k -exec ls -l {} \; | \
    cut -d " " -f 5- | sort -k4 >> 3.huge-fil.`hostname`

4.huge-dir

#!/bin/sh

# 4.huge-dir seeks the system for the huge directories.
> 4.huge-dir.`hostname`
echo "Huge directories at `hostname`:" >> 4.huge-dir.`hostname`
du -B 1000 / | grep -E '^[0-9]{7,}' >> 4.huge-dir.`hostname`

5.nobody

#!/bin/sh

# 5.nobody seeks the system for the files without any association to the
# user or group existing in the system.
> 5.nobody.`hostname`
echo "Nouser or nogroup files at `hostname`:" >> 5.nobody.`hostname`
find / -mount -nouser -o -nogroup | egrep -v "/dev/" >> 5.nobody.`hostname`

w1k0 I added those to my mini script folder. Thanks'

Here's another cool trick I read in Linux Server Hacks:

Filename "ssh-to"

#!/bin/sh

ssh `basename $0` $*;

Then within PATH make symlinks to that script with your favorite hostname. My bin folder looks like:

-rwxr-xr-x 1 micxz users   ssh-to
lrwxrwxrwx 1 micxz users   mars -> ssh-to
lrwxrwxrwx 1 micxz users   pluto -> ssh-to

Then you can simply type "hostname" and bang you have a shell on that server. (tip: even cooler if you've setup keys)

I just learned you can browse remote file systems via ssh in konqueror with the url fish://user@host

To hide splash screen:
sudo sed -i 's/Logo=1/Logo=0/g' /etc/openoffice/sofficerc
To unhide the splash screen:
sudo sed -i 's/Logo=0/Logo=1/g' /etc/openoffice/sofficerc

# download to stdout and extract with tar
curl <url> | tar czf -

This is one liner formatting text files with long lines. It breaks each line after 76 column and inserts empty lines between paragraphs. I called that script format:

#!/bin/sh
# formats text data
cat "$1" | perl -pe "s/^/.ll 76\n.na\n.nh\n/;s/\n/\n\n/" | nroff | cat -s

To format file.txt use the commands format file.txt in order to display the contents of the file or format file.txt > file-after-formatting.txt in order to store the contents in a new file.

I used tricky Perl command inserting nroff commands to the file before processing it with nroff. I’m sure there’s easier method of achieving the same result with AWK.

A quick way to check which device node your USB stick is using. Plug in your USB stick and enter code in terminal.

dmesg | egrep hd.\|sd. | grep 'logical'

Example output from my terminal

sd 0:0:0:0: [sda] 180941040 512-byte logical blocks: (92.6 GB/86.2 GiB)
sd 0:0:1:0: [sdb] 29886400 512-byte logical blocks: (15.3 GB/14.2 GiB)
sd 4:0:0:0: [sdc] 7837696 512-byte logical blocks: (4.01 GB/3.73 GiB)

As you can see, sdc is the USB stick because it's the capacity of the USB stick eg. 4 GB.

Of course there are other ways to determine which device node your USB is using.

- Cheers

dv502,

Your one liner informs about device but omits an information about partitions. So I improved it slightly:

dmesg | egrep 'hd.|sd.' | grep 'logical' | grep `sed 's/.*\[//;s/\].*//' | tail -n 1` /var/log/messages | tail -n 5