Category: bash

Bash keyboard shortcuts

15 February 201815 February 2018 Fabien bash, Linux Leave a comment

When you are spending your day in the terminal , you will find that navigating using only the arrows keys although fine at the the start will frustrate you because it’s quite slow.
Bash come with a lot of keyboard shortcuts design to gain precious time.
CTRL+A # will move the cursour to the beginning of the line CTRL+E # moves to end of line CTRL+C # halts the current command and back to prompt CTRL+D # deletes one character backward or logs out of current session, similar to exit CTRL+K # deletes all the text forward to the end of the line CTRL+L # clears screen and redisplay the line CTRL+R # searches history entering keyword , CTRL+T # transposes two characters CTRL+U # kills backward from point to the beginning of line CTRL+W # kills the word behind the cursor

i’ve put in bold the shortcuts that i use every day , i don’t use CTRL+A and CTRL+E because the keyboard has keys that are foing the same thing.

Bash brace expansion. (to delete files)

3 February 2018 Fabien bash, Linux Leave a comment

In bash , it is common to have to do some action on a numéric serie of files , like deleting or renaming ,

Using brace expansion you will be able to générate a single line that will act on multiple targets ,

For our example , let’s think about a list of ten files

log1.log log2.log log3.log log4.log log5.log
log6.log log7.log log8.log log9.log log10.log

in the simple case where i want to delete the complete list of files, i just have to run de command,

rm log*.log

The 10 files will be deleted, but , this is not practical if you want to keep the last file , and delete the other nine,

In that case we can use a usefull tool called brace expansion,

To delete the files 1 to 9 , i just have to run in my bash terminal!

rm log{1..9}.log

This last one is extremely practical , I use it very often in a lot of different uses cases.

But in other cases , you may want to delete even on odd files , or one every three files,

This is also doable with braces expansions

rm log{1..10..2} .log
#to delete  1 3 5 7 9
rm log{1..10..3}.log
#to delete 1 4 7 10

Redirect stdout and stderr to a file

22 November 201722 November 2017 Fabien bash Leave a comment

il you need to run a script unattended and wish to log the output of that script you must already know that you can simply do

 ./myscript.sh > mylog.log

the problem with that is you will not log the error messages , only the output messages ,
the error are going to be displayed on the terminal but , not logged
and the error message are often as important as the output message , we do not want to dismiss them from the log ,

the solution to that issue is to run the script using this

 ./myscript.sh > mylog.log 2>&1

With this , both the output will be logged in to the log file.

grafana service doesn’t auto start with the system

21 August 201721 August 2017 Fabien bash, Linux Leave a comment

When installing grafana and influxDB. I encountered the folowing issue

The service of grafana and influxDB where not starting .

service grafana-server status
grafana-server.service - Starts and stops a single grafana instance on this system
Loaded: loaded (/usr/lib/systemd/system/grafana-server.service; disabled; vendor preset: enabled)
Active: inactive (dead)
Docs: http://docs.grafana.org

I checked using rcconf that my service where enabled , and they where !

I then tried sysv-rc-conf and my service where correctly configured !

the services had absolutely no issue starting when called manually and started fine !

loaded (/usr/lib/systemd/system/grafana-server.service; disabled;

this line was telling me that the service startup was disabled !

after seaching a bit on the web ,the solution to this issue is using systemctl to enables the services.

systemctl enable grafana-server.service
systemctl enable influxdb.service

InfluxDB insert multiples values at once using curl

17 August 201717 August 2017 Fabien bash, Orange PI 2 Comments

If you have to insert multiples values in an influxdb database. You can use curl and insert the values one at the time.

But there is a better way like in this example

Let’s say you want to insert your tree loadaverage values in you influxdb database.

First put them into variables using these 3 lines.

onemin=$(cat /proc/loadavg |awk -F' ' '{print $1}')
fivemin=$(cat /proc/loadavg |awk -F' ' '{print $2}')
fifteenmin=$(cat /proc/loadavg |awk -F' ' '{print $3}')

Then using this curl line, insert them all at once in your influxdb database

curl -i -XPOST 'http://localhost:8086/write?db=database' --data-binary "loadavg 1min=$onemin,5min=$fivemin,15min=$fifteenmin"

OrangePI : Install OpenRTSP on a ARM device

13 August 201713 August 2017 Fabien bash, Linux, Orange PI 1 Comment

One of my project with my OrangePI is to setup the device as a basic IP security camera server ,

the OrangePI zero , will connect to RTSP flux of the IP camera , and dump the stream on to an external disk , where it will be kept for 7 days before getting automaticly deleted
At a latter date i will run a motion detection program on these files and send a mail if motion is detected on a specific part of the frame.

after a little research , i found that a lot of people are using a program called openRTSP to record their RTSP stream to disk

But , after connecting on my OrangePi zero and trying to look for the program i found out that it’s not part of the default installation of armbian stable.

I tried to install the package using apt install openRTSP

apt install openRTSP
Reading package lists... Done
Building dependency tree
Reading state information... Done
E: Unable to locate package openRTSP

I went to google and searched , openRTSP package ,and found out , on the debian website that openRTSP is part of the livemedia-utils package,

I went back to my SSH session and typed apt-get install livemedia-utils

root@orangepizero:~# apt-get install livemedia-utils
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following additional packages will be installed:
libbasicusageenvironment1 libgroupsock8 liblivemedia50 libusageenvironment3
The following NEW packages will be installed:
libbasicusageenvironment1 libgroupsock8 liblivemedia50 libusageenvironment3 livemedia-utils
0 upgraded, 5 newly installed, 0 to remove and 0 not upgraded.
Need to get 356 kB of archives.
After this operation, 1,187 kB of additional disk space will be used.
Do you want to continue? [Y/n] y

There is no man pages associated with the program , you have to rely on the developer website to understand all the different options . The documentation is available here : http://www.live555.com/openRTSP/ or here if the website is down.
but as usual , in the officials repositories you only get dated version

you might want to compile the source code to have the last version available:

as root

Go to /usr/src: cd /usr/src
Get the live555 liveMedia source code: wget http://www.live555.com/liveMedia/public/live555-latest.tar.gz
Unpack it: tar -xzf live555-latest.tar.gz
Go into the unpacked directory: cd live
Generate the make files: ./genMakefiles linux
Build the code: make
Install the latest version: make install

Bash script to clean movies name

22 June 201722 June 2017 Fabien bash, Linux 3 Comments

The issue when you download movies online, is that most of the time the file name is going to be complete trash , filled with useless info like the team that made the rip or the website from where the movie was downloaded.
This is very annoying when you sort by name and the website made the idiotic decision to put their name in front of the actual film name.

for example here a typical movie name
[ Torrent9.info ] Dont.Think.Twice.2016.FRENCH.WEBRip.XviD-NEWCiNE.avi
the actual useful data that i wish to keep is Dont.Think.Twice.2016.avi

for this purpose i’ve developped a small bash script that will list all the files in a chosen directory then , if necessary , will rename the file.

#!/bin/bash
#the location of the folder full of movies to rename.
for file in /media/freebox/*
do
if [[ -f $file ]]; then
#that line will delete all the text after the date but keep the file extension.
nfile=$(echo $file | sed "s/\(20[0-1][0-8]\).*\(mkv\|avi\|mp4\)/\1.\2/")

#this line is going to try to delete a space a the start of the filename if it exist
nfile=$(echo $nfile | sed  "s/\/ /\//")

#this line is going to delete double dots in the filename and replace them with simple dots
nfile=$(echo $nfile | sed -e "s/\.\./\./g")

#that line is going to delete everything between [] including the [] 
nfile=$(echo $nfile | sed -e 's/\[[^][]*\]//g')


#after all this work on the filename
#you can now check if the any changes have been made , 

if [[ "$file" == "$nfile" ]]
then
   #if no change display this
   echo "Not touched : $file"
else
   #if change detected , rename the file
   echo ______
   echo Renamed :
   echo "Original name : $file"
   echo "Renamed to  : $nfile"
   echo ______
#doing the actual renaming
mv "$file" "$nfile"
fi
fi
done

this is not a very efficient script
but it’s fulfill it’s purpose and very rarely encounter a filename that he is unable to clean.