Skip to content

Latest commit

 

History

History
344 lines (265 loc) · 7.41 KB

Linux Session #5.md

File metadata and controls

344 lines (265 loc) · 7.41 KB

Session #5

System Administration

Sed utility :

  • Sed stands for streamline editor

  • It does not change your file unless the output is saved with shell redirection by default

  • Sed editor process a file (input) one line at a time and send output to the screen

  • Sed stores the line it process in a buffer and once processing is finished the line is sent to the screen (unless command wad to delete the line) and the next line is read and the process is repeated until the last line is reached

Addressing :

  • Its used to determine whcih lines to be edited Addressing format can be:
  1. Number
  2. Regular Expression
  3. Both
  • Sed command tells sed what to do with line whether:
  1. Print it
  2. Remove it
  3. Change it
  • Sed command format
sed 'Command' FileName
  • Print lines that contains word "root"
sed '/root/p' myfile

word is put inside backward slashes [/word/]
p -> For Printing the line that contains word

  • Prevent default behaviour of Sed to print each line
sed -n '/root/p' myfile

-n : a flag

  • Print only Line #1
sed -n '1p' /etc/password
  • Print range from Line #1 to Line #10
sed -n '1,10p' /etc/password
  • Print from Line that contains word "games" to line that contains word "root"
sed -n '/games/,/root/p' /etc/password
  • Print all Lines except Line #3
sed -n '3!p' Myfile
  • Delete Line #3 {Just in terminal not actual file}
sed '3d' Myfile
  • Nothing will be shown in terminal
sed -n '3d' Myfile
  • Delete all Lines that contain word "Hello"
sed '/Hello/d' Myfile
  • Delete all Lines that starts with word "Hello"
sed '/^Hello/d' Myfile
  • Delete all Lines that contain word "Hello" and then quite
sed '/Hello/d;/^Hello/q' /etc/password

[;] : for more than one command
q : to quite Note: not working with him

  • Delete all Lines that ends with word "Hello"
sed '/Hello$/d' File2.txt
  • To save file with new edit
sed -i '/^Hello/d' File2.txt

Note: Not recommended to use it, as you deal with sensitive files like passwd, so save your changes in another file {redirection}

sed -i '/^Hello/d' File2.txt > File2_Temp.txt
  • To edit a pattern using sed, this command will change every "everybody" word to "everyone"
sed 's/everybody/everyone' Myfile2

s: stands for substitute

  • If there are two "everybody" in the same line {ex: .... everybody .... everybody} its going to just change first word and then break to next line without changing other word in same line, to fix this
sed 's/everybody/everyone/g' Myfile2

g: stands for global

  • To print lines that are changed
sed -n 's/everybody/everyone/gp' Myfile2

p : for print

  • To delete from line #3 till the end
sed '3,$d' Myfile
  • To make two Sed command in one line, this command will print from line #1 to line #3, then from line #6 to line #7
sed -n -e '1,3p' -e '6,7p' Myfile

AWK Utility :

  • Named AWK for first letter of name of the three developer of it

  • its a programming language used for manipulating data and generating reports

  • AWK scans a file line by line, search for lines that match a specific patterns and then performing selected action

  • Awk command format

awk 'Instructions' FileName

  • In Awk, Line = Record

  • Important variables:

    1. FS -> you can control field separator {default is space but you can change it to colon [:] as in [/etc/passwd] }
    2. RS -> record separator which is '\n' by default
    3. $0 -> Means entire record {$1 for field #1 , $2 for field #2 , etc ...}
    4. NR -> Means record number
    5. NF -> Number of fields in a record {7 as in passwd}

  • Print first field for each line

awk -F: '{print $1}' /etc/passwd

F: specify that field separator is [:]

  • Print field 1,3 for each line
awk -F: '{print $1,$3}' /etc/passwd
  • Print "Log name:" text with first field for each line
awk -F: '{print "Log name:",$1}' /etc/passwd
  • Display whole file {like $ cat}
awk '{print $0}' /etc/passwd
  • Display Number of each line along with line itself
awk '{print NR,$0}' /etc/passwd
  • Display each line + Number of fields for each one
awk '{print $0,NF}' /etc/passwd
  • BEGIN is executed one time before awk process any line from the input file
  • It determines first value of FS,RS variables
awk 'BEGIN {FS = ":";RS="\n"} {print $0,NR}' 2-empty-spaces.txt
  • END is executed one time after awk process all lines of input file
  • It does not match any input line
awk 'END {print NR} {print NR}' myfile

Conditional Expression :

if(expression1) {statement1;statement2}
else if(expression2) {statement1;statement2}
else {statement1}

Operators: < <= == != >= > ~ !~
~ -> Match a regular expression
!~ -> Not match a regular expression

Loops:

  • While Loop
awk -F: '{i=1;while(i<=NF){print NF,$i;i++}}' /etc/passwd

It will print 7 fields for each line, each in a new line

  • For Loop
awk -F: '{for(i=1;i<=NF;i++){print NF,$i}}' /etc/passwd

awk -F: '{if($1>$2)max=$1;else max=$2;print max}' awk.txt

awk -F: '{ if($1 + $2 > 100) print $0 }' awk.txt

awk '{if(NR == 4 || NR == 5) print NR "->" $0}' /etc/passwd

Shell Script :

  • Standard Shells :

    1. Bourne shell (sh)
    2. C shell (csh)
    3. korn shell (ksh)

  • Shell program is a combination of Unix commands + programming constructions and comments

  • To execute the script use [$ chmod] command to turn on the execute permission for current user

-> First line starts with-> #! /bin/bash
-> Comment -># Calculating x
-> Extension of it is [.sh]
-> $ nano day5.sh

  • To run the script use:
    1. ./days.sh
    2. source dat5.sh {this method is better}

Note: any file by default does not have execute permissions, so we have first to [$ chmod] command

chmod u+x day5.sh

Remember : echo $PATH -> Shows all Directories that stores commands of Linux, each time you enter a command then hitting enter button system check if the entered command is exist in any file in these Directories

  • To make your script global {just enter his name and the press enter button}, use this command
export PATH = $PATH:/home/arafat/bash.script

[:] as field separator of file $PATH is [:] To make this global to all terminals add above command to [~/.bashrc]

Variables:

  • Local Variables -> x=foo;echo $x

Note: x = foo is wrong {NO SPACES}

  • Environment Variables -> export y=bat; echo $y
  • Predefined variables -> Defined already in bash script

echo $? -> output of last command [0/1]
0 is for true {as ant c program return 0: when executed successfully}
1 is false

$# -> Number of argument
$* -> list all arguments
$1,$2 ,... -> first,second argument
$0 -> script name

EX: inside script.sh file

echo "script is $0"
echo "Number of arguments are $#"
echo "List arguments $*"
echo "Argument 1 = $1"

run : $ script.sh arg1 arg2 arg3

  • O/P:
script.sh
3
arg1 arg2 arg3
arg1

$ i=5                          O/P
$ echo $i                       5
$ i=$i+1; echo $i             '5+1'
$ let i=$i+2
$ echo $i                       7
$ typeset -i Num
$ Num=5:echo $Num               5
$ Num=$Num+5:echo $Num          10