• The goal is
• Have a good command of shell Definition and acquisition of variables
• Be able to carry on shell Four simple operations

Disk related commands Review :

Programming language classification introduction

Computers can only recognize machine language （ Such as :01010101001 such ）, Programmers can't write directly 01 This code , Therefore, the program language written by the programmer should be translated into machine language . A tool for translating other languages into machine languages , be called ** Compiler or interpreter **.

Such as ： Chinese —（ translate ）---- foreigners

There are two ways of compiler translation , One is compile , One is explain . The difference between the following :

• Compiler language ：

​ The program needs a special compilation process before execution , Compiling programs into machine language files , No retranslation is required at runtime , Just use the results of the compilation . High efficiency of program execution , Dependent compiler , Poor cross platform performance . Such as C、C++

• Explanatory language ：

​ The program does not need to be compiled , The program is run by ** Interpreter ** Translate into machine language , Translate every time you execute . So it's less efficient . such as Python/JavaScript/ Perl /ruby/Shell And so on are all explanatory languages .

• summary ：

Compiled languages are better than interpreted languages Faster , But it's not as good as explanatory language Good cross platform . If you do the underlying development or large-scale application or operating system development i They use compiled languages ; If it's some server scripts and some auxiliary interfaces , The speed requirement is not high 、 For each platform Compatibility is required In general, I use Explanatory language .

shell Introduce

summary ：

• shell It's a bridge of human-computer interaction
• shell The type of

# cat /etc/shells 
/bin/sh			 yes bash shell A shortcut to 
/bin/bash		bash shell It's most Linux default shell, The included functions can almost cover shell All functions 
/sbin/nologin	 Means non interactive , Can't log in to the operating system 
/bin/dash		 small , Efficient , Less features 

• /bin/bash or /bin/sh Is the most commonly used shell The interpreter of the program

What is? shell Script ？

The programming of interpretive languages is generally called script Programming , That is what script means .shell script Namely shell Script .

Several orders + The basic format + Specific grammar + thought = shell Script

When to use scripts ？

Repetition 、 Complicated work , By scripting work orders , In the future, you only need to execute scripts to complete these tasks .

① Automated analytical processing

② Automated backup

③ Automate batch deployment installation

④ wait …

How to learn shell Script ？

1. Be familiar with as many commands as possible

2. Master the standard format of script （ Declaration type 、 Run the script using the standard execution method ）

3. Be familiar with the basic syntax of scripts

4. See more （ understand ）——> Imitate more （ More practice ）——> Think more

shell Script

first shell Script

[[email protected] ~]# mkdir /shell/shell01 -p
[[email protected] ~]# vim /shell/shell01/1.sh

#!/bin/bash
 Interpretive language declaration type , Express /bin/bash For the interpreter ( use /bin/sh It's fine too )

echo "hello world"

Script execution method ：

Method 1: （ The script does not need execution permission , Declaring types doesn't work , Because it is directly designated sh To execute this script )

[[email protected] ~]# cd /shell/shell01/
[[email protected] shell01]# sh 1.sh
hello world

 Add :sh -x 1.sh Executing the script in this way has certain debugging function , You can view the script execution process 

Method 2:（ The script does not need execution permission , Declaring types doesn't work , Because it is directly designated bash To execute this script )

[[email protected] shell01]# bash 1.sh
hello world

Method 3: （ The script needs execution permission , Declaration types work )

[[email protected] shell01]# ./1.sh
-bash: ./1.sh: Permission denied
[[email protected] shell01]# chmod a+x 1.sh
[[email protected] shell01]# ./1.sh
hello world

Pure command shell Script

That is, you don't need to learn any other programming syntax （ But you need to be proficient in commands ), Arrange the commands in order from top to bottom . When executing, it will be executed from top to bottom .

Example : Print 1,2,3

#!/bin/bash

echo 1
echo 2
echo 3

Example : Delete other yum The configuration file , Configure local yum( Suppose the image is mounted to /mnt Catalog )

#!/bin/bash

rm /etc/yum.repos.d/* -rf
cat > /etc/yum.repos.d/local.repo <<EOF
[local]
name=local
baseurl=file:///mnt
enabled=1
gpgcheck=0
EOF

review

# Notice the following two differences 
cat > /tmp/abc <<EOF				# > Symbol , It means to cover the original content 
 Hello 
EOF

cat >> /tmp/abc <<EOF				# >> Symbol , It means adding 
 ha-ha 
EOF

key word : Programming code translate ( Interpreter /bin/bash or /bin/sh) Write multiple commands into the program in sequence , It can help us execute in batch at one time , Increase of efficiency .

practice : use shell Realization httpd Access logs rotate every day (/var/log/httpd/access_log)

requirement :

1. take /var/log/httpd/access_log Rotate to /backup/ year / month / year - month - date .access_log
2. Create a new one after rotation /var/log/httpd/access_log
3. After each log rotation , All in /var/log/message Record a message that the log has been rotated ( Information customization )
4. Put the written script into crond In the service

preparation :

# yum install httpd httpd-devel elinks -y
# /etc/init.d/httpd restart
# elinks -dump 127.0.0.1

# cat /var/log/httpd/access_log
127.0.0.1 - - [12/Apr/2019:14:55:15 +0800] "GET / HTTP/1.1" 403 4961 "-" "ELinks/0.12pre5 (textmode; Linux; -)"

#!/bin/bash
mkdir /backup/$(date +%Y)/$(date +%m)/   -p

mv /var/log/httpd/access_log /backup/$(date +%Y)/$(date +%m)/$(date +%F).access_log
touch /var/log/httpd/access_log
/etc/init.d/httpd reload  &> /dev/null
logger -t " log rotation " "$(date +%F) success "

59 23 * * * sh /path/xxx.sh

Variable

What is a variable ?

Popular said , Variables are used to temporarily store data .( notes : learn python Variables will be discussed more deeply )

Format of variable definition

Variable name = A variable's value

Want to get the value of the variable , Add... Before the variable $ Symbol

# a=1
# echo $a
1

When do I need to define variables ？

• If a content needs to be used more than once , And in the code Recurring , Then we can use variables to represent the content . In this way, when modifying the content , You just need to change the value of the variable .
• In the process of code operation , The execution result of some commands may be saved , Subsequent code needs to use these results , You can use this variable directly .

#!/bin/bash

year=$(date +%Y)
month=$(date +%m)

mkdir /backup/$year/$month/   -p

mv /var/log/httpd/access_log /backup/$year/$month/$(date +%F).access_log
touch /var/log/httpd/access_log
/etc/init.d/httpd reload  &> /dev/null
logger -t " log rotation " "$(date +%F) success "

Definition rules and methods of variables ( a key )

Basic rules of variable definition

1. Case sensitive , Variable names with the same name but different case are different variables

# a=2
# A=3
# echo $a
2
# echo $A
3

2. Pay attention to the format when defining , There can be no spaces on either side of the equal sign , When assigning a value to a string with spaces , Put it in quotation marks

 B="hello world"
 B='hello world haha'	
  Single quotation marks and double quotation marks are OK here , The latter assignment will overwrite the previous assignment 

3. The difference between single quotation marks and double quotation marks : Variables or special characters in single quotation marks are only general characters , But variables or special characters in double quotation marks can maintain their variable characteristics

# echo '$B'
$B
# echo "$B"
hello world haha

4. The variable name can be a combination of letters or numbers or underscores , But it can't start with a number

# c123=aaa
# echo $c123
aaa											 correct 

# 123c=aaa
bash: 123c=aaa: command not found			 error 

# _abc11122=aaa
# echo $_abc11122
aaa											 correct 

5. Variable names should be known as well as possible ( Don't let all the variables in a script be a,b,c,d And so on. , Not easy to read )

Basic method of variable definition

1. Assign values directly to variables

a=1

2. The execution result of the command can be assigned to the variable as the value of the variable

 You can use multiple times in a command $(), But use it many times `` The execution symbol will report an error 
# a=`rpm -qf `which mount``  Report errors 

# a=$(rpm -qf `which mount`)  correct 
# a=$(rpm -qf $(which mount))  correct 

3. adopt read Define variables interactively

# read -p " Type in your name :" name
 Type in your name : Zhang San 
# echo $name
 Zhang San 

Example :

#!/bin/bash

read -p " Type in your name :" name

echo " Hello ,$name"

read -s -p " Enter your mobile number :" num
echo											#  direct echo Means line break 
echo " Your mobile number is $num" 

read -n 2 -p " Enter your age :" age
echo
echo " you $age Year old "

read -t 3 -p "214546+23446＋46645-126456=?" result
echo
echo " Mental arithmetic is not fast enough "

Variable acquisition

adopt $ Variable name or ${ Variable name } To get the value of a variable

# num=12345
# echo $num
12345
# echo ${num}
12345

What's the difference between enlarging brackets and not enlarging brackets ?

answer : Enlarging parentheses can realize interception ( section ) Or other operations .

 The first 1 The number represents the starting position ,0 On behalf of the 1 position ,1 On behalf of the second , And so on 
 The first 2 The number represents how many digits are intercepted 
# echo ${num:0:3}
123
# echo ${num:1:2} 
23
# echo ${num:2:2}
34
# echo ${num:2:3}
345

Example :

read -s -p " Enter your mobile number :" num
echo										
echo " Your mobile number has a mantissa of ${num:7:4}" 

After class development ( Understanding can , Don't remember , Learn later awk,sed There's a better way )

file=/dir1/dir2/dir3/my.file.txt
 We can use  ${
     }  Replace them separately to get different values ：
${
    file#*/}： Take off the first one  /  And the string to the left of it ：dir1/dir2/dir3/my.file.txt
${
    file#*1/}： Remove the second  /  And the string to the left of it ：dir2/dir3/my.file.txt
${
    file##*/}： Take off the last one  /  And the string to the left of it ：my.file.txt
${
    file#*.}： Take off the first one  .  And the string to the left of it ：file.txt
${
    file##*.}： Take off the last one  .  And the string to the left of it ：txt
${
    file%/*}： Take off the last one  /  And the string to the right of it ：/dir1/dir2/dir3
${
    file%%/*}： Take off the first one  /  And the string to the right of it ：( Null value )
${
    file%.*}： Take off the last one  .  And the string to the right of it ：/dir1/dir2/dir3/my.file
${
    file%%.*}： Take off the first one  .  And the string to the right of it ：/dir1/dir2/dir3/my

 We can also replace the string in the variable value ：
${
    file/dir/path}： The first one. dir Replace with  path：/path1/dir2/dir3/my.file.txt
${
    file//dir/path}： Will all dir Replace with  path：/path1/path2/path3/my.file.txt

The elimination of variables

Use unset Variable name You can cancel the variable ( Be careful ： After a script is executed , Variable auto release , Not every variable unset Cancel )

# aaa=haha
# echo $aaa
haha
# unset aaa
# echo $aaa

linux Classification of variables

• Local temporary variables ： Variables temporarily customized by the current user . Effective in the current process , Invalid in other processes and child processes .
• environment variable ： The current process works , And can be called by child process .
  • View the current user's environment variables env
  • Query all variables of the current user ( Local temporary variables and environment variables ) set
  • export The command can change the current local temporary variable into an environment variable

Example : Local temporary variables

# bbb=haha

# env |grep bbb env I can't find it in the library , Because it's not an environment variable 
# set |grep bbb set We can find out 
bbb=haha

# echo $bbb 
haha
# bash  Intron bash( Subprocesses )
# echo $bbb  In the child bash( Subprocesses ) Invalid in 

# exit  Quitters bash( Subprocesses )
exit
# echo $bbb
haha

Example : environment variable

# export ccc=haha

# env |grep ccc env We can find out 
ccc=haha
# set |grep ccc
ccc=haha

# echo $ccc
haha
# bash  Intron bash( Subprocesses )
# echo $ccc  In the child bash( Subprocesses ) Effective in 
haha
# exit
exit

• Global environment variable ： All users and programs can call , And inherit , New users can also call by default . Such as $HOME,$PATH,$USER etc.

Example : Custom global environment variables

# vim /etc/profile
export abc=haha					 Add this sentence in the blank space at the end of the document , And save exit 

# source /etc/profile
 perhaps 
# . /etc/profile

 Use source or . Let it take effect , Any terminal and sub terminal of the operating system bash Can call this variable ( Other terminals that have been opened need to log in again )

• System variables ( built-in bash Medium variable )： shell Its name and function have been fixed , Can be directly in shell Call in script .

Example :

[[email protected] shell01]# vim 5.sh
#!/bin/bash

echo '$0' yes "$0"
echo '$1' yes "$1"
echo '$2' yes "$2"
echo '$3' yes "$3"
echo '$$' yes "$$"
echo '$#' yes "$#"
echo '$*' yes "$*"
echo '[email protected]' yes "[email protected]"


[[email protected] shell01]# sh 5.sh a b c
$0 yes 5.sh
$1 yes a
$2 yes b
$3 yes c
$$ yes 2183
$# yes 3
$* yes a b c
[email protected] yes a b c

Four simple operations

Arithmetic operations ： By default ,shell We can only support simple Integers operation

Bash shell The way of arithmetic operation :

1. Use $[ ] ( It is recommended to use this type mainly , Others can be understood )
2. Use $(( ))
3. Use expr External programs
4. Use let command
5. With the help of bc command

#!/bin/bash

a=1
let a++                         #  Use let Perform an operation , There is no need to add $
b=$[$a-1]
c=$(($a**3))
d=`expr $a + 3`                 # + There should be a space on both sides of the sign 

echo $a
echo $b
echo $c
echo $d

With the help of bc Command to realize decimal operation

# echo 1.1+2.2 | bc
3.3

# echo "sqrt(100)" | bc 
10							 take a square root 

Example : When an ordinary user logs out , Show how long you have logged in

Such as abc When the user logs out , It will automatically display " Hello ,abc, You have logged in xx when xx branch xx second ,bye bye"

# vim /home/abc/.bash_profile 
export logintime=$(date +%s)

# vim /home/abc/.bash_logout

clear
logouttime=$(date +%s)
alltime=$[$logouttime-$logintime]
hours=$[$alltime/3600]
minutes=$[$alltime%3600/60]
seconds=$[$alltime%60]
echo " Hello ,$USER, As soon as you log in $hours when $minutes branch $seconds second ,bye bye"

The order added

wc

Statistical orders

# echo 1234 |wc -c
5
# echo 1234 |wc -L
4
# cat /etc/passwd |wc -l
 Count the number of users 

cut

Intercept order , characteristic : Only a single character can be used as a separator . Later on awk and sed In many aspects of interception, it is better than cut Have more advantages .

# head -1 /etc/passwd | cut -d: -f7
/bin/bash
# head -1 /etc/passwd | cut -d: -f2
x

# echo 12345 |cut -c2-4
234
# echo 12345 |cut -c2-
2345

sort

Sort order

# cat /etc/passwd |sort -t: -k3 -n

# cat /etc/passwd |sort -t: -k3 -n -r

uniq

To repeat the order

# cut -d: -f7 /etc/passwd |sort |uniq -c

grep

Line find command ( Later, regular expressions will focus on )

 Show the line number of the original document plus n Parameters 	grep -n root /etc/passwd		
 Reverse lookup plus v Parameters   grep -v bin /etc/passwd
 Case insensitive plus i Parameters   grep -ni root grep.txt 

 Find yes root and ftp Keyword line 
# grep root /etc/passwd | grep ftp

 Find yes root or ftp Keyword line 
# egrep "root|ftp" /etc/passwd
# grep -E "root|ftp" /etc/passwd


 Statistics root stay /etc/passwd There are several times in 
grep -o root /etc/passwd |wc -l


[[email protected] shell03]# cat 1.txt
111
222
333
444
555
666
[[email protected] shell03]# cat 2.txt
aaa
bbb
333
ccc
ddd
555

 Method 1 ：
# grep -f 1.txt 2.txt -- Find the duplicate lines in the two files 
333
555

 Method 2 :
# sort 1.txt 2.txt -o 3.txt
cat 3.txt | uniq -c |awk '$1>1 {print $2}'

 Method 3 :
 Put a line in a file , Take it out line by line , Compare circularly in another file , If there is , It means that both files have .



# ifconfig |grep -A 1 vmnet 
vmnet1    Link encap:Ethernet  HWaddr 00:50:56:C0:00:01  
          inet addr:1.1.1.1  Bcast:1.1.1.255  Mask:255.255.255.0
--
vmnet8    Link encap:Ethernet  HWaddr 00:50:56:C0:00:08  
          inet addr:192.168.56.1  Bcast:192.168.56.255  Mask:255.255.255.0

Exercises :

 Operate on the following files 
# cat 1.txt
http://a.domain.com/1.html
http://b.domain.com/1.html
http://c.domain.com/1.html
http://a.domain.com/2.html
http://a.domain.com/3.html
http://b.domain.com/3.html
http://c.domain.com/2.html
http://c.domain.com/3.html
http://a.domain.com/1.html
 Here are the results 

	4 a.domain.com
	3 c.domain.com
	2 b.domain.com