sed The format of the command is

sed [options] 'command' file(s)  #  With command Instructions to process file(s)
sed [options] -f script file(s)  #  With script Script to handle file(s)

among ,options Mainly used in the following ：

-i： Modify the original document directly , Instead of output to the screen

sed -i 's/test/TEST' file01

-n： namely --quiet, Only the processed results are displayed , Not the whole content of the file
And p Instructions can be used together to print only the lines where the replacement occurs .

sed -n 's/test/TEST/p' file01

-e ： namely --expression, Allow multiple commands to be executed on the same line

sed -e '1,5d' -e 's/test/check/' file01

delimiter

sed In general use / As a delimiter , You can also use other symbols . When the delimiter appears in the string to match , You need to use a backslash to escape .

#  take test Replace all with TEST
sed 's/test/TEST/g' file01
sed 's#test#TEST#g' file01
sed 's:test:TEST:g' file01
sed 's|test|TEST|g' file01
#  take /bin Replace all with /usr/bin
sed 's/\/bin/\/usr\/bin/g' file01
sed 's#/bin#/usr/bin#g' file01

Print ：p

And -n Use it together , Print only rows that match

#  Print match to test The line of 
sed -n '/test/p' file01
#  Print page 99 Everything after the line 
sed -n '100,$p' file02

Read ：r

Read file02 The content in , Add to file01 Match to test After each line of the string

sed '/test/r file02' file01

write in ：w

take file01 Match to test All lines of the string , Write to file02 in （ Will overwrite the original contents of the file ）

sed '/test/w file02' file01

Delete ：d

#  Delete blank lines 
sed '/^$/d' file01
#  Delete the first 2 That's ok 
sed '2d' file01
#  Delete the first 2 Line to all lines at the end of the file 
sed '2,$d' file01
#  Delete last line 
sed '$d' file01
#  Delete test All the lines at the beginning 
sed '/^test/d' file01

Replace ：s

Replace Every line Match to first character string ：

sed 's/test/TEST/' file01

Replace Every line Match to all character string ：

sed 's/test/TEST/g' file01

matching

Common matching rules include the following ：

^： Match line begins , Such as ：/^sed/ Match all to sed Beginning line .
$： End of match line , Such as ：/sed$/ Match all to sed The line at the end .
.： matching One Any character that is not a newline character , Such as ：/s.d/ matching s Followed by an arbitrary character , And finally d.
*： matching 0 One or more character , Such as ：/*sed/ Matching all templates is one or more spaces followed by sed The line of .
\+： matching 1 One or more character , Such as ：/se\+d/ matching s and d At least one letter in the middle e String .
[]： Match a specified range of characters , Such as /[sS]ed/ matching sed and Sed.
[^]： Match a character that is not in the specified range , Such as ：/[^A-RT-Z]ed/ Match does not contain A-R and T-Z The beginning of a letter , Following the ed The line of .
\(..\)： Match the strings , Save matching characters , Such as s/\(love\)able/\1rs,loveable Replaced with lovers.
\1： Substring match mark , And \(..\) In combination with .\1 Represents the first substring matched , And so on ,\2 Represents the second substring that matches ,….
&： Save search characters to replace other characters , Such as s/love/ **&** /,love That's it **love**.
\<： Match the beginning of the word , Such as :/\<love/ The match consists of love The line at the beginning of the word .
\>： Match the end of the word , Such as /love\>/ The match consists of love The lines of the ending words .
x\{m\}： Repeat characters x,m Time , Such as ：/0\{5\}/ Matching inclusion 5 individual 0 The line of .
x\{m,\}： Repeat characters x, At least m Time , Such as ：/0\{5,\}/ Match at least 5 individual 0 The line of .
x\{m,n\}： Repeat characters x, At least m Time , Not more than n Time , Such as ：/0\{5,10\}/ matching 5~10 individual 0 The line of .
[:space:]： Match single whitespace , Including Spaces 、 Box drawings, etc .

Example ：
A simple example

# \w\+ Match every word ,"&" Put double quotation marks on the matched string 
echo this is a test line | sed 's/\w\+/"&"/g'
"this" "is" "a" "test" "line"

#  Match two English letter strings , Then change the position. 
echo aaa BBB | sed 's/\([a-z]\+\) \([A-Z]\+\)/\2 \1/'
BBB aaa
# \([a-z]\+\) Match to aaa, Marked as \1
# \([A-Z]\+\) Match to BBB, Marked as \2

A complex example

cat file03
# logfile /var/log/nscd.log
# threads 4
     enable-cache      group               yes
     enable-cache      hosts               yes

#  Remove the comment on the first line （ The delimiter is /）
sed -i 's/^#\([[:space:]]*logfile[[:space:]]*\/var\/log\/nscd.log\)/\1/g' file03  
#  Remove the comment on the first line （ The delimiter is |）
sed -i 's|^#\([[:space:]]*logfile[[:space:]]*/var/log/nscd.log\)|\1|g' file03 
#  among ^# Said to # Beginning line 
# [[:space:]]* Represents a match 0 One or more blanks 
# \(...\) Indicates that the matching substring is reserved ,\1 Represents the first string matched 

cat file03
     logfile           /var/log/nscd.log
# threads 4
     enable-cache      group               yes
     enable-cache      hosts               yes
     
#  Put the third line of yes Replace with no
sed -i 's/\([[:space:]]*enable-cache[[:space:]]*group[[:space:]]*\)yes/\1no/g' file03

cat file03
     logfile           /var/log/nscd.log
# threads 4
     enable-cache      group               no
     enable-cache      hosts               yes

References:
【1】https://g-ghy.github.io/linux-command/c/sed.html
【2】https://www.twle.cn/c/yufei/sed/sed-basic-regular-expressions.html