sed常用命令

语法一

sed [options] {sed-commands} {input-file}

例子：

# -n表示取消默认输出,p表示打印行
[root@yyl ~]# sed -n 'p' passwd
# 只打印第三行
[root@yyl ~]# sed -n '3p' passwd
# 打印1，3行
[root@yyl ~]# sed -n '1,3p' passwd

语法二

$sed [options] -f {sed-commands-in-a-file} {input-file}

例子：

# 打印以root开头或者nobody开头的行
[root@yyl ~]# cat sed_exp1.sed
/^root/ p
/^nobody/ p
[root@yyl ~]# sed -n -f sed_exp1.sed passwd
root:x:0:0:root:/root:/bin/bash
nobody:x:99:99:Nobody:/:/sbin/nologin
[root@yyl ~]#

语法三

sed [options] -e {sed-command-1} -e {sed-command-2} {input-file}

例子：

# 打印以root开头或者nobody开头的行
[root@yyl ~]# sed -n -e '/^root/ p' -e '/^nobody/ p' passwd
root:x:0:0:root:/root:/bin/bash
nobody:x:99:99:Nobody:/:/sbin/nologin
# 或者
[root@yyl ~]# sed -n \
> -e '/^root/ p' \
> -e '/^nobody/ p' \
> passwd
root:x:0:0:root:/root:/bin/bash
nobody:x:99:99:Nobody:/:/sbin/nologin
[root@yyl ~]#

语法四

sed [options] '{
sed-command-1
sed-command-2
}' input-file

例子：

# 打印以root开头或者nobody结尾的行
[root@yyl ~]# sed -n '{
> /^root/ p
> /nobody$/ p
> }' passwd
root:x:0:0:root:/root:/bin/bash

行后增加语法

sed '[address] a the-line-to-append' input-file

例子：

[root@yyl ~]# sed '2 a 108,Donald Stufft, Nebula' example2.txt
101,Ian Bicking,Mozilla
102,Hakim El Hattab,Whim
108,Donald Stufft, Nebula
103,Paul Irish,Google
104,Addy Osmani,Google
105,Chris Wanstrath,Github
106,Mattt Thompson,Heroku
107,Ask Solem Hoel,VMware

行前插入语法

sed '[address] i the-line-to-insert' input-file

例子：

[root@yyl ~]# sed '2 i 108,Donald Stufft, Nebula' example2.txt
101,Ian Bicking,Mozilla
108,Donald Stufft, Nebula
102,Hakim El Hattab,Whim
103,Paul Irish,Google
104,Addy Osmani,Google
105,Chris Wanstrath,Github
106,Mattt Thompson,Heroku
107,Ask Solem Hoel,VMware

修改行语法

sed '[address] c the-line-to-insert' input-file

例子：修改符合Paul行为…

[root@yyl ~]# sed '/Paul/ c 108,Donald Stufft, Nebula' example2.txt
101,Ian Bicking,Mozilla
102,Hakim El Hattab,Whim
108,Donald Stufft, Nebula
104,Addy Osmani,Google
105,Chris Wanstrath,Github
106,Mattt Thompson,Heroku
107,Ask Solem Hoel,VMware

其他

1. l 会显示隐藏字符比如 '\\t制表符' $ \\r

   #    显示第一行的隐藏字符
   [root@yyl ~]# sed -n '1l' example2.txt
   \t101,Ian Bicking,Mozilla$

2. l5 会在5个字符数之后进行换行。

   [root@yyl ~]# sed -n 'l5' example2.txt
   \t10\
   1,Ia\
   n Bi\
   ckin\
   g,Mo\
   zill\

3. = 会先打印每行行号

   [root@yyl ~]# sed '=' example2.txt
   1
   	101,Ian Bicking,Mozilla
   2
   102,Hakim El Hattab,Whim

4. y或翻译你要转换的字符,这里I会转化成i，B转换成b

   [root@yyl ~]# sed 'y/IB/ib/' example2.txt | head -1
   101,ian bicking,Mozilla

sed 流

1. 读
2. 执行
3. 打印
4. 重复

测试文件 example.txt

[root@yyl ~]# cat example2.txt
101,good good study, day day up
102,Come on, chop-chop, we are running late
103,He always blah blah and I've heard it all before
104,He is a bit of a goody-goody
105,I think you might pooh-pooh this idea
106,Let's divide the prize fifty-fifty
107,In the end he was forced to resign but it was all very hush-hush

测试文件 example2.txt

101,Ian Bicking,Mozilla
102,Hakim El Hattab,Whim
103,Paul Irish,Google
104,Addy Osmani,Google
105,Chris Wanstrath,Github
106,Mattt Thompson,Heroku
107,Ask Solem Hoel,VMware

文本替换

语法：

sed '[address-range|pattern-range] s/original-string/replacement-string/[substitute-flags]' inputfile

例子

1. 替换Google为Github(默认只会替换每一行的第一个 替换全部使用 ‘s/good/bad/g’)

   [root@yyl ~]# sed 's/good/bad/' example.txt | head -1
   101,bad good study, day day up

2. 替换匹配Come的行里面的chop为chops

   [root@yyl ~]# sed '/Come/s/chop/chops/' example.txt
   101,good good study, day day up
   102,Come on, chops-chop, we are running late
   103,He always blah blah and I've heard it all before
   104,He is a bit of a goody-goody
   105,I think you might pooh-pooh this idea
   106,Let's divide the prize fifty-fifty
   107,In the end he was forced to resign but it was all very hush-hush

3. 默认只会替换每行的第一个

   [root@yyl ~]# sed '1s/g/G/' example.txt | head -1
   101,Good good study, day day up

4. g可以替换每行的全部符合

   [root@yyl ~]# sed '1s/g/G/g' example.txt | head -1
   101,Good Good study, day day up

5. 可以直接指定想要替换的第N个匹配项,这里是第二个

   [root@yyl ~]# sed '1s/g/G/2' example.txt | head -1
   101,good Good study, day day up

6. 使用w将能够替换的行重定向写到output.txt

   [root@yyl ~]# sed '1s/g/G/g w output.txt' example.txt | head -1
   101,Good Good study, day day up
   [root@yyl ~]# cat output.txt
   101,Good Good study, day day up

7. 还可以使用i忽略匹配的大小写,有的版本可能不能用

   [root@yyl ~]# sed '1s/Good/GOOD/i' example.txt | head -1
   101,GOOD good study, day day up

8. 给每行前和后都添加点字符 echo “(当前行)”;

   [root@yyl ~]# sed 's/\(.*\)/echo \"\1\";/' example.txt
   echo "101,good good study, day day up";
   echo "102,Come on, chop-chop, we are running late";
   echo "103,He always blah blah and I've heard it all before";
   echo "104,He is a bit of a goody-goody";
   echo "105,I think you might pooh-pooh this idea";
   echo "106,Let's divide the prize fifty-fifty";
   echo "107,In the end he was forced to resign but it was all very hush-hush";

9. 执行命令

   [root@yyl ~]# cat commond.txt
   /etc/passwd
   /etc/group
   [root@yyl ~]# sed 's/^/ls -l /e' commond.txt
   -rw-r--r--. 1 root root 879 5月  25 12:39 /etc/passwd
   -rw-r--r--. 1 root root 476 6月  12 23:10 /etc/group

10. sed分隔符不只可以使用’/‘ 这里使用’#’

    [root@yyl ~]# cat passwd | head -1
    root:x:0:0:root:/root:/bin/bash
    [root@yyl ~]# cat passwd | head -1 | sed 's#/bin/bash#/sbin/bash#'
    root:x:0:0:root:/root:/sbin/bash

11. 替换覆盖

    [root@yyl ~]# sed '{
    > s/good/GOOD/
    > s/GO/go/
    > }' example.txt | head -1
    101,goOD good study, day day up

12. 神奇的 & 符号

    [root@yyl ~]# sed 's/^[0-9][0-9][0-9]/[&]/g' example.txt
    [101],good good study, day day up
    [102],Come on, chop-chop, we are running late
    [103],He always blah blah and I've heard it all before
    [104],He is a bit of a goody-goody
    [105],I think you might pooh-pooh this idea
    [106],Let's divide the prize fifty-fifty
    [107],In the end he was forced to resign but it was all very hush-hush
    [root@yyl ~]# sed 's/good/&test/' example.txt | head -1
    101,goodtest good study, day day up

13. 元字符\u \l \U \L ：转换为大写/小写字符
    \U 将所有自负转换成大写
    \u 将首字符转成大写
    \L 将所有字符转成小写
    \l 将首字符转成小写

    [root@yyl ~]# sed 's/good/\U&/' example.txt | head -1
    101,GOOD good study, day day up
    [root@yyl ~]# sed 's/good/\U&/g' example.txt | head -1
    101,GOOD GOOD study, day day up
    [root@yyl ~]# sed 's/good/\u&/g' example.txt | head -1
    101,Good Good study, day day up
    [root@yyl ~]# sed 's/^.*$/\U&/' example.txt | head -1
    101,GOOD GOOD STUDY, DAY DAY UP
    [root@yyl ~]# sed 's/good/\U&/g' example.txt | head -1 | sed 's/GOOD/\L&/'
    101,good GOOD study, day day up
    [root@yyl ~]# sed 's/good/\U&/g' example.txt | head -1 | sed 's/GOOD/\l&/'
    101,gOOD GOOD study, day day up

14. \E能打断\L或者\U改变大小写

    [root@yyl ~]# sed 's/Ian/\Uian/' example2.txt|head -1
    101,IAN Bicking,Mozilla
    [root@yyl ~]# sed 's/Ian/\Uia\En/' example2.txt|head -1
    101,IAn Bicking,Mozilla

15. 使用()获取正则匹配的内容，使用\1 获取第一个括号的内容，使用\2获取第二个括号里面的内容
    例如：获取passwd文件的用户名

    [root@yyl ~]# sed 's/\(^[a-z-]\+\):.*/\1/' passwd | head -1
    root

16. 使用（）替换部分字符串

    [root@yyl ~]# echo "w1w2w3"|sed 's/\(w1\)\(w2w3\)/\2\1/'
    w2w3w1

17. ^ 表示匹配以什么开头

    [root@yyl ~]# sed -n '/^101/ p' example2.txt
    101,Ian Bicking,Mozilla

18. $表示匹配以什么结尾

    [root@yyl ~]# sed -n '/Github$/ p' example2.txt
    105,Chris Wanstrath,Github

19. .表示单个字符,下面的匹配一个逗号然后I然后2个单字符

    [root@yyl ~]# sed -n '/,I../ p' example2.txt
    101,Ian Bicking,Mozilla

20. *表示匹配0个或者多个, \+表示匹配一个或者多个, \?表示匹配0个或者1个
    [0-9]表示匹配数字,下面匹配包含3或者4的行

    [root@yyl ~]# sed -n '/att*/ p' example2.txt
    102,Hakim El Hattab,Whim
    105,Chris Wanstrath,Github
    106,Mattt Thompson,Heroku
    [root@yyl ~]# sed -n '/t\+/ p' example2.txt
    102,Hakim El Hattab,Whim
    105,Chris Wanstrath,Github
    106,Mattt Thompson,Heroku
    [root@yyl ~]# sed -n '/Ho\?e/ p' example2.txt
    106,Mattt Thompson,Heroku
    107,Ask Solem Hoel,VMware
    [root@yyl ~]# sed -n '/[34]/ p' example2.txt
    103,Paul Irish,Google
    104,Addy Osmani,Google

21. • 表示范围,这里匹配3,4,5

      [root@yyl ~]# sed -n '/[3-5]/ p' example2.txt
      103,Paul Irish,Google
      104,Addy Osmani,Google
      105,Chris Wanstrath,Github

22. | 表示或者的关系

    [root@yyl ~]# sed -n '/3\|5/ p' example2.txt
    103,Paul Irish,Google
    105,Chris Wanstrath,Github

23. {m} 表示前面的匹配的重复次数

    [root@yyl ~]# cat numbers.txt
    1
    12
    123
    1234
    12345
    123456
    [root@yyl ~]# sed -n '/^[0-9]\{3\}$/ p' numbers.txt
    123

24. {m,n } 表示匹配m-n的次数都算

    [root@yyl ~]# sed -n '/^[0-9]\{3,4\}$/ p' numbers.txt
    123
    1234

25. 删除所有注释和空行

    [root@yyl ~]# sed -e 's/#.*//' -e '/^$/ d' /etc/profile

26. 转化windows文件到unix格式(这个没看懂)

    sed 's/.$//' filename

sed 可执行脚本

[root@yyl ~]# cat testscript.sed
#!/usr/bin/sed -nf
/root/ p
/nobody/ p
[root@yyl ~]# chmod u+x testscript.sed
[root@yyl ~]# ./testscript.sed passwd
root:x:0:0:root:/root:/bin/bash
operator:x:11:0:operator:/root:/sbin/nologin
nobody:x:99:99:Nobody:/:/sbin/nologin

sed 修改源文件和备份

1. -i 会修改源文件,但是可以同时使用bak备份

   [root@yyl ~]# sed -ibak 's/Ian/IAN/' example2.txt
   [root@yyl ~]# ls -l | grep example2.txt
   -rw-r--r--.  1 root root        173 6月  13 15:12 example2.txt
   -rw-r--r--.  1 root root        173 6月  13 10:37 example2.txtbak
   [root@yyl ~]# cat example2.txt
   101,IAN Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google
   105,Chris Wanstrath,Github
   106,Mattt Thompson,Heroku
   107,Ask Solem Hoel,VMware
   [root@yyl ~]# cat example2.txtbak
   101,Ian Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google
   105,Chris Wanstrath,Github
   106,Mattt Thompson,Heroku
   107,Ask Solem Hoel,VMware

2. 1还可以使用这种语法 --in-place=后缀

   [root@yyl ~]# sed --in-place=_bak 's/IAN/Ian/' example2.txt
   [root@yyl ~]# ls -l | grep example2.txt
   -rw-r--r--.  1 root root        173 6月  13 15:16 example2.txt
   -rw-r--r--.  1 root root        173 6月  13 10:37 example2.txtbak
   -rw-r--r--.  1 root root        173 6月  13 15:12 example2.txt_bak

###

sed命令—行定位

1. 打印出文件的第三行 (定位单行)

   [root@yyl ~]# nl example2.txt | sed -n '3p'
        3	103,Paul Irish,Google

2. 打印出文件的第2到第3行 （多行定位）

   [root@yyl ~]# nl example2.txt | sed -n '2,3p'
        2	102,Hakim El Hattab,Whim
        3	103,Paul Irish,Google

3. 打印出包含Paul的行 （单行定位）

   [root@yyl ~]# nl example2.txt | sed -n '/Paul/p'
        3	103,Paul Irish,Google

4. 打印出包含Paul开始的行到包含Chris结束的行 （多行定位）

   [root@yyl ~]# nl example2.txt | sed -n '/Paul/,/Chris/p'
        3	103,Paul Irish,Google
        4	104,Addy Osmani,Google
        5	105,Chris Wanstrath,Github

5. 从Paul开始的行到Chris结束的行，都不要

   [root@yyl ~]# nl example2.txt | sed -n '/Paul/,/Chris/!p'
        1	101,Ian Bicking,Mozilla
        2	102,Hakim El Hattab,Whim
        6	106,Mattt Thompson,Heroku
        7	107,Ask Solem Hoel,VMware

6. 打印出文件非第5行的内容 (单行取反)

   [root@yyl ~]# nl example2.txt | sed -n '5!p'
        1	101,Ian Bicking,Mozilla
        2	102,Hakim El Hattab,Whim
        3	103,Paul Irish,Google
        4	104,Addy Osmani,Google
        6	106,Mattt Thompson,Heroku
        7	107,Ask Solem Hoel,VMware

7. 打印出文件的非第2到五行内容 （多行取反）

   [root@yyl ~]# nl example2.txt | sed -n '2,5!p'
        1	101,Ian Bicking,Mozilla
        6	106,Mattt Thompson,Heroku
        7	107,Ask Solem Hoel,VMware

8. 步进获取行的内容（从第2行开始 步长为3）

   [root@yyl ~]# nl example2.txt | sed -n '2~3p'
        2	102,Hakim El Hattab,Whim
        5	105,Chris Wanstrath,Github
   结果为2,5,8,11......行

9. 从找到开始打印到第五行(Paul 在第3行)

   [root@yyl ~]# nl example2.txt | sed -n '/Paul/,5p'
        3	103,Paul Irish,Google
        4	104,Addy Osmani,Google
        5	105,Chris Wanstrath,Github

10. 从找到的地方开始 在多输出3行

    [root@yyl ~]# nl example2.txt | sed -n '/Paul/,+3p'
         3	103,Paul Irish,Google
         4	104,Addy Osmani,Google
         5	105,Chris Wanstrath,Github
         6	106,Mattt Thompson,Heroku

删除行

1. 删除所有行

   [root@yyl ~]# nl example2.txt | sed 'd'

2. 只删除第二行

   [root@yyl ~]# nl example2.txt | sed '2d'
        1	101,Ian Bicking,Mozilla
        3	103,Paul Irish,Google
        4	104,Addy Osmani,Google
        5	105,Chris Wanstrath,Github
        6	106,Mattt Thompson,Heroku
        7	107,Ask Solem Hoel,VMware

3. 删除第1到4行

   [root@yyl ~]# nl example2.txt | sed '1,4d'
        5	105,Chris Wanstrath,Github
        6	106,Mattt Thompson,Heroku
        7	107,Ask Solem Hoel,VMware

4. 按步长删除（删除奇数行）

   [root@yyl ~]# nl example2.txt | sed '1~2d'
        2	102,Hakim El Hattab,Whim
        4	104,Addy Osmani,Google
        6	106,Mattt Thompson,Heroku

5. 删除匹配的行(同样支持 ‘/Paul/,4d’ | ‘/Paul/,+3d’ | ‘/Paul/,/Chris/d’ | ‘/Paul/,/Chris/!d’)

   [root@yyl ~]# nl example2.txt | sed '/Paul/d'
        1	101,Ian Bicking,Mozilla
        2	102,Hakim El Hattab,Whim
        4	104,Addy Osmani,Google
        5	105,Chris Wanstrath,Github
        6	106,Mattt Thompson,Heroku
        7	107,Ask Solem Hoel,VMware

6. 删除空行 注释行(这块没有用nl命令 nl命令会影响 sed执行效果)

   [root@yyl ~]# cat example2.txt
   101,Ian Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google
   105,Chris Wanstrath,Github
   #1234567777
   
   106,Mattt Thompson,Heroku
   107,Ask Solem Hoel,VMware
   
   [root@yyl ~]# sed '/^#/d' example2.txt // 删除注释行
   101,Ian Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google
   105,Chris Wanstrath,Github
   
   106,Mattt Thompson,Heroku
   107,Ask Solem Hoel,VMware
   
   [root@yyl ~]# sed '/^$/d' example2.txt  // 删除空行
   101,Ian Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google
   105,Chris Wanstrath,Github
   #1234567777
   106,Mattt Thompson,Heroku
   107,Ask Solem Hoel,VMware

重定向

1. 将example.txt内容重定向写到output.txt

   [root@yyl ~]# ls -l | grep txt
   -rw-r--r--.  1 root root        186 6月  13 09:55 example.txt
   [root@yyl ~]# sed 'w output.txt' example2.txt
   101,Ian Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google
   105,Chris Wanstrath,Github
   #1234567777
   
   106,Mattt Thompson,Heroku
   107,Ask Solem Hoel,VMware
   [root@yyl ~]# ls -l | grep txt
   -rw-r--r--.  1 root root        186 6月  13 09:55 example2.txt
   -rw-r--r--.  1 root root        186 6月  13 09:55 output.txt

2. 将example.txt内容重定向写到output1.txt(不会在终端输出)

   [root@yyl ~]# sed -n 'w output1.txt' example2.txt
   [root@yyl ~]# ls -l | grep txt
   -rw-r--r--.  1 root root        186 6月  13 09:55 example2.txt
   -rw-r--r--.  1 root root        186 6月  13 09:56 output1.txt
   -rw-r--r--.  1 root root        186 6月  13 09:55 output.txt

3. 只写第二行

   [root@yyl ~]# sed -n '2w output.txt' example2.txt
   [root@yyl ~]# cat output.txt
   102,Hakim El Hattab,Whim

4. 将第一到四行写到output.txt

   [root@yyl ~]# sed -n '1,4w output.txt' example2.txt
   [root@yyl ~]# cat output.txt
   101,Ian Bicking,Mozilla
   102,Hakim El Hattab,Whim
   103,Paul Irish,Google
   104,Addy Osmani,Google

5. 将匹配Mat的行到结尾行写到output.txt(正则匹配区分大小写)
   同样支持

   '1,4w output.txt'
   '/Paul/,5w output.txt'
   '/Paul/,+1w output.txt' 
   '/Paul/,/Chris/w output.txt'
   '/Paul/,/Chris/!w output.txt'
   

[root@yyl ~]# sed -n '/mat/,$w output.txt' example2.txt
[root@yyl ~]# cat output.txt
[root@yyl ~]# sed -n '/Mat/,$w output.txt' example2.txt
[root@yyl ~]# cat output.txt
106,Mattt Thompson,Heroku
107,Ask Solem Hoel,VMware