【Shell】sed编辑器实例

sed是用来解析和转换文本的工具，它使用简单，是简洁的程序设计语言。

sed编辑器

（一） sed编辑器基础
- 1. 简介
- 2. sed的模式空间
（二）基本的sed编辑命令
（三）sed命令实例
- 1. 向文件中添加或插入行
- 2. 更改文件中的指定的行
- 3. 删除文件中的行
- 4. 替换文件中内容
- 5. 打印文件中的行

（一） sed编辑器基础

1. 简介

自动化地编辑一个或多个文件
简化在多个文件中执行相同编辑地任务
编写转换程序

2. sed的模式空间

sed维护一种模式空间，即一个工作区或临时缓冲区，当使用编辑命令时，将在那存储单个输入行。下图展示了进行模式空间转换的一个两行的sed脚本。它将“The linux system”转换成“The LINUX Operating System”

在这里插入图片描述
sed一次处理一行输入的优点是在读取非常庞大的文件时不会出现问题，一般的文本编辑器必须将整个文件读入内存，这将会产生内存溢出或者在处理庞大的文件时速度非常慢。

（二）基本的sed编辑命令

a ：新增， a 的后面可以接字串，而这些字串会在新的一行出现(目前的下一行)～
c ：取代， c 的后面可以接字串，这些字串可以取代 n1,n2 之间的行！
d ：删除，因为是删除啊，所以 d 后面通常不接任何东东；
i ：插入， i 的后面可以接字串，而这些字串会在新的一行出现(目前的上一行)；
p ：打印，亦即将某个选择的数据印出。通常 p 会与参数 sed -n 一起运行～
s ：取代，可以直接进行取代的工作哩！通常这个 s 的动作可以搭配正则表达式！

（三）sed命令实例

1. 向文件中添加或插入行

实例1：在文件的指定行之后添加一行内容。

$ cat info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available

使用sed的追加编辑命令（a）在上述文件的第5行添加一行命令：

[ley@ley-web ~]$ sed '5a\
> Solaris - Sysadmin, Recovery etc.' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Solaris - Sysadmin, Recovery etc.
Productivity - Too many technologies to explore, no much time available

实例2：在匹配模式的行之后添加一行内容

$ sed '/Databases/a\
> Solaris - Sysadmin, Recovery etc.' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available

实例3：在文本的最后一行后添加多行内容
在info.txt文件后面添加两行

$ sed '$a\
>  Solaris - Sysadmin, Recovery etc.\
> Windows - Sysadmin etc.' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available
 Solaris - Sysadmin, Recovery etc.
Windows - Sysadmin etc.

实例4：在文件中的指定行之前插入一行内容。

$ sed '3i\
> Solaris - Sysadmin, Recovery etc.' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Solaris - Sysadmin, Recovery etc.
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available

实例5：在匹配指定模式的行之前插入一行内容。

在文件info.txt中匹配/Security/的行之前插入两行内容。

$ sed '/Security/i\
> Solaris - Sysadmin, Recovery etc.\
> Windows - Sysadmin etc.' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Solaris - Sysadmin, Recovery etc.
Windows - Sysadmin etc.
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available

实例6：在文件的最后一行之前插入一行内容

$ sed '$i\
> Solaris - Sysadmin, Recovery etc.' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Solaris - Sysadmin, Recovery etc.
Productivity - Too many technologies to explore, no much time available

2. 更改文件中的指定的行

实例1：修改文件的第一行

$ sed '1c<Change line>' info.txt
<Change line>
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available

实例2：修改匹配指定模式的行
匹配/Cool/的行修改为

$ sed '/Cool/c<Change line>' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
<Change line>
Storage - NetApp, EMC etc
Productivity - Too many technologies to explore, no much time available

实例3：更改文件的最后一行。

$ sed '$c<Change line>' info.txt
Linux - Sysadmin
Databased - Oracle, MySQL etc
Security - Firewall, Network, Oline Security etc
Cool - Websities
Storage - NetApp, EMC etc
<Change line>

3. 删除文件中的行

实例1：删除文件中的指定的行
删除文件所在的第四行

$ sed '4d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc
5       Storage - NetApp, EMC etc
6       Productivity - Too many technologies to explore, no much time available

实例2：从指定的行开始删除，并每个固定行删除一行

从第四行开始删除，并每隔两行就删掉一行

$ sed '4~2d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc
5       Storage - NetApp, EMC etc

实例3：删除指定范围的行

$ sed '3,6d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc

实例4：删除指定范围以外的行
删除3到6行以外的行

$ sed '3,6!d' info.txt
3       Security - Firewall, Network, Oline Security etc
4       Cool - Websities
5       Storage - NetApp, EMC etc
6       Productivity - Too many technologies to explore, no much time available

实例5：删除文件中的最后一行

$ sed '$d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc
4       Cool - Websities
5       Storage - NetApp, EMC etc

实例6：删除文件中匹配指定模式的行。

匹配/Productivity/的行并删除

$ sed '/Productivity/d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc
4       Cool - Websities
5       Storage - NetApp, EMC etc

实例7：从匹配指定模式的行删到文件的最后一行

$ sed '/Productivity/,$d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc
4       Cool - Websities
5       Storage - NetApp, EMC etc

实例8：删除文件中匹配指定模式的行，及其后面的n行内容
删除info.txt文件中的/Security/，及其后面的一行

$ sed '/Security/,+1d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
5       Storage - NetApp, EMC etc
6       Productivity - Too many technologies to explore, no much time available

实例9：删除文件中的空行

$ sed '/^$/d' info.txt
1       Linux - Sysadmin
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc
4       Cool - Websities
5       Storage - NetApp, EMC etc
6       Productivity - Too many technologies to explore, no much time available

实例10：删除文件中不匹配指定模式的行

删除文件中不匹配/Databased/或/Security/的行

$ sed '/Databased\|Security/!d' info.txt
2       Databased - Oracle, MySQL etc
3       Security - Firewall, Network, Oline Security etc

实例11：删除文件的指定范围内的行中匹配指定模式的行。

删除含有etc的行

$ sed '1,4{/etc/d}' info.txt
1       Linux - Sysadmin
4       Cool - Websities
5       Storage - NetApp, EMC etc
6       Productivity - Too many technologies to explore, no much time available

4. 替换文件中内容

$ cat techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows

实例1：替换一行中第一个匹配模式的字符串

将每一行的第一个/Virtualization替换成/Virt

$ sed 's/Virtualization/Virt./' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virt.:Server Virtualization,Storage Virtualization,     Desktop Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows
# Additional class

实例2：替换文件中的匹配指定模式的所有字符串

将每一行的所有/Virtualization替换成/Virt

$ sed 's/Virtualization/Virt./g' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virt.:Server Virt.,Storage Virt.,       Desktop Virt.
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows
# Additional class

实例3：替换文件中每行第n个匹配指定模式的字符串

将每一行的第三个/Virtualization替换成/Virt

$ sed 's/Virtualization/Virt./3' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virt.,     Desktop Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows
# Additional class

实例4：将发生字符串替换的行写入指定的文件，并只打印发生替换的行

将文件中的“Network”字符串替换为“Net.”

$ sed -n 's/Network/Net./gpw /tmp/sedOutput' techClass.txt
1.      Net.: Route, Switch, Wireless,Communicate, Device
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Net.

实例5：只替换文件中匹配指定模式的行中的字符串

文件中匹配/:/的行中的逗号“,”之后的字符串清空

$ sed '/:/s/,.*//g' techClass.txt
1.      Network: Route
2.      Security: Data Protection
3.      Server: Blade
4.      Virtualization:Server Virtualization
5.      Database: SQLServer
6.      OS: Linux
# Additional class

实例6：删掉每行的最后n个字符

删掉文件中每行最后两个字符

$ sed 's/..$//g' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Devic
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Securit
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Networ
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop Virtualizati
5.      Database: SQLServer,MySQL,Oracle, DB
6.      OS: Linux, Unix, Window
# Additional cla

实例7：删除文件中的注释

$ sed 's/^#.*//' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows

实例8：删除文件中的注释及其空行

$ sed 's/^#.*//;/^$/d' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows

实例9：使用符号“&”获得匹配的字符串

每一行数字编号加一个小括号（）

$ sed 's/^[0-9]\./(&)/' techClass.txt
(1.)    Network: Route, Switch, Wireless,Communicate, Device
(2.)    Security: Data Protection,Terminal Security,C1oud Security,WEB Security
(3.)    Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
(4.)    Virtualization:Server Virtualization,Storage Virtualization,    Desktop Virtualization
(5.)    Database: SQLServer,MySQL,Oracle, DB2
(6.)    OS: Linux, Unix, Windows
# Additional class

5. 打印文件中的行

实例1：打印文件中的第n行

$ sed -n '4p' techClass.txt
Virtualization:Server Virtualization,Storage Virtualization,    Desktop Virtuali                                                                                                        zation

实例2：文件从第n行开始打印，并每隔m-1行就打印一行。

$ sed -n '3~2p' techClass.txt
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
5.      Database: SQLServer,MySQL,Oracle, DB2
# Additional class

实例3：打印文件的最后一行

$ sed -n '$p' techClass.txt
# Additional class

实例4：打印文件的第n~m行

$ sed -n '2,6p' techClass.txt
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows

实例5：打印文件的第n行到最后一行

$ sed -n '3,$p' techClass.txt
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows
# Additional class

实例6：打印文件中匹配指定模式的行

$ sed -n '/Network/p' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network

实例7：打印文件中从匹配指定模式的行到第n行的内容

$ sed -n '/Security/,6p' techClass.txt
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows

实例8：打印文件中从第n行道匹配指定模式的行的内容

$ sed -n '1,/Database/p' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2

实例9：打印文件中从匹配指定模式的行到最后一行

$ sed -n '/Server/,$p' techClass.txt
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows
# Additional class

实例10：打印文件中匹配指定模式的行及后面的n行

$ sed -n '/Network/,+1p' techClass.txt
1.      Network: Route, Switch, Wireless,Communicate, Device
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization

实例11：打印文件中从匹配模式的行到匹配另一个指定模式的行的内容

$ sed -n '/Security/,/OS/p' techClass.txt
2.      Security: Data Protection,Terminal Security,C1oud Security,WEB Security
3.      Server: Blade,Mini Computer,Mainframes,HPC,Disaster Recovery, Network
4.      Virtualization:Server Virtualization,Storage Virtualization,    Desktop                                                                                                         Virtualization
5.      Database: SQLServer,MySQL,Oracle, DB2
6.      OS: Linux, Unix, Windows