Linux sed常规用法小结

sed是一个非交互性的流编辑器，是stream editor的缩写。sed每次只处理一行内容，可以对文本或标准输入流进行处理。需要注意的是，sed并不直接操作初始数据，它操作的是一份原始数据的拷贝。sed处理时，把当前处理的行存储在临时缓冲区中，然后处理缓冲区中的内容，处理完成后，如果没有重定向到文件， 将把缓冲区中的内容送往屏幕，接着处理下一行直到处理完毕。
理论的东东，这里不谈了，下面谈些常规应用吧。
在演示之前，我们先构筑个初始文件orig.txt 和append.txt

[plain] view plain copy

1. [root@localhost.shell]#catorig.txt
2. idnamepriceaddress
3. --------------------------------
4. 1apple$50USA
5. 2pear$60Canda
6. 3pen$1China
7. --------------------------------
8. total$111
9. [root@localhost.shell]#catappend.txt
10. 4apple$50USA
11. 5$60Canda
12. 6pen$1China

1、打印指定行

a)打印第一行

[plain] view plain copy

1. [root@localhost.shell]#sed-n'1p'orig.txt
2. idnamepriceaddress

b)打印3-5行

[plain] view plain copy

1. [root@localhost.shell]#sed-n'3,5p'orig.txt
2. 1apple$50USA
3. 2pear$60Canda
4. 3pen$1China

c)打印最后一行

[plain] view plain copy

1. [root@localhost.shell]#sed-n'$p'orig.txt
2. total$111

d)打印所有行

[plain] view plain copy

1. [root@localhost.shell]#sed-n'1,$p'orig.txt
2. idnamepriceaddress
3. -------------------------------
4. 1apple$50USA
5. 2pear$60Canda
6. 3pen$1China
7. --------------------------------
8. total$111

e)打印含有pen的行

[plain] view plain copy

1. [root@localhost.shell]#sed-n'/pen/p'orig.txt
2. 3pen$1China

f)从第一行开始打印，打印到第一个含有$行

[plain] view plain copy

1. [root@localhost.shell]#sed-n'1,/\$/'porig.txt
2. idnamepriceaddress
3. -------------------------------
4. 1apple$50USA

2、打印含有元数据$行的行号

[plain] view plain copy

1. [root@localhost.shell]#sed-n'/\$/='orig.txt
2. 3
3. 4
4. 5
5. 7

3、替换

a)把-替换为空

[plain] view plain copy

1. [root@localhost.shell]#sed-e's/-*//g'orig.txt
2. idnamepriceaddress
4. 1apple$50USA
5. 2pear$60Canda
6. 3pen$1China
8. total$111

b)把元数据$替换为￥

[plain] view plain copy

1. [root@localhost.shell]#sed's/\$/￥/g'orig.txt
2. idnamepriceaddress
3. -------------------------------
4. 1apple￥50USA
5. 2pear￥60Canda
6. 3pen￥1China
7. --------------------------------
8. total￥111

c)-替换为空，并删除空行

[plain] view plain copy

1. [root@localhost.shell]#sed-e's/-*//g'-e'/^$/'dorig.txt
2. idnamepriceaddress
3. 1apple$50USA
4. 2pear$60Canda
5. 3pen$1China
6. total$111

d)把含有total的行，替换为******

[plain] view plain copy

1. [root@localhost.shell]#sed'/total/c\******'orig.txt
2. idnamepriceaddress
3. -------------------------------
4. 1apple$50USA
5. 2pear$60Canda
6. 3pen$1China
7. --------------------------------
8. ******

4、删除，删除前两行

[plain] view plain copy

1. [root@localhost.shell]#sed'1,2d'orig.txt
2. 1apple$50USA
3. 2pear$60Canda
4. 3pen$1China
5. --------------------------------
6. total$111

5、添加

a)在含有apple的行前插入一行 0 nothing $5 USA

[plain] view plain copy

1. [root@localhost.shell]#sed'/apple/i\0nothing$5USA'orig.txt
2. idnamepriceaddress
3. -------------------------------
4. 0nothing$5USA
5. 1apple$50USA
6. 2pear$60Canda
7. 3pen$1China
8. --------------------------------

b)在含有pen的行后插入一行4 four $5 USA

[plain] view plain copy

1. [root@localhost.shell]#sed'/pen/a\4four$5USA'orig.txt
2. idnamepriceaddress
3. -------------------------------
4. 1apple$50USA
5. 2pear$60Canda
6. 3pen$1China
7. 4four$5USA
8. --------------------------------
9. total$111

c)在含有pen的行后面追加append.txt 文件的内容

[plain] view plain copy

1. [root@localhost.shell]#sed'/pen/rappend.txt'orig.txt
2. idnamepriceaddress
3. -------------------------------
4. 1apple$50USA
5. 2pear$60Canda
6. 3pen$1China
7. 4apple$50USA
8. 5$60Canda
9. 6pen$1China
10. --------------------------------
11. total$111

6、导出文件

a)把含有$的行导出到out.txt文件中

[plain] view plain copy

1. [root@localhost.shell]#sed-n'/\$/wout.txt'orig.txt
2. [root@localhost.shell]#catout.txt
3. 1apple$50USA
4. 2pear$60Canda
5. 3pen$1China
6. total$111

b)把含有apple的行输出到apple.txt文件中

[plain] view plain copy

1. [root@localhost.shell]#sed-n'/apple/'porig.txt>apple.txt
2. [root@localhost.shell]#catapple.txt
3. 1apple$50USA

转载自： http://blog.youkuaiyun.com/love__coder/article/details/7024739