Toggle source

sed命令相关技巧

sed 替换回车

sed '/^$/!{:n;N;/\n$/!s/\n//;tn;}' xxx

回车替换为空格

sed ':a;N;$!ba;s/\n +/ /g' xxx

sed 替换单引号

例如下面的例子中通过 sed 命令将 xx'yy 替换为 zz@tt：

sed -i 's/xx'\''yy/zz@tt/g' xxx

输出匹配行后面的行

输出匹配行加上下一行：

sed -n '/pattern/{N;p;}' xxx

如果需要附加更多的行，例如下面表示再附加 3 行：

sed -n '/pattern/,+3p' xxx

输出匹配行前面的行

输出匹配行的前面一行：

sed -n '/pattern/{x;p;d;}; x' xxx

工作机制：

sed has a hold space and a pattern space. New lines are read into the pattern space. The idea of this script is the the previous line is saved in the hold space.

This the current line contains pattern, then do:

x : swap the pattern and hold space so that the prior line is in the pattern space

p : print the prior line

d : delete the pattern space and start processing next line
x

This is executed only on lines which do not contain age is : 10. In this case, save the current line in the hold space.

输出匹配行的前面一行加上匹配行：

sed -n '/pattern/{x;p;x;p}; h' xxx

查找匹配行之后多次替换

多个匹配命令以分号隔开，匹配后的所有命令用大括号括起来：

sed -n '/dc-version/{s/^.*value="//;s/".*$//;p}' xxx

替换中使用匹配内容

替换块中使用 & 表示完整匹配内容：

~ # echo "BEGIN (254:6) END" | sed -E 's/\([0-9]*:([0-9]*\))/, &/'
BEGIN , (254:6) END

替换块中使用 \1 \2 等表示对应的子匹配块

~ # echo "BEGIN (254:6) END" | sed -E 's/:([0-9]*\))/, \1/'
BEGIN (254, 6) END
~ # echo "BEGIN (254:6) END" | sed -E 's/\(([0-9]*):([0-9]*)\)/[\1, \2]/'
BEGIN [254, 6] END

替换块之后使用数字表示替换第几个匹配项

~ # echo "BEGIN (254:6:117) END" | sed -E 's/([0-9]+)/666/2'
BEGIN (254:666:117) END

# sed命令相关技巧

## sed 替换回车

```bash
sed '/^$/!{:n;N;/\n$/!s/\n//;tn;}' xxx
```

## 回车替换为空格

```bash
sed ':a;N;$!ba;s/\n +/ /g' xxx
```

## sed 替换单引号

例如下面的例子中通过 sed 命令将 `xx'yy` 替换为 `zz@tt`：

```bash
sed -i 's/xx'\''yy/zz@tt/g' xxx
```

## 输出匹配行后面的行

输出匹配行加上下一行：

```bash
sed -n '/pattern/{N;p;}' xxx
```

如果需要附加更多的行，例如下面表示再附加 3 行：

```bash
sed -n '/pattern/,+3p' xxx
```

## 输出匹配行前面的行

输出匹配行的前面一行：

```bash
sed -n '/pattern/{x;p;d;}; x' xxx
```

工作机制：

sed has a hold space and a pattern space. New lines are read into the pattern space. The idea of this script is the the previous line is saved in the hold space.

* This the current line contains pattern, then do:

**x** : swap the pattern and hold space so that the prior line is in the pattern space

**p** : print the prior line

**d** : delete the pattern space and start processing next line
* **x**

This is executed only on lines which do not contain age is : 10. In this case, save the current line in the hold space.

输出匹配行的前面一行加上匹配行：

```bash
sed -n '/pattern/{x;p;x;p}; h' xxx
```

## 查找匹配行之后多次替换

多个匹配命令以分号隔开，匹配后的所有命令用大括号括起来：

```bash
sed -n '/dc-version/{s/^.*value="//;s/".*$//;p}' xxx
```

## 替换中使用匹配内容

* 替换块中使用 `&` 表示完整匹配内容：

```bash
~ # echo "BEGIN (254:6) END" | sed -E 's/$[0-9]*:([0-9]*$)/, &/'
BEGIN , (254:6) END
 ```

* 替换块中使用 `\1` `\2` 等表示对应的子匹配块

```bash
~ # echo "BEGIN (254:6) END" | sed -E 's/:([0-9]*\))/, \1/'
BEGIN (254, 6) END
~ # echo "BEGIN (254:6) END" | sed -E 's/$([0-9]*):([0-9]*)$/[\1, \2]/'
BEGIN [254, 6] END
 ```

* 替换块之后使用数字表示替换第几个匹配项

```bash
~ # echo "BEGIN (254:6:117) END" | sed -E 's/([0-9]+)/666/2'
BEGIN (254:666:117) END
 ```

Uranus' Wiki