正则表达式是一种字符模式，用于在查找过程中匹配制定的字符。
元字符通常在 Linux 中分为两类：
-
Shell 元字符，由 Linux Shell 进行解析。
-
正则表达式元字符，由 vi/grep/sed/awk 等文本处理工具进行解析。
正则表达式一般以文本行进行处理，在进行下面实例之前，先为 grep 命令设置 --color 参数：
$ alias grep='grep --color=auto'
这样每次过滤出来的字符串都会带色彩了。
在开始之前还需要做一件事情，就是创建两个测试用的文件re-file 和 linux.txt，内容如下：
$ cat re-fileI had a lovely time on our little picnic.Lovers were all around us. It is springtime. Ohlove, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.is forever. I live for you. It's hard to get back in thegroove.$ cat linux.txtLinux is a goodgod assdxw bcvnbvbjkgreatttttt wexcvxcoperaaaating dhfghfvxgooodfs awrerdxxhklgdsystem awxxxxgladgood
正则表达式元字符
| 元字符 | 功能 |
|---|---|
| ^ | 以什么开头 |
| $ | 以什么结尾 |
| . | 匹配一个字符 |
| * | 匹配0个或多个 |
| [] | 匹配集合中的 |
| [x-y] | 匹配集合范围内的 |
| [^ ] | 匹配不在集合中的 |
| \ | 转义 |
-
特殊的元字符
| 元字符 | 功能 | 实例 | 怎么匹配 |
|---|---|---|---|
| \< | 以什么开头 | '\<love' | 匹配以love开头的所有行 |
| \> | 以什么结尾 | 'love\>' | 匹配love结尾的所有行 |
| \(..\) | 标签匹配以后使用的字符 | '\(love\)able \1er' | 用位置\1\2引导前面做好的标签，最大支持9个 |
| x\{m\} or x\{m,\} or x\{m,n\} | 重复字符x，m次，至少m次，至少m且不超过n次 | o\{5,10\} | o字符重复5到10次的行 |
-
扩展的正则表达式
| 元字符 | 说明 |
|---|---|
| + | 重复前一个字符一个或一个以上 |
| ？ | 0个或者一个字符 |
| | | 表示或，查找多个字符串 |
| () | 分组过滤匹配 |
实操
-
匹配以 love 开头的所有行
$ grep '^love' re-filelove, how much I adore you. Do you know
-
匹配 love 结尾的所有行
$ grep 'love$' re-fileclover. Did you see them? I can only hope love.
-
匹配以 l 开头，中间包含两个字符，结尾是 e 的所有行
$ grep 'l..e' re-fileI had a lovely time on our little picnic.love, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.is forever. I live for you. It's hard to get back in the
-
匹配 0 个或多个空行，后面是 love 的字符
$ grep ' *love' re-fileI had a lovely time on our little picnic.love, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.
-
匹配 love 或 Love
$ grep '[Ll]ove' re-file # 对l不区分大小写I had a lovely time on our little picnic.Lovers were all around us. It is springtime. Ohlove, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.
-
匹配 A-Z 的字母，其次是 ove
$ grep '[A-Z]ove' re-fileLovers were all around us. It is springtime. Oh
-
匹配不在 A-Z 范围内的任何字符行，所有的小写字符
$ grep '[^A-Z]' re-fileI had a lovely time on our little picnic.Lovers were all around us. It is springtime. Ohlove, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.is forever. I live for you. It's hard to get back in thegroove.
-
匹配 love.
$ grep 'love\.' re-fileclover. Did you see them? I can only hope love.
-
匹配空格
$ grep '^$' re-file
-
匹配任意字符
$ grep '.*' re-fileI had a lovely time on our little picnic.Lovers were all around us. It is springtime. Ohlove, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.is forever. I live for you. It's hard to get back in thegroove.
-
前面 o 字符重复 2 到 4 次
$ grep 'o\{2,4\}' re-filegroove.
-
重复 o 字符至少 2 次
$ grep 'o\{2,\}' re-filegroove.
-
重复 o 字符最多 2 次
$ grep 'o\{,2\}' re-fileI had a lovely time on our little picnic.Lovers were all around us. It is springtime. Ohlove, how much I adore you. Do you knowthe extent of my love? Oh, by the way, I thinkI lost my gloves somewhere out in that field ofclover. Did you see them? I can only hope love.is forever. I live for you. It's hard to get back in thegroove.
-
重复前一个字符一个或一个以上
$ egrep "go+d" linux.txtLinux is a goodgod assdxw bcvnbvbjkgooodfs awrerdxxhklgood
-
0 个或者一个字符
$ egrep "go?d" linux.txtgod assdxw bcvnbvbjkgdsystem awxxxx
-
同时查找多个字符串
$ egrep "gd|good" linux.txtLinux is a goodgdsystem awxxxxgood
-
分组过滤匹配
$ egrep "g(la|oo)d" linux.txtLinux is a goodgladgood
