Linux 如何删除字符串中的`{{i：`and `}}`

发布于04月06日

我想替换所有出现的

{{i:very long text with some [special](https://google.com) nested text *here*}}

到

very long text with some [special](https://google.com) nested text *here*

That is, in pseudocode, I want 到 get <text> from {{i:<text>}} in a text file.

I don't mind which 到ol you use: grep, awk, ... This could be a beginning.

推荐答案

您可以使用Perl中的递归正则表达式来实现您想要的结果.这个正则表达式将匹配你的字符串，允许嵌套{{i:—}}对:

{{i:((?>(?:(?!{{i:|}}).)+|(?R))*)}}

它匹配:

第{{i:章:文字{{i:
((?>(?:(?!{{i:|}}).)+|(?R))*) : zero or more (captured in group 1) of either
- (?:(?!{{i:|}}).)+:一个或多个不在{{i:或}}序列开头的字符(这是tempered greedy token)；or
- (?R):整个模式的递归
第}}章:文字}}

Regex演示regex101

假设test.txt包含:

{{i:very long text with {some} [special](https://google.com) nested text *here*}}
{{i:hello {{i:hola}}}}

然后您可以:

perl -pe 's/{{i:((?>(?:(?!{{i:|}}).)+|(?R))*)}}/\1/' test.txt

输出:

very long text with {some} [special](https://google.com) nested text *here*
hello {{i:hola}}

实用课程推荐