两个第n个位置字符之间的正则表达式

Question

I'm trying to fetch some data depending from a text string that lies between two characters (_) but could be a word in a nth position.

Currently I have the following

!((?:.*?(_)){2})_(.+?)$

working on the following data

D20_Mbps_U10_Mbps_TC4_P

where I would expect to get

U10

but get nothing as the first part captures

D20_Mbps_

and thus leaves nothing for the second part to capture

I've tried

_\s*(.*?)(?=\s*_)

But this only gives me the first occurance where I need it to be nth position. Where I can supply n at runtime.

any ideas?

Thanks

Answer 1

Let me try answering this in detail.

When you want to match some Nth occurrence of a substring within a delimited string, you should really think of some String.Split function. In your case, splitting with _ and getting the values you need is a trivial task.

Now, when you cannot use a programming means to extract that value, you can only do this with a limiting quantifier , grouping and capturing (in Java and .NET, it is possible to achieve the same even without capturing).

So, the main idea is to match 0 or more characters other than your delimiter and then match the delimiters itself, and then repeat the same N-1 times. Then, just match the delimiter again and capture following non-delimiter characters.

^(?:[^_]*_){2}([^_]*)

See demo . Group 1 will contain U10 .

Or another variation :

^(?:[^_]*_){2}([^_]*)_(.+)$

This will capture the 3rd _ -delimited element into Group 1. Group 2 in this case is the 4th+ elements, the rest of the string up to the end.

Note that in some regex flavors { and ( must be escaped (vim, sed with non-EGREP versions, etc.).