正则表达式让我字符串，给定开始和结束

Question

I want to build a function with 3 parameters:

var function = searchMyRegex(begin_marker, end_marker, text) { ... }

I need a RegExp that will give me certain string, given the beginning and end, but it will me the shortest version, meaning that neither the beginning nor the end will be in between the 2 markers

For instance

var text = "Lorem ipsum ipsum sit amet amet adipisicing elit"

If BEGIN is ipsum and END is amet , I want to get in return ipsum sit amet , not ipsum ipsum sit amet amet

The way I tried it to solve myself was like this:

new RegExp(begin_marker + "[^" + end_marker + "]*\ " + end_marker);

But this does not seem to work, and Im not sure why.

Answer 1

You can use this kind of construct:

ipsum(?:(?!ipsum).)*?amet

(?:(?!ipsum).)*? checks if each position after ipsum and before amet isn't the start of an other ipsum .

With the non-greedy quantifier *? , the match stop at the first amet occurrence.

---

Note that the previous technique needs to perform a test with a lookahead for each position. If you deal with a long string, you can also play with character classes to be more efficient:

ipsum[^ia]*(?:i(?!psum)[^ia]*|a(?!met)[^ia]*)*amet

However it's less easy to build this pattern dynamically.

---

When the markers are words (ie that start and end with word characters), you can limit the number of positions tested with the lookahead like this:

\bipsum\W+(?:(?!ipsum\b)\w+\W+)*?amet\b

The lookahead is only tested at the start of words. This one can be easily build dynamically.