Question

I am trying the following:

.+\?[^\?].+\?

That is, match everything til a “?”, then, if no other “?” after that, match everything til another “?”. I guess it is not working because the first .+ matches the entire string already.

Answer 1

Your description isn’t precise, but here’s the rules I assume you need your regex to satisfy:

1. Zero or more characters that are not ‘?’
2. A ‘?’
3. One or more characters that are not ‘?’
4. A ‘?’

This is actually pretty simple.

[^?]*\?[^?]+\?

Note that this will match a substring of a larger string. If you need to ensure the entire string exactly matches this, throw in ^ and $ anchors:

^[^?]*\?[^?]+\?$

---

Explanation:

1. ^

   Start of the string. In a multiline context this also matches start of the line, but you’re probably not in a multiline context.

2. [^?]

   Match anything that’s not the literal character ‘?‘.

3. *

   Match zero or more of the previous token.

4. \?

   Match a literal ‘?‘.

5. [^?]

   Match anything that’s not the literal character ‘?‘.

6. +

   Match one or more of the previous token. This ensures that you cannot have two ‘?‘ in a row.

7. \?

   Match a literal ‘?‘.

8. $

   Match the end of the string (or the end of the line in a multiline context).

---

Note: I assumed zero or more non-‘?‘ before the first ‘?‘. This will match something like ?abc?. If this is illegal, change the first * to a +.