In the following string blah blahTESTblab la14blah-15S rebla I'm trying to capture TEST and 15S if present.

I wrote the following regex : (TEST).*?(\d{1,2}S)?

Everything is going well until the last ?. The last group really needs to be optional. And groups can't be anchored, as strings I'm searching in are human written, ie tot reliable.

What's wrong ?

In the following string blah blahTESTblab la14blah-15S rebla I'm trying to capture TEST and 15S if present.

I wrote the following regex : (TEST).*?(\d{1,2}S)?

Everything is going well until the last ?. The last group really needs to be optional. And groups can't be anchored, as strings I'm searching in are human written, ie tot reliable.

What's wrong ?

• regex-group

2 Answers 2

The following pattern seems to do what you want:

(TEST)(?:.*?(\d{1,2}S)|)

This says to match:

• (TEST) a required "TEST" in the first capture group
• (?: do not capture
  • .*?(\d{1,2}S) match either any string ending in e.g. "15S" in the second capture group
  • | OR
  • (match nothing at all beyond the first "TEST")
• ) end non capture group

Demo

Thanks again for this answer. I modified your regex a bit in order to work correctly with the libreoffice regex function. I added .* twice, at the beginning and at the end of the regex.

With blah blahTESTblab la14blah-15S reblain cell A1 :

• =REGEX(A1;"(?i)(TEST)(?:.*(\d{1,2}S)|)";"★$1$2") gives blah blah★TEST15S rebla
• =REGEX(A1;"(?i).*(TEST)(?:.*(\d{1,2}S)|).*";"★$1$2") gives ★TEST15S