Match all after "word" if exists, if "word" does not exist, match all

Question

I have the two following sentences:

1: KEYWORD this is a random sentence
2: this is another random sentence

I would like to match all after KEYWORD if KEYWORD exists, and if it doesn't, I want to match all.
So the output would be:

1: this is a random sentence
2: this is another random sentence

I tried a few things that did not lead me anywhere close:

(?<=KEYWORD)[\s\S]+

This would only match everything after the keyword, but will fail to capture anything if the keyword is not present...

PS: I use java

Add a language tag. Show code. Regexes are not standalone concepts in Java. You have to show how you match them, since there are a number of options available, many of which are not what you want. — Mad Physicist, Feb 28 '18 at 16:22
@rock321987. Because then you wouldn't need the sledgehammer that is regex. — Mad Physicist, Feb 28 '18 at 16:24
[You may use this regex `(?<=KEYWORD |^(?!.*KEYWORD)).*`](https://regex101.com/r/TdyA7Y/1) — anubhava, Feb 28 '18 at 16:27
@ctwheels This is exactly what I wanted, thanks a lot. Do you want to add it as an answer? — Kylke, Feb 28 '18 at 16:27

score 3 · Answer 1 · answered Feb 28 '18 at 16:27

3

See regex in use here

(?<=KEYWORD).*|^(?:(?!KEYWORD).)*$

Option 1
- (?<=KEYWORD) Positive lookbehind ensuring what precedes is KEYWORD
- .* Match any character (except \n) any number of times
Option 2
- ^ Assert position at the start of the line
- (?:(?!KEYWORD).)* Tempered greedy token matching any character except where KEYWORD matches
- $ Assert position at the end of the line

Result:

this is a random sentence
this is another random sentence

answered Feb 28 '18 at 16:27

ctwheels

21,901
9
42
77

Tempered greedy token pattern isn't really needed, I left a comment above. – anubhava Feb 28 '18 at 16:29
2

@anubhava interesting, I like it you should post as another option :) – ctwheels Feb 28 '18 at 16:30
Thanks, I did not know about tempered greedy token. It looks usefull in some cases. – Kylke Feb 28 '18 at 16:41

score 3 · Accepted Answer · answered Feb 28 '18 at 16:35

3

As an alternative to ctwheel's nice working solution you may be able to use this regex that may perform faster because of no tempered greedy token that asserts a lookahead before every character.

(?<=KEYWORD |^(?!.*KEYWORD)).*

RegEx Demo

Details:

(?<=KEYWORD: If we have "KEYWORD " at previous position
|: OR
^(?!.*KEYWORD): We don't have KEYWORD anywhere in the line
.*: Match 0 or more of any characters

answered Feb 28 '18 at 16:35

anubhava

761,203
64
569
643

1

Thanks :) That makes a lot of sense! No idea, why I haven't thought of this when I looked for a solution... – Kylke Feb 28 '18 at 16:39

Match all after "word" if exists, if "word" does not exist, match all

2 Answers2