PCRE pattern to detect nested matches

Question

I use the following pattern to match and replace a simple IF-ELSE-ENDIF block using preg_replace_callback:

##IF (.*?)##(.*?)(?:##ELSE##(.*?))?##ENDIF##

The following two examples are matched and replaced correctly

1. ##IF x()## foo ##ENDIF##

Result:

Group 1: `x()`
Group 2: ` foo `

2. ##IF y()## foo ##ELSE## bar ##ENDIF##

Result:

Group 1: `y()`
Group 2: ` foo `
Group 3: ` bar `

Now I want to enhance this functionality to allow nested IF-ELSE blocks, e.g.

##IF g()## pre-outer-if ##IF h()## inner-if ##ENDIF## post-outer-if ##ELSE## outer-else ##ENDIF##

Here is a demo on regex101.com.

I played around to enhance the pattern with the recursion operator and positive lookaheads but I did not succeed. The goal is to match and replace related IF-ELSE-ENDIF blocks to allow a user to define conditional textblocks.

If possible, I want to avoid writing a language parser for this enhancement.

Thanks a lot for any suggestion!

What exactly are you doing with the matches? Please share the method with preg_replace_callback. — Wiktor Stribiżew
– Wiktor Stribiżew, Commented Jul 26, 2018 at 13:03
@Wiktor: The blocks are part of a bigger text and allows the users of the system to create dynamic text modules based on those conditions. The whole condition is replaced by either the ##IF## or the ##ELSE## part. The functions x(), y() or g() are just examples for a set of user defined functions. — Andre
– Andre, Commented Jul 26, 2018 at 13:10

Michail · Accepted Answer · 2018-07-28 04:05:39Z

2

Try this:

##IF (?>(.*?)##)((?>(?0)|.)*?)(?:##ELSE##((?>(?0)|.)*?))?##ENDIF##

It will find a outer if-block groups.

edited Jul 28, 2018 at 4:05

answered Jul 28, 2018 at 3:56

Michail

8734 silver badges11 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Andre Over a year ago

Deleted my last comment because it was wrong. This regex seems to work fine. I use it inside a loop until no more ##IF## statements are found. Could you explain why the first atomic group (?>(.*?)##) is necessary?

Michail Over a year ago

It effectively reduce backtracking in no-match case (for example, ##IF witout ##ENDIF##). If no-matches can't occur then atomic group can be omitted.

Collectives™ on Stack Overflow

PCRE pattern to detect nested matches

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related