Regex capturing within a group

Question

I want to be able to delete all instances of newlines within <p> tags, but not the ones outside. Example:

<p dir="ltr">Test<br>\nA\naa</p>\n<p dir="ltr">Bbb</p>

This is the regex I came up with:

(<p[^>]*?>)(?:(.*)\n*)*(.*)(</p[^>]*?>)

and I replace with:

$1$2$3$4

I was hoping that this would work but (?:(.*)\n*)* seems to be causing issues. Is there any way to do repeated matches like this, with a capturing group?

Thanks in advance!

there are two p tags? you want \n to be removed separately for them? — rock321987
– rock321987, Commented May 23, 2016 at 18:12
Separately for p tags is fine. Its just that I'm hoping to replace all the \n within the p tags in one fell swoop. I was hoping that its possible with regex without nested loops. — thisbytes
– thisbytes, Commented May 23, 2016 at 18:16
I would recommend using something like JSoup for this kind of work. — ThePerson
– ThePerson, Commented May 23, 2016 at 18:25

rock321987 · Accepted Answer · 2016-05-24 09:13:10Z

2

Solution

You can use this regex(works in PCRE but not in Java. For Java version refer below)

(?s)(?:<p|\G(?!\A))(?:(?!<\/p>).)*?\K[\n\r]+

Regex Demo

Regex Breakdown

(?s) #Enable . to match newlines

(?:
   <p #this part is to assure that whatever we find is inside <p tag
    | #Alternation(OR)
   \G(?!\A) #Find the position of starting of previous match.
)

(?:
  (?!<\/p>). #Till it is impossible to match </p>, match .
)*? #Do it lazily

\K #Whatever is matched till now discard it

[\n\r]+ #Find \n or \r

Java Code

With a bit of modification, I was able to do it in Java

String line = "<p dir=\"ltr\">Test<br>\nA\naa</p>\nabcd\n<p dir=\"ltr\">Bbb</p>"; 
System.out.println(line.replaceAll("(?s)((?:<p|\\G(?!\\A))(?:(?!<\\/p>).)*?)[\\n\\r]+", "$1"));

Ideone Demo

edited May 24, 2016 at 9:13

answered May 23, 2016 at 18:28

rock321987

11.1k1 gold badge34 silver badges44 bronze badges

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

17 Comments

thisbytes Over a year ago

Holy... Wow. that's pretty darn amazing.

zec Over a year ago

damn my regex noobness! good job rock - i was too slow to be the savior.

rock321987 Over a year ago

@Jun first let me check it in JAVA

bobble bubble Over a year ago

Just was about to add answer "(?s)\\n+(?=(?:(?!<p).)*?</p)" as a Java idea when I saw you did update for Java. I like your answer.

bobble bubble Over a year ago

@rock321987 Your pattern is more accurate and if there is a long html input with many \n outside of <p I think your approach will outperform my idea as mine would trigger the lookahead at any \n+.

|

Collectives™ on Stack Overflow

Regex capturing within a group

1 Answer 1

17 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

17 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related