Java ReplaceAll Regular Expression With Exclusions

Question

I am trying to replace all instances of sentence terminators such as '.', '?', and '!', but I do not want to replace strings like "dr." and "mr.".

I have tried the following:

text = text.replaceAll("(?![mr|mrs|ms|dr])(\\s*[\\.\\?\\!]\\s*)", "\n");

...but that does not seem to work. Any suggestions would be appreciated.

Edit: After the feedback here and a bit of tweeking this is the working solution to my problem.

private String convertText(String text) {
  text = text.replaceAll("\\s+", " ");
  text = text.replaceAll("[\n\r\\(\\)\"\\,\\:]", "");
  text = text.replaceAll("(?i)(?<!dr|mr|mrs|ms|jr|sr|\\s\\w)(\\s*[\\.\\?\\!\\;](?:\\s+|$))","\r\n");
  return text.trim();
}

The code will extract all* compound and single sentences from an excerpt of text, removing all punctuation and extraneous white-space.
*There are some exceptions...

Try removing the brackets, [], from around the list of exceptions: (?!mr|mrs|ms|dr). They stand for "character set", not "full strings" as you're using them. Don't know if it will entirely solve your problem, but it's worth a start — newfurniturey
– newfurniturey, Commented Dec 6, 2012 at 5:20
There's several problems with trying to do that though. How are you going to handle sequences like J. H. Ronaldo says that the train is running on time.... Is he right?. — Anthill
– Anthill, Commented Dec 6, 2012 at 5:25
@Anthill, I have added support for ignoring single characters that precede a period. Is this the correct way of is there an even easier method? — Mr. Polywhirl
– Mr. Polywhirl, Commented Dec 7, 2012 at 1:24

Narendra Yadala · Accepted Answer · 2012-12-06 05:20:36Z

2

You need to use negative lookbehind instead of negative lookahead like this

String x = "dr. house.";
System.out.println(x.replaceAll("(?<!mr|mrs|ms|dr)(\\s*[\\.\\?\\!]\\s*)","\n"));

Also the list of mr/dr/ms/mrs should not be inside character classes.

answered Dec 6, 2012 at 5:20

Narendra Yadala

9,6641 gold badge31 silver badges44 bronze badges

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

1 Comment

Mr. Polywhirl Over a year ago

I was so close, I remember negative look-behinds vaguely. Thanks.

The111 · Accepted Answer · 2012-12-06 05:19:09Z

-1

You're going to need to have a complete list of the letter combinations which are allowed to precede .. Then, you can replace dr. and mr. (and any other allowed combos) with something unique like dr28dsj458sj and mr28dsj458sj. Ideally you should check that your temp substitute value exists nowhere else in the document. Then go through and remove all your sentence terminators, then go through again and replace the occurrences of 28dsj458sj with . again.

answered Dec 6, 2012 at 5:19

The111

5,8674 gold badges42 silver badges56 bronze badges

1 Comment

Mr. Polywhirl Over a year ago

I like this hacky workaround, but it may slow down performance. The negative look-behind was what I was going for. Thanks for your time.

Collectives™ on Stack Overflow

Java ReplaceAll Regular Expression With Exclusions

2 Answers 2

1 Comment

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related