Seperating words in between dash with a line break

Matthew Habash

Fellow Notepad++ Users,

Could you please help me the the following regular expression problem I am having?

I am trying to separate words between dashes with a line break but I’ve scoured forums and message threads and haven’t turned up much.

Here is a sample text line.

This is  a sample text---these are the words in between dashes--this is the end of the sample text

Here is how I would like it to look.

This is  a sample text
---these are the words in between dashes---
this is the end of the sample text

Notice how the words in between dashes are in seperate lines.

To accomplish this, I have tried using the following Find/Replace expressions and settings

Find What = (?=---)(.*)(?=---)
Replace With = \n
Search Mode = REGULAR EXPRESSION
Dot Matches Newline = NOT CHECKED

I can only get it to look like this:
This is a sample text
—these are the words in between dashes–this is the end of the sample text

Unfortunately, this did not produce the output I desired, and I’m not sure why. Could you please help me understand what went wrong and help me find the solution?

PeterJones

@Matthew-Habash said in Seperating words in between dash with a line break:

Unfortunately, this did not produce the output I desired, and I’m not sure why. Could you please help me understand what went wrong and help me find the solution?

I do not think you understand lookahead, nor fully understand your data.

In your Find What, the (?=---) says “the next three characters must be hypens, but DON’T SELECT THEM YET OR MOVE THE CURSOR” – so that puts the search cursor at the beginning of the ---these are the words in between dashes---. Then (.*) says “grab 0 or more of any type of character…”. Then the second (?=---) says "… until you reach a point where the next three characters are hyphens, but don’t select those final three hyphens

In your example text, there is only one triple-hyphen; the second set of dashes are just a double hyphen. This means that the (.*)(?=---) matches 0 characters – because if the .* doesn’t “eat up” any characters, then the lookahead will find the original triple-hyphen as the matching point. So you’ve got a 0-width selection just before the first and only ---.

Here’s what it matches, using Find Next instead of Replace, so you can see the “selection” (which it indicates with ^ zero length match):

Then your replacement says “replace everything that was matched” (ie, the emptiness just before the ---) “with a LF sequence”, which gives you:

Please note: if your sample text had the text you thought it had three hyphens at the end of the phrase:

This is  a sample text---these are the words in between dashes---this is the end of the sample text

then your regex would have selected the first three hyphens, and the words in between:

And then the replace would have replaced all that selected text with a newline,

… which still wouldn’t be what you want.

Based on your “how I would like it to look”, I think what you really intended to do was to put a newline before the first --- and another newline after the second --- (if there really were a second ---. To do that, I would do the following:

Find What: ---.*---
Replace With: \r\n$0\r\n
Search Mode: Regular Expression
. matches newline not checked

The Find What is simpler: you want to match everything, including the hyphens, so no reason for the lookaheads and groups.
In the Replace With, I am using $0 to say “include the contents of everything that was matched”, so that your original text isn’t lost. I am also using \r\n instead of just \n, because I am assuming you really have Windows-style newline (CRLF), not Unix/Linux newlines (LF only). If you actually do have Linux LF only, then Replace With: \n$0\n

Results:
FIND NEXT =>

REPLACE ALL =>

This is  a sample text
---these are the words in between dashes---
this is the end of the sample text

But my regex will not work for your original single-line, because that one only has two hyphens, not three, so it will find no matches and do no replacements.

Matthew Habash

@PeterJones Thank you for the explanation and my apologies for the confusion. I meant for both sides of lines to have the same number of dashes. I guess I did not catch that.

Your regex code worked and that was what I needed. Thank you very much!