We have text like this:


Start: {paragraph return}
line 1; {paragraph return}

line 2; {paragraph return}
ending line.{paragraph return}


That is, everything but the ending line could have a colon, semicolon, comma..., but ONLY the "ending line" has a period. This is the only period.


I want to end up with:


Start: line 1; line 2; ending line.{paragraph return}



I already have code that manipulates this for various reasons. I will add code that searches for "space ^p" (a space and then a paragraph return) and replaces with just "^p" (a paragraph return). I'll run this a few times, so that I should end up with "character ^p", each line ending with some type of character then an immediate paragraph return.


What I would like to implement is the following:


If the line does not end with a ".^p" but does end with a "^p", replace the "^p" with a space.



This removes the paragraph return from every line but those that end with a period and paragraph return.



I could "brute force" this by replacing all ";^p" with a space, all ",^p" with a space, ":^p" with a space.



But is there a more elegant way of doing this?