Matching sentences can be tricky. There are a lot of rules and edge cases, so you may need to make additional adjustments.
If your sentences are well-formed, ending in a period, question, exclamation, and the start of the next sentence is capitalized, you can do something like this:
![image.png](https://groups.google.com/group/textsoap/attach/4be7511941b0e/image.png?part=0.2&view=1)
The first part (?<! .. ) is a negative look-behind. This helps prevent the period in an abbreviation from adding a return.
([.?!]\s) this second part (capture $1) looks for an end punctuation and whitespace, capturing it.
Finally ([:uppercase:]) looks for and captures the first capitalized character (which should be the start of the next sentence).
The replacement then puts a return between the end of the sentence and the starting character, including the captured strings so no text is lost.
The second match looks for a sentence ending within a quotation. Something like, "Is this the right answer?"
This is meant as a starting point. There may be additional edge cases, depending on the text you are working with.