You're talking about sentence boundary detection. In some writing forms (e.g. Chinese), the end-of-sentence marker is unambiguous, and you can just read whole files or whole paragraphs into one scalar variable and use "split" with the distinctive end-of-sentence character (or use that character as the input record separator $/).

But in others (e.g. English), the character used for the end-of-sentence marker is also used for lots of other things -- it is ambiguous, and it can be hard to tell, in any "algorithmic" way, whether a given period marks the end of a sentence or not. It's easy to spot all occurrences of the period character, but it takes a little more work to know which ones are sentence boundaries (and in some situations it takes a lot more work).