Split Documents into Retrieval-Safe Chunks

Implement chunk_text(text, chunk_size, overlap) that splits text into overlapping chunks. - Use a sliding window: advance chunk_size - overlap characters each step. - Each chunk (except possibly the last) should be exactly chunk_size characters. - Include the last chunk even if it is shorter than chunk_size. - If overlap >= chunk_size, no overlap is applied (stride equals chunk_size). - Return a list of strings.