I have a text transcription of someone recalling a story. Provide a word-for-word segmentation of the following recall into linguistic clauses, numbered in order of appearance in the recall. Do not change or remove any words from the text. Once you have done that, output a CSV with each row being a clause. When segmenting, focus on punctuation, topic shifts, and new noun-verb units. Also, be sure to leave phrases such as "oh my god" intact and in their own events. Segment into clauses where every new finite verb with its own (even implied) subject starts a new clause; coordinators that launch a fresh verb = split; keep fillers intact; preserve exact wording. Include lead in words (like "Well,") with the clause that immediately follows it. Make sure that filler or transition words ("and", "but", "oh") are not their own clauses. 
Independent clauses that represent a single thought should be kept together. For instance, “He said that he was sad” is a single thought. Do NOT break it up into two clauses (i.e., “He said” and “that he was sad”). As another example, “I think the time jumped” is a single thought. Do NOT break it up into two clauses (i.e., “I think” and “the time jumped”).
Even if you think that a word was duplicated (for example the same word is repeated twice), keep both instances in. The provided clause segmentation should be 100% identical to the original transcription and the text words. Once you're done, please check that the transcript you generated is word for word identical with the ones originally provided to you. If so, please give me a CSV that I can download, which lists out the clauses you generated. If not, edit it to make sure the words are identical. Thanks!