It would be nice if the parser didn't include NL for whenever the enter key was used (this happens when making haiku, since there are multiple lines). Also, the parser is detecting a 三点リーダー as an unmatched word, instead of treating it like other punctuation.
I need to visually stick something in there so the parser knows where to break it up - is there a character or characters that you think would not jump out as much?
Edit: the ellipsis will be handled in the next update, most likely tomorrow.
I just think it's weird that the parser actually displays NL in the vocabulary list for the sentence, since punctuation isn't listed; it doesn't seem like something the user needs to see.
Also, got another one for you. For a sentence with 2020年, the parser breaks up the numbers by default as 2 - ふ, 0, 2 - ふ, 0, instead of reading it as one number.
I did a word garden sentence that included the phrase 鰐はしりとりが嫌いだ, and the parser changed some of the hiragana into katakana and parsed it as (鰐)(は)(しり)(トリガ)(嫌いだ).