Properly distribute furigana over keywords and example sentences

[stephenmk]

This deck does not correctly distribute furigana over keywords and sentences. Beginners may not be able to tell that 恐怖 is きょう + ふ rather than きょ + うふ, for example.

I used a furigana solver to fix the distributions.

The data is posted here: https://gist.github.com/stephenmk/3f7fb2d36b5c990298f2319b4f06c8fb

The solver is somewhat opinionated. For example, it doesn't consider 日本 to be separable into に + ほん; it considers にほん to be an indivisible idiomatic reading. If you disagree, you could replace all instances of 日本[にほん] with 日[に] 本[ほん].

Before	After
	
	
	

I only found a few errors in the deck data. There was the こわばった error that you can see in the example sentence for 恐怖 displayed above. The "Word Furigana" values for 戦い and 入り口 also had errors.

[donkuri]

This is absolutely fantastic! Thank you very much, I have updated the cards accordingly. This is now Kaishi 1.5k v1.3.0.