Get the right data in Kanji

[dper]

In the class Kanji, get all the necessary information available. We need:

• The literal. One character.
• A one word meaning, probably in all caps.
• The rest of the definition, as a truncated string
• The onyomi, as a truncated string.
• The kunyomi, as a truncated string.
• One to three example words and associated reading and meaning, as a truncated string.

We might want:

• JLPT level, except that right now Kanjidic2 doesn't have it.
• Grade level. Not all kanji have one, but if they do, we can display it on the front or back of the card. It is not strictly necessary, but I think it might be rather motivating.
• Stroke count. We can display this on the front or the back of the card. But it might not be useful.

Most of this is already there from the old code, but it needs to be confirmed.

[dper]

This is an example of my old paper flash cards. This shows the back of the card. The new cards should look similar to this. We could also add the grade level. We don't know the JLPT levels any more (since the 2010 changes), so we can't display those.

[dper]

All of the information is now there, with the following caveats.

• There might be a bug where kanji with no grade level lead to badness.
• The definitions are not split. Should be done by the displaying class.
• Readings are not truncated. Should be done by the displaying class.
• No long examples are included. There's no good way to push this off until display time, because if we exclude an example, we want to replace it with a different example. We can't just truncate it like we do with readings and definitions.
• Capitalization is not touched here. Should be done by the displaying class.