I love spaced-repetition systems. I've been an extremely early Anki user, all the way back to 2006.

Yet nothing beats rote language drills to build automaticity.

So I've previously written language drilling apps to help me learn languages.

I've experimented with many possible schemas; yet I've come to prefer storing simple sentence-level units in SQL databases.

This to me is the simplest single-table definition schema to hold learning materials across different languages. (It's great because it also works as a reference for Anki deck fields.)

Explaining the Schema

1. expression

   • Represents a grammatical phrase, expression or a sentence fragment.

   • Written in Japanese/other languages.

2. expression_meaning

   Closest translation. Not the 'explanation' (which goes into notes).

3. expression_reading

   Phonetic reading. (For Japanese, in hiragana/katakana.)

4. expression_source

   Reference - e.g. specific texts/documents/websites.

5. focus_phrase

   • A specific word, or target grammar point, within the expression.

   • If present, it is the focal point of the expression.

6. focus_phrase_meaning

   • Meaning of the highlighted phrase/grammar point.

   • Different from expression_meaning.

7. notes

   Supplementary details; usage tips, exceptions, or cultural nuances.