halkyon/halkyon_learn_curator/test/kanjidic_test.dart

import 'dart:convert';
import 'dart:io';

import 'package:halkyon_learn_curator/src/sources/kanjidic.dart';
import 'package:test/test.dart';

void main() {
  late List<Map<String, Object?>> lines;

  setUpAll(() async {
    final jlpt = await readKanjiJlpt(
      File('test/fixtures/kanji-data-sample.json'),
    );
    lines = [
      for (final line in kanjidicDictionary(
        await File('test/fixtures/kanjidic2-sample.xml').readAsString(),
        jlpt: jlpt,
      ))
        jsonDecode(line) as Map<String, Object?>,
    ];
  });

  Map<String, Object?> entry(String character) =>
      lines.skip(1).firstWhere((line) => line['id'] == character);
  Map<String, Object?> fields(String character) =>
      entry(character)['entry']! as Map<String, Object?>;

  test('the header says the file holds characters, and whose they are', () {
    final header = lines.first;
    expect(header['source'], 'kanjidic');
    expect(header['kind'], 'characters');
    expect(header['sourceVersion'], '2026-282');
    expect((header['language'], header['glossLanguage']), ('ja', 'en'));
    expect(header['license'], 'CC-BY-SA-4.0');
    expect(
      [
        for (final credit in header['credits']! as List)
          (credit as Map)['name'],
      ],
      ['KANJIDIC2', 'JLPT kanji lists'],
    );
  });

  test('a kanji: its readings, English meanings only, grade, strokes, '
      'frequency and JLPT level', () {
    final eat = entry('食');
    expect(eat['headword'], '食');
    expect(eat['reading'], isNull);
    expect(eat['labels'], ['jlpt:n5', 'grade:2', 'radical:184']);
    expect(fields('食'), {
      'meanings': ['eat', 'food'],
      'on': ['ショク', 'ジキ'],
      'kun': ['く.う', 'く.らう', 'た.べる', 'は.む'],
      'grade': 2,
      'strokes': 9,
      'frequency': 328,
      'jlpt': 5,
    });
    expect(fields('一')['meanings'], isNot(contains('un')));
  });

  test('a search finds a kanji by its on readings in either kana and in '
      'romaji, by whole kun readings and by its meanings; the kanji part '
      'of a kun reading only loosely', () {
    final eat = entry('食');
    expect(
      eat['keys'],
      containsAll([
        '食',
        'ショク',
        'しょく',
        'shoku',
        'たべる',
        'taberu',
        'くらう',
        'eat',
        'food',
      ]),
    );
    expect(eat['keys'], isNot(contains('た')));
    expect(eat['looseKeys'], containsAll(['た', 'ta', 'く', 'ku']));
    // ひと- and ひと.つ: the suffix's dash goes, its kanji part is whole.
    expect(entry('一')['keys'], containsAll(['ひと', 'ひとつ', 'hitotsu']));
    expect(entry('一')['looseKeys'], isNot(contains('ひと')));
  });

  test('levels, grades and radicals label the kanji; level and grade rank '
      'them: JLPT level, then school grade, then frequency', () {
    expect(entry('亜')['labels'], ['jlpt:n1', 'grade:secondary', 'radical:7']);
    // KANJIDIC2's classical radical, not Nelson's.
    expect(entry('唖')['labels'], ['radical:30']);
    final ranks = [
      for (final kanji in ['一', '食', '亜', '唖']) entry(kanji)['rank']! as int,
    ];
    expect(ranks, orderedEquals([...ranks]..sort()));
  });

  test('a character with no Japanese reading and no English meaning is '
      'left out', () {
    expect(lines.skip(1).where((line) => line['id'] == '僲'), isEmpty);
  });

  test('hiragana and katakana follow, the kana table each, before every '
      'kanji, found by their romaji', () {
    final kana = [
      for (final line in lines.skip(1))
        if ((line['labels']! as List).any(['hiragana', 'katakana'].contains))
          line,
    ];
    expect(kana, hasLength(142));
    expect(kana.first['id'], 'あ');
    expect(kana[71]['id'], 'ア');
    expect(kana.last['id'], 'ポ');
    expect(entry('ア')['labels'], ['katakana']);
    expect(entry('ア')['keys'], containsAll(['ア', 'あ', 'a']));
    expect(fields('ぢ'), {'romaji': 'ji', 'script': 'hiragana'});
    // を is "wo" in the table, found by "o" too.
    expect(fields('を')['romaji'], 'wo');
    expect(entry('を')['keys'], containsAll(['wo', 'o']));
    final lastKana = kana
        .map((line) => line['rank']! as int)
        .reduce((a, b) => a > b ? a : b);
    expect(entry('食')['rank']! as int, greaterThan(lastKana));
  });
}