halkyon/halkyon_learn_curator/test/mecab_test.dart

import 'dart:io';

import 'package:archive/archive.dart';
import 'package:halkyon_learn_curator/src/mecab.dart';
import 'package:test/test.dart';

void main() {
  late Directory compiled;
  late Directory into;

  setUp(() async {
    compiled = await Directory.systemTemp.createTemp('compiled');
    into = await Directory.systemTemp.createTemp('dictionaries');
    for (final name in [
      'char.bin',
      'dicrc',
      'matrix.bin',
      'sys.dic',
      'unk.dic',
    ]) {
      await File('${compiled.path}/$name').writeAsString('$name of IPADIC');
    }
  });
  tearDown(() async {
    await compiled.delete(recursive: true);
    await into.delete(recursive: true);
  });

  test('packs the files MeCab reads, the licence and an empty mecabrc into '
      'one gzipped tar beside the dictionary files', () async {
    final licence = File('${compiled.path}/copyright')
      ..writeAsStringSync('Copyright 2000, 2001, 2002, 2003 Nara Institute');

    final packed = await packIpadic(compiled, into, licence: licence);

    expect(packed?.path, '${into.path}/$ipadicFile');
    final files = {
      for (final file in TarDecoder().decodeBytes(
        const GZipDecoder().decodeBytes(packed!.readAsBytesSync()),
      ))
        file.name: String.fromCharCodes(file.content),
    };
    expect(files.keys, {
      'char.bin',
      'dicrc',
      'matrix.bin',
      'sys.dic',
      'unk.dic',
      'COPYING',
      'mecabrc',
    });
    expect(files['sys.dic'], 'sys.dic of IPADIC');
    expect(files['COPYING'], startsWith('Copyright'));
    expect(files['mecabrc'], isEmpty);
    // Nothing the server would load as a dictionary.
    expect(
      into.listSync().where(
        (file) =>
            file.path.endsWith('.jsonl') || file.path.endsWith('.jsonl.gz'),
      ),
      isEmpty,
    );
  });

  test('leaves the pack as it is when the dictionary is the same, so devices '
      'download it again only once it changed', () async {
    final first = await packIpadic(compiled, into);
    final bytes = first!.readAsBytesSync();

    expect(await packIpadic(compiled, into), isNull);
    expect(File('${into.path}/$ipadicFile').readAsBytesSync(), bytes);

    await File('${compiled.path}/sys.dic').writeAsString('a newer IPADIC');
    expect(await packIpadic(compiled, into), isNotNull);
  });

  test('refuses a dictionary missing a file MeCab reads', () async {
    await File('${compiled.path}/matrix.bin').delete();

    await expectLater(
      packIpadic(compiled, into),
      throwsA(isA<FileSystemException>()),
    );
    expect(File('${into.path}/$ipadicFile').existsSync(), isFalse);
  });
}