WIP: add kanjivg data

search: let radical search API take RadkfileRadical inputs
const_data/radicals: store both formal and informal variant of radical
2026-06-04 00:54:03 +09:00 · 2026-06-03 20:18:02 +09:00 · 2026-06-03 17:42:09 +09:00 · 2026-06-03 17:01:53 +09:00 · 2026-06-03 17:01:17 +09:00 · 2026-06-02 03:19:52 +09:00
29 changed files with 1194 additions and 325 deletions
@@ -16,15 +16,14 @@ jobs:
      uses: https://github.com/cachix/install-nix-action@v31
      with:
        extra_nix_config: |
-          experimental-features = nix-command flakes
          show-trace = true
          max-jobs = auto
          trusted-users = root
          experimental-features = nix-command flakes
          build-users-group =

-    - name: Update database inputs
-      run: nix flake update datasources
+    # - name: Update database inputs
+    #   run: nix flake update datasources

    - name: Build database
      run: nix build .#database -L
@@ -7,11 +7,11 @@
        ]
      },
      "locked": {
-        "lastModified": 1776081209,
-        "narHash": "sha256-zR1115tcOPnYLk6NznSf7YslyaJLc/MGayEHShitx18=",
+        "lastModified": 1780302182,
+        "narHash": "sha256-IfC+dpdjjlkzrWlm+p851T43GsR04wMAPqGn63jisJ4=",
        "ref": "refs/heads/main",
-        "rev": "7fe3552bb16e1d315c0b27b243e5eb53cd9e86fc",
-        "revCount": 13,
+        "rev": "c116674dd1e0b879660e6237e54904aa825d4511",
+        "revCount": 29,
        "type": "git",
        "url": "https://git.pvv.ntnu.no/Mugiten/datasources.git"
      },
@@ -20,13 +20,29 @@
        "url": "https://git.pvv.ntnu.no/Mugiten/datasources.git"
      }
    },
+    "kanjivg-src": {
+      "flake": false,
+      "locked": {
+        "lastModified": 1775218066,
+        "narHash": "sha256-iYv9xakgoGt/JwwdKDUCpSAF36hBtKlX9oN7xiLowjs=",
+        "ref": "refs/heads/master",
+        "rev": "544d319f79348c092d567b662f27f33dacfa60cd",
+        "revCount": 2215,
+        "type": "git",
+        "url": "https://git.pvv.ntnu.no/mugiten/kanjivg.git"
+      },
+      "original": {
+        "type": "git",
+        "url": "https://git.pvv.ntnu.no/mugiten/kanjivg.git"
+      }
+    },
    "nixpkgs": {
      "locked": {
-        "lastModified": 1775423009,
-        "narHash": "sha256-vPKLpjhIVWdDrfiUM8atW6YkIggCEKdSAlJPzzhkQlw=",
+        "lastModified": 1779560665,
+        "narHash": "sha256-tpyBcxPpcQb8ukyNF7DoCwfSY3VPsxHoYwj00Cayv5o=",
        "owner": "NixOS",
        "repo": "nixpkgs",
-        "rev": "68d8aa3d661f0e6bd5862291b5bb263b2a6595c9",
+        "rev": "64c08a7ca051951c8eae34e3e3cb1e202fe36786",
        "type": "github"
      },
      "original": {
@@ -38,6 +54,7 @@
    "root": {
      "inputs": {
        "datasources": "datasources",
+        "kanjivg-src": "kanjivg-src",
        "nixpkgs": "nixpkgs"
      }
    }
@@ -8,12 +8,18 @@
      url = "git+https://git.pvv.ntnu.no/Mugiten/datasources.git";
      inputs.nixpkgs.follows = "nixpkgs";
    };
+
+    kanjivg-src = {
+      url = "git+https://git.pvv.ntnu.no/mugiten/kanjivg.git";
+      flake = false;
+    };
  };

  outputs = {
    self,
    nixpkgs,
    datasources,
+    kanjivg-src,
  }: let
    inherit (nixpkgs) lib;
    systems = [
@@ -45,9 +45,7 @@ class KanjiElement extends Element {
  });

  @override
-  Map<String, Object?> get sqlValue => {
-    ...super.sqlValue,
-  };
+  Map<String, Object?> get sqlValue => {...super.sqlValue};
 }

 class ReadingElement extends Element {
@@ -129,6 +127,19 @@ class XRefParts {
    'readingRef': readingRef,
    'senseOrderNum': senseOrderNum,
  };
+
+  @override
+  bool operator ==(Object other) {
+    if (identical(this, other)) return true;
+
+    return other is XRefParts &&
+        other.kanjiRef == kanjiRef &&
+        other.readingRef == readingRef &&
+        other.senseOrderNum == senseOrderNum;
+  }
+
+  @override
+  int get hashCode => Object.hash(kanjiRef, readingRef, senseOrderNum);
 }

 class XRef {
@@ -168,9 +179,7 @@ class Sense extends SQLWritable {
  });

  @override
-  Map<String, Object?> get sqlValue => {
-    'senseId': senseId,
-  };
+  Map<String, Object?> get sqlValue => {'senseId': senseId};

  bool get isEmpty =>
      antonyms.isEmpty &&
@@ -15,6 +15,12 @@ class ResolvedXref {
  const ResolvedXref(this.entry, this.ambiguous);
 }

+// A constant map of xref parts to jmdict id for unresolvable xrefs.
+final xrefExceptions = {
+  // NOTE: see https://www.edrdg.org/jmwsgi/entr.py?svc=jmdict&g=2870981.1~2369718 for details
+  XRefParts(kanjiRef: 'プレストレスト', readingRef: 'コンクリート'): 2472380,
+};
+
 /// Resolves an xref (pair of kanji, optionally reading, and optionally sense number) to an a specific
 /// JMdict entry, if possible.
 ///
@@ -27,15 +33,65 @@ ResolvedXref resolveXref(
  SplayTreeMap<String, Set<Entry>> entriesByReading,
  XRefParts xref,
 ) {
-  List<Entry> candidateEntries = switch ((xref.kanjiRef, xref.readingRef)) {
-    (null, null) => throw Exception(
-      'Xref $xref has no kanji or reading reference',
-    ),
-    (final String k, null) => entriesByKanji[k]!.toList(),
-    (null, final String r) => entriesByReading[r]!.toList(),
-    (final String k, final String r) =>
-      entriesByKanji[k]!.intersection(entriesByReading[r]!).toList(),
-  };
+  late List<Entry> candidateEntries;
+
+  if (xrefExceptions.containsKey(xref)) {
+    final exceptionEntryId = xrefExceptions[xref]!;
+    // NOTE: this is slow, but we have few exceptions. Let's wait for JMdict XML-NG to be released so we can delete this :)
+    final exceptionEntry =
+        entriesByKanji.values
+            .expand((set) => set)
+            .firstWhereOrNull((entry) => entry.entryId == exceptionEntryId) ??
+        entriesByReading.values
+            .expand((set) => set)
+            .firstWhereOrNull((entry) => entry.entryId == exceptionEntryId);
+
+    if (exceptionEntry != null) {
+      return ResolvedXref(exceptionEntry, false);
+    } else {
+      throw Exception(
+        'Xref $xref matches an exception entry ID $exceptionEntryId, but that entry was not found among the candidates.',
+      );
+    }
+  }
+
+  switch ((xref.kanjiRef, xref.readingRef)) {
+    case (null, null):
+      throw Exception('Xref $xref has no kanji or reading reference');
+
+    case (final String k, null):
+      if (!entriesByKanji.containsKey(k)) {
+        throw Exception(
+          'Xref $xref has kanji reference "$k" but no entries found with that kanji',
+        );
+      }
+      candidateEntries = entriesByKanji[k]!.toList();
+      break;
+
+    case (null, final String r):
+      if (!entriesByReading.containsKey(r)) {
+        throw Exception(
+          'Xref $xref has reading reference "$r" but no entries found with that reading',
+        );
+      }
+      candidateEntries = entriesByReading[r]!.toList();
+      break;
+
+    case (final String k, final String r):
+      if (!entriesByKanji.containsKey(k)) {
+        throw Exception(
+          'Xref $xref has kanji reference "$k" but no entries found with that kanji',
+        );
+      }
+      if (!entriesByReading.containsKey(r)) {
+        throw Exception(
+          'Xref $xref has reading reference "$r" but no entries found with that reading',
+        );
+      }
+      candidateEntries = entriesByKanji[k]!
+          .intersection(entriesByReading[r]!)
+          .toList();
+  }

  // Filter out entries that don't have the number of senses specified in the xref
  if (xref.senseOrderNum != null) {
@@ -0,0 +1,92 @@
+import 'package:jadb/_data_ingestion/sql_writable.dart';
+
+/// Enum set in the kvg:position attribute, used by `<g>` elements in the KanjiVG SVG files.
+enum KanjiPathGroupPosition {
+  bottom,
+  kamae,
+  kamaec,
+  left,
+  middle,
+  nyo,
+  nyoc,
+  right,
+  tare,
+  tarec,
+  top,
+}
+
+/// Contents of a \<g> element in the KanjiVG SVG files.
+class KanjiPathGroupTreeNode extends SQLWritable {
+  final String id;
+  final List<KanjiPathGroupTreeNode> children;
+  final String? element;
+  final String? original;
+  final KanjiPathGroupPosition? position;
+  final String? radical;
+  final int? part;
+
+  KanjiPathGroupTreeNode({
+    required this.id,
+    this.children = const [],
+    this.element,
+    this.original,
+    this.position,
+    this.radical,
+    this.part,
+  });
+
+  @override
+  Map<String, Object?> get sqlValue => {
+    'id': id,
+    'element': element,
+    'original': original,
+    'position': position?.name,
+    'radical': radical,
+    'part': part,
+  };
+}
+
+/// Contents of a `<text>` element in the StrokeNumber's group in the KanjiVG SVG files
+class KanjiStrokeNumber extends SQLWritable {
+  final int num;
+  final double x;
+  final double y;
+
+  KanjiStrokeNumber(this.num, this.x, this.y);
+
+  @override
+  Map<String, Object?> get sqlValue => {'num': num, 'x': x, 'y': y};
+}
+
+/// Contents of a `<path>` element in the KanjiVG SVG files
+class KanjiVGPath extends SQLWritable {
+  final String id;
+  final String type;
+  final String svgPath;
+
+  KanjiVGPath({required this.id, required this.type, required this.svgPath});
+
+  @override
+  Map<String, Object?> get sqlValue => {
+    'id': id,
+    'type': type,
+    'svgPath': svgPath,
+  };
+}
+
+class KanjiVGItem extends SQLWritable {
+  final String character;
+  final List<KanjiVGPath> paths;
+  final List<KanjiStrokeNumber> strokeNumbers;
+  final List<KanjiPathGroupTreeNode> pathGroups;
+
+  KanjiVGItem({
+    required this.character,
+    required this.paths,
+    required this.strokeNumbers,
+    required this.pathGroups,
+  });
+
+  @override
+  Map<String, Object?> get sqlValue => {'character': character};
+}
@@ -0,0 +1,7 @@
+import 'package:sqflite_common/sqflite.dart';
+
+Future<void> seedKanjiVGData(Iterable<String> xmlContents, Database db) async {
+  final b = db.batch();
+
+  await b.commit(noResult: true);
+}
@@ -1,217 +1,310 @@
-const Map<int, List<String>> radicals = {
-  1: ['一', '｜', '丶', 'ノ', '乙', '亅'],
+import 'package:collection/collection.dart';
+
+class RadkfileRadical {
+  /// The formal variant of the radical
+  ///
+  /// This is the one you want to use for searching.
+  final String formalVariant;
+
+  /// The informal variant of the radical (if it differs from the formal one).
+  ///
+  /// This is the one you should display to the user.
+  final String? informalVariant;
+
+  /// The number of strokes in this radical.
+  final int strokeCount;
+
+  const RadkfileRadical({
+    required this.formalVariant,
+    this.informalVariant,
+    required this.strokeCount,
+  });
+}
+
+const Map<int, List<RadkfileRadical>> radicals = {
+  1: [
+    RadkfileRadical(formalVariant: '一', strokeCount: 1),
+    RadkfileRadical(formalVariant: '｜', strokeCount: 1),
+    RadkfileRadical(formalVariant: '丶', strokeCount: 1),
+    RadkfileRadical(formalVariant: 'ノ', strokeCount: 1),
+    RadkfileRadical(formalVariant: '乙', strokeCount: 1),
+    RadkfileRadical(formalVariant: '亅', strokeCount: 1),
+  ],
  2: [
-    '二',
-    '亠',
-    '人',
-    '⺅',
-    '𠆢',
-    '儿',
-    '入',
-    'ハ',
-    '丷',
-    '冂',
-    '冖',
-    '冫',
-    '几',
-    '凵',
-    '刀',
-    '⺉',
-    '力',
-    '勹',
-    '匕',
-    '匚',
-    '十',
-    '卜',
-    '卩',
-    '厂',
-    '厶',
-    '又',
-    'マ',
-    '九',
-    'ユ',
-    '乃',
-    '𠂉',
+    RadkfileRadical(formalVariant: '二', strokeCount: 2),
+    RadkfileRadical(formalVariant: '亠', strokeCount: 2),
+    RadkfileRadical(formalVariant: '人', strokeCount: 2),
+    RadkfileRadical(formalVariant: '化', informalVariant: '⺅', strokeCount: 2),
+    RadkfileRadical(formalVariant: '个', informalVariant: '𠆢', strokeCount: 2),
+    RadkfileRadical(formalVariant: '儿', strokeCount: 2),
+    RadkfileRadical(formalVariant: '入', strokeCount: 2),
+    RadkfileRadical(formalVariant: 'ハ', strokeCount: 2),
+    RadkfileRadical(formalVariant: '并', strokeCount: 2),
+    RadkfileRadical(formalVariant: '冂', strokeCount: 2),
+    RadkfileRadical(formalVariant: '冖', strokeCount: 2),
+    RadkfileRadical(formalVariant: '冫', strokeCount: 2),
+    RadkfileRadical(formalVariant: '几', strokeCount: 2),
+    RadkfileRadical(formalVariant: '凵', strokeCount: 2),
+    RadkfileRadical(formalVariant: '刀', strokeCount: 2),
+    RadkfileRadical(formalVariant: '刈', strokeCount: 2),
+    RadkfileRadical(formalVariant: '力', strokeCount: 2),
+    RadkfileRadical(formalVariant: '勹', strokeCount: 2),
+    RadkfileRadical(formalVariant: '匕', strokeCount: 2),
+    RadkfileRadical(formalVariant: '匚', strokeCount: 2),
+    RadkfileRadical(formalVariant: '十', strokeCount: 2),
+    RadkfileRadical(formalVariant: '卜', strokeCount: 2),
+    RadkfileRadical(formalVariant: '卩', strokeCount: 2),
+    RadkfileRadical(formalVariant: '厂', strokeCount: 2),
+    RadkfileRadical(formalVariant: '厶', strokeCount: 2),
+    RadkfileRadical(formalVariant: '又', strokeCount: 2),
+    RadkfileRadical(formalVariant: 'マ', strokeCount: 2),
+    RadkfileRadical(formalVariant: '九', strokeCount: 2),
+    RadkfileRadical(formalVariant: 'ユ', strokeCount: 2),
+    RadkfileRadical(formalVariant: '乃', strokeCount: 2),
+    RadkfileRadical(formalVariant: '乞', strokeCount: 2),
  ],
  3: [
-    '⻌',
-    '口',
-    '囗',
-    '土',
-    '士',
-    '夂',
-    '夕',
-    '大',
-    '女',
-    '子',
-    '宀',
-    '寸',
-    '小',
-    '⺌',
-    '尢',
-    '尸',
-    '屮',
-    '山',
-    '川',
-    '巛',
-    '工',
-    '已',
-    '巾',
-    '干',
-    '幺',
-    '广',
-    '廴',
-    '廾',
-    '弋',
-    '弓',
-    'ヨ',
-    '彑',
-    '彡',
-    '彳',
-    '⺖',
-    '⺘',
-    '⺡',
-    '⺨',
-    '⺾',
-    '⻏',
-    '⻖',
-    '也',
-    '亡',
-    '及',
-    '久',
+    RadkfileRadical(formalVariant: '込', informalVariant: '⻌', strokeCount: 3),
+    RadkfileRadical(formalVariant: '口', strokeCount: 3),
+    RadkfileRadical(formalVariant: '囗', strokeCount: 3),
+    RadkfileRadical(formalVariant: '土', strokeCount: 3),
+    RadkfileRadical(formalVariant: '士', strokeCount: 3),
+    RadkfileRadical(formalVariant: '夂', strokeCount: 3),
+    RadkfileRadical(formalVariant: '夕', strokeCount: 3),
+    RadkfileRadical(formalVariant: '大', strokeCount: 3),
+    RadkfileRadical(formalVariant: '女', strokeCount: 3),
+    RadkfileRadical(formalVariant: '子', strokeCount: 3),
+    RadkfileRadical(formalVariant: '宀', strokeCount: 3),
+    RadkfileRadical(formalVariant: '寸', strokeCount: 3),
+    RadkfileRadical(formalVariant: '小', strokeCount: 3),
+    RadkfileRadical(formalVariant: '尚', informalVariant: '⺌', strokeCount: 3),
+    RadkfileRadical(formalVariant: '尢', strokeCount: 3),
+    RadkfileRadical(formalVariant: '尸', strokeCount: 3),
+    RadkfileRadical(formalVariant: '屮', strokeCount: 3),
+    RadkfileRadical(formalVariant: '山', strokeCount: 3),
+    RadkfileRadical(formalVariant: '川', strokeCount: 3),
+    RadkfileRadical(formalVariant: '巛', strokeCount: 3),
+    RadkfileRadical(formalVariant: '工', strokeCount: 3),
+    RadkfileRadical(formalVariant: '已', strokeCount: 3),
+    RadkfileRadical(formalVariant: '巾', strokeCount: 3),
+    RadkfileRadical(formalVariant: '干', strokeCount: 3),
+    RadkfileRadical(formalVariant: '幺', strokeCount: 3),
+    RadkfileRadical(formalVariant: '广', strokeCount: 3),
+    RadkfileRadical(formalVariant: '廴', strokeCount: 3),
+    RadkfileRadical(formalVariant: '廾', strokeCount: 3),
+    RadkfileRadical(formalVariant: '弋', strokeCount: 3),
+    RadkfileRadical(formalVariant: '弓', strokeCount: 3),
+    RadkfileRadical(formalVariant: 'ヨ', strokeCount: 3),
+    RadkfileRadical(formalVariant: '彑', strokeCount: 3),
+    RadkfileRadical(formalVariant: '彡', strokeCount: 3),
+    RadkfileRadical(formalVariant: '彳', strokeCount: 3),
+    RadkfileRadical(formalVariant: '忙', informalVariant: '⺖', strokeCount: 3),
+    RadkfileRadical(formalVariant: '扎', informalVariant: '⺘', strokeCount: 3),
+    RadkfileRadical(formalVariant: '汁', informalVariant: '⺡', strokeCount: 3),
+    RadkfileRadical(formalVariant: '犯', informalVariant: '⺨', strokeCount: 3),
+    RadkfileRadical(formalVariant: '艾', informalVariant: '⺾', strokeCount: 3),
+    RadkfileRadical(formalVariant: '邦', informalVariant: '⻏', strokeCount: 3),
+    RadkfileRadical(formalVariant: '阡', informalVariant: '⻖', strokeCount: 3),
+    RadkfileRadical(formalVariant: '也', strokeCount: 3),
+    RadkfileRadical(formalVariant: '亡', strokeCount: 3),
+    RadkfileRadical(formalVariant: '及', strokeCount: 3),
+    RadkfileRadical(formalVariant: '久', strokeCount: 3),
  ],
  4: [
-    '⺹',
-    '心',
-    '戈',
-    '戸',
-    '手',
-    '支',
-    '攵',
-    '文',
-    '斗',
-    '斤',
-    '方',
-    '无',
-    '日',
-    '曰',
-    '月',
-    '木',
-    '欠',
-    '止',
-    '歹',
-    '殳',
-    '比',
-    '毛',
-    '氏',
-    '气',
-    '水',
-    '火',
-    '⺣',
-    '爪',
-    '父',
-    '爻',
-    '爿',
-    '片',
-    '牛',
-    '犬',
-    '⺭',
-    '王',
-    '元',
-    '井',
-    '勿',
-    '尤',
-    '五',
-    '屯',
-    '巴',
-    '毋',
+    RadkfileRadical(formalVariant: '老', informalVariant: '⺹', strokeCount: 4),
+    RadkfileRadical(formalVariant: '心', strokeCount: 4),
+    RadkfileRadical(formalVariant: '戈', strokeCount: 4),
+    RadkfileRadical(formalVariant: '戸', strokeCount: 4),
+    RadkfileRadical(formalVariant: '手', strokeCount: 4),
+    RadkfileRadical(formalVariant: '支', strokeCount: 4),
+    RadkfileRadical(formalVariant: '攵', strokeCount: 4),
+    RadkfileRadical(formalVariant: '文', strokeCount: 4),
+    RadkfileRadical(formalVariant: '斗', strokeCount: 4),
+    RadkfileRadical(formalVariant: '斤', strokeCount: 4),
+    RadkfileRadical(formalVariant: '方', strokeCount: 4),
+    RadkfileRadical(formalVariant: '无', strokeCount: 4),
+    RadkfileRadical(formalVariant: '日', strokeCount: 4),
+    RadkfileRadical(formalVariant: '曰', strokeCount: 4),
+    RadkfileRadical(formalVariant: '月', strokeCount: 4),
+    RadkfileRadical(formalVariant: '木', strokeCount: 4),
+    RadkfileRadical(formalVariant: '欠', strokeCount: 4),
+    RadkfileRadical(formalVariant: '止', strokeCount: 4),
+    RadkfileRadical(formalVariant: '歹', strokeCount: 4),
+    RadkfileRadical(formalVariant: '殳', strokeCount: 4),
+    RadkfileRadical(formalVariant: '比', strokeCount: 4),
+    RadkfileRadical(formalVariant: '毛', strokeCount: 4),
+    RadkfileRadical(formalVariant: '氏', strokeCount: 4),
+    RadkfileRadical(formalVariant: '气', strokeCount: 4),
+    RadkfileRadical(formalVariant: '水', strokeCount: 4),
+    RadkfileRadical(formalVariant: '火', strokeCount: 4),
+    RadkfileRadical(formalVariant: '杰', strokeCount: 4),
+    RadkfileRadical(formalVariant: '爪', strokeCount: 4),
+    RadkfileRadical(formalVariant: '父', strokeCount: 4),
+    RadkfileRadical(formalVariant: '爻', strokeCount: 4),
+    RadkfileRadical(formalVariant: '爿', strokeCount: 4),
+    RadkfileRadical(formalVariant: '片', strokeCount: 4),
+    RadkfileRadical(formalVariant: '牛', strokeCount: 4),
+    RadkfileRadical(formalVariant: '犬', strokeCount: 4),
+    RadkfileRadical(formalVariant: '礼', strokeCount: 4),
+    RadkfileRadical(formalVariant: '王', strokeCount: 4),
+    RadkfileRadical(formalVariant: '元', strokeCount: 4),
+    RadkfileRadical(formalVariant: '井', strokeCount: 4),
+    RadkfileRadical(formalVariant: '勿', strokeCount: 4),
+    RadkfileRadical(formalVariant: '尤', strokeCount: 4),
+    RadkfileRadical(formalVariant: '五', strokeCount: 4),
+    RadkfileRadical(formalVariant: '屯', strokeCount: 4),
+    RadkfileRadical(formalVariant: '巴', strokeCount: 4),
+    RadkfileRadical(formalVariant: '毋', strokeCount: 4),
  ],
  5: [
-    '玄',
-    '瓦',
-    '甘',
-    '生',
-    '用',
-    '田',
-    '疋',
-    '疒',
-    '癶',
-    '白',
-    '皮',
-    '皿',
-    '目',
-    '矛',
-    '矢',
-    '石',
-    '示',
-    '禸',
-    '禾',
-    '穴',
-    '立',
-    '⻂',
-    '世',
-    '巨',
-    '冊',
-    '母',
-    '⺲',
-    '牙',
+    RadkfileRadical(formalVariant: '玄', strokeCount: 5),
+    RadkfileRadical(formalVariant: '瓦', strokeCount: 5),
+    RadkfileRadical(formalVariant: '甘', strokeCount: 5),
+    RadkfileRadical(formalVariant: '生', strokeCount: 5),
+    RadkfileRadical(formalVariant: '用', strokeCount: 5),
+    RadkfileRadical(formalVariant: '田', strokeCount: 5),
+    RadkfileRadical(formalVariant: '疋', strokeCount: 5),
+    RadkfileRadical(formalVariant: '疔', strokeCount: 5),
+    RadkfileRadical(formalVariant: '癶', strokeCount: 5),
+    RadkfileRadical(formalVariant: '白', strokeCount: 5),
+    RadkfileRadical(formalVariant: '皮', strokeCount: 5),
+    RadkfileRadical(formalVariant: '皿', strokeCount: 5),
+    RadkfileRadical(formalVariant: '目', strokeCount: 5),
+    RadkfileRadical(formalVariant: '矛', strokeCount: 5),
+    RadkfileRadical(formalVariant: '矢', strokeCount: 5),
+    RadkfileRadical(formalVariant: '石', strokeCount: 5),
+    RadkfileRadical(formalVariant: '示', strokeCount: 5),
+    RadkfileRadical(formalVariant: '禹', strokeCount: 5),
+    RadkfileRadical(formalVariant: '禾', strokeCount: 5),
+    RadkfileRadical(formalVariant: '穴', strokeCount: 5),
+    RadkfileRadical(formalVariant: '立', strokeCount: 5),
+    RadkfileRadical(formalVariant: '初', strokeCount: 5),
+    RadkfileRadical(formalVariant: '世', strokeCount: 5),
+    RadkfileRadical(formalVariant: '巨', strokeCount: 5),
+    RadkfileRadical(formalVariant: '冊', strokeCount: 5),
+    RadkfileRadical(formalVariant: '母', strokeCount: 5),
+    RadkfileRadical(formalVariant: '買', strokeCount: 5),
+    RadkfileRadical(formalVariant: '牙', strokeCount: 5),
  ],
  6: [
-    '瓜',
-    '竹',
-    '米',
-    '糸',
-    '缶',
-    '羊',
-    '羽',
-    '而',
-    '耒',
-    '耳',
-    '聿',
-    '肉',
-    '自',
-    '至',
-    '臼',
-    '舌',
-    '舟',
-    '艮',
-    '色',
-    '虍',
-    '虫',
-    '血',
-    '行',
-    '衣',
-    '西',
+    RadkfileRadical(formalVariant: '瓜', strokeCount: 6),
+    RadkfileRadical(formalVariant: '竹', strokeCount: 6),
+    RadkfileRadical(formalVariant: '米', strokeCount: 6),
+    RadkfileRadical(formalVariant: '糸', strokeCount: 6),
+    RadkfileRadical(formalVariant: '缶', strokeCount: 6),
+    RadkfileRadical(formalVariant: '羊', strokeCount: 6),
+    RadkfileRadical(formalVariant: '羽', strokeCount: 6),
+    RadkfileRadical(formalVariant: '而', strokeCount: 6),
+    RadkfileRadical(formalVariant: '耒', strokeCount: 6),
+    RadkfileRadical(formalVariant: '耳', strokeCount: 6),
+    RadkfileRadical(formalVariant: '聿', strokeCount: 6),
+    RadkfileRadical(formalVariant: '肉', strokeCount: 6),
+    RadkfileRadical(formalVariant: '自', strokeCount: 6),
+    RadkfileRadical(formalVariant: '至', strokeCount: 6),
+    RadkfileRadical(formalVariant: '臼', strokeCount: 6),
+    RadkfileRadical(formalVariant: '舌', strokeCount: 6),
+    RadkfileRadical(formalVariant: '舟', strokeCount: 6),
+    RadkfileRadical(formalVariant: '艮', strokeCount: 6),
+    RadkfileRadical(formalVariant: '色', strokeCount: 6),
+    RadkfileRadical(formalVariant: '虍', strokeCount: 6),
+    RadkfileRadical(formalVariant: '虫', strokeCount: 6),
+    RadkfileRadical(formalVariant: '血', strokeCount: 6),
+    RadkfileRadical(formalVariant: '行', strokeCount: 6),
+    RadkfileRadical(formalVariant: '衣', strokeCount: 6),
+    RadkfileRadical(formalVariant: '西', strokeCount: 6),
  ],
  7: [
-    '臣',
-    '見',
-    '角',
-    '言',
-    '谷',
-    '豆',
-    '豕',
-    '豸',
-    '貝',
-    '赤',
-    '走',
-    '足',
-    '身',
-    '車',
-    '辛',
-    '辰',
-    '酉',
-    '釆',
-    '里',
-    '舛',
-    '麦',
+    RadkfileRadical(formalVariant: '臣', strokeCount: 7),
+    RadkfileRadical(formalVariant: '見', strokeCount: 7),
+    RadkfileRadical(formalVariant: '角', strokeCount: 7),
+    RadkfileRadical(formalVariant: '言', strokeCount: 7),
+    RadkfileRadical(formalVariant: '谷', strokeCount: 7),
+    RadkfileRadical(formalVariant: '豆', strokeCount: 7),
+    RadkfileRadical(formalVariant: '豕', strokeCount: 7),
+    RadkfileRadical(formalVariant: '豸', strokeCount: 7),
+    RadkfileRadical(formalVariant: '貝', strokeCount: 7),
+    RadkfileRadical(formalVariant: '赤', strokeCount: 7),
+    RadkfileRadical(formalVariant: '走', strokeCount: 7),
+    RadkfileRadical(formalVariant: '足', strokeCount: 7),
+    RadkfileRadical(formalVariant: '身', strokeCount: 7),
+    RadkfileRadical(formalVariant: '車', strokeCount: 7),
+    RadkfileRadical(formalVariant: '辛', strokeCount: 7),
+    RadkfileRadical(formalVariant: '辰', strokeCount: 7),
+    RadkfileRadical(formalVariant: '酉', strokeCount: 7),
+    RadkfileRadical(formalVariant: '釆', strokeCount: 7),
+    RadkfileRadical(formalVariant: '里', strokeCount: 7),
+    RadkfileRadical(formalVariant: '舛', strokeCount: 7),
+    RadkfileRadical(formalVariant: '麦', strokeCount: 7),
  ],
-  8: ['金', '長', '門', '隶', '隹', '雨', '青', '非', '奄', '岡', '免', '斉'],
-  9: ['面', '革', '韭', '音', '頁', '風', '飛', '食', '首', '香', '品'],
-  10: ['馬', '骨', '高', '髟', '鬥', '鬯', '鬲', '鬼', '竜', '韋'],
-  11: ['魚', '鳥', '鹵', '鹿', '麻', '亀', '啇', '黄', '黒'],
-  12: ['黍', '黹', '無', '歯'],
-  13: ['黽', '鼎', '鼓', '鼠'],
-  14: ['鼻', '齊'],
-  17: ['龠'],
+  8: [
+    RadkfileRadical(formalVariant: '金', strokeCount: 8),
+    RadkfileRadical(formalVariant: '長', strokeCount: 8),
+    RadkfileRadical(formalVariant: '門', strokeCount: 8),
+    RadkfileRadical(formalVariant: '隶', strokeCount: 8),
+    RadkfileRadical(formalVariant: '隹', strokeCount: 8),
+    RadkfileRadical(formalVariant: '雨', strokeCount: 8),
+    RadkfileRadical(formalVariant: '青', strokeCount: 8),
+    RadkfileRadical(formalVariant: '非', strokeCount: 8),
+    RadkfileRadical(formalVariant: '奄', strokeCount: 8),
+    RadkfileRadical(formalVariant: '岡', strokeCount: 8),
+    RadkfileRadical(formalVariant: '免', strokeCount: 8),
+    RadkfileRadical(formalVariant: '斉', strokeCount: 8),
+  ],
+  9: [
+    RadkfileRadical(formalVariant: '面', strokeCount: 9),
+    RadkfileRadical(formalVariant: '革', strokeCount: 9),
+    RadkfileRadical(formalVariant: '韭', strokeCount: 9),
+    RadkfileRadical(formalVariant: '音', strokeCount: 9),
+    RadkfileRadical(formalVariant: '頁', strokeCount: 9),
+    RadkfileRadical(formalVariant: '風', strokeCount: 9),
+    RadkfileRadical(formalVariant: '飛', strokeCount: 9),
+    RadkfileRadical(formalVariant: '食', strokeCount: 9),
+    RadkfileRadical(formalVariant: '首', strokeCount: 9),
+    RadkfileRadical(formalVariant: '香', strokeCount: 9),
+    RadkfileRadical(formalVariant: '品', strokeCount: 9),
+  ],
+  10: [
+    RadkfileRadical(formalVariant: '馬', strokeCount: 10),
+    RadkfileRadical(formalVariant: '骨', strokeCount: 10),
+    RadkfileRadical(formalVariant: '高', strokeCount: 10),
+    RadkfileRadical(formalVariant: '髟', strokeCount: 10),
+    RadkfileRadical(formalVariant: '鬥', strokeCount: 10),
+    RadkfileRadical(formalVariant: '鬯', strokeCount: 10),
+    RadkfileRadical(formalVariant: '鬲', strokeCount: 10),
+    RadkfileRadical(formalVariant: '鬼', strokeCount: 10),
+    RadkfileRadical(formalVariant: '竜', strokeCount: 10),
+    RadkfileRadical(formalVariant: '韋', strokeCount: 10),
+  ],
+  11: [
+    RadkfileRadical(formalVariant: '魚', strokeCount: 11),
+    RadkfileRadical(formalVariant: '鳥', strokeCount: 11),
+    RadkfileRadical(formalVariant: '鹵', strokeCount: 11),
+    RadkfileRadical(formalVariant: '鹿', strokeCount: 11),
+    RadkfileRadical(formalVariant: '麻', strokeCount: 11),
+    RadkfileRadical(formalVariant: '亀', strokeCount: 11),
+    RadkfileRadical(formalVariant: '滴', strokeCount: 11),
+    RadkfileRadical(formalVariant: '黄', strokeCount: 11),
+    RadkfileRadical(formalVariant: '黒', strokeCount: 11),
+  ],
+  12: [
+    RadkfileRadical(formalVariant: '黍', strokeCount: 12),
+    RadkfileRadical(formalVariant: '黹', strokeCount: 12),
+    RadkfileRadical(formalVariant: '無', strokeCount: 12),
+    RadkfileRadical(formalVariant: '歯', strokeCount: 12),
+  ],
+  13: [
+    RadkfileRadical(formalVariant: '黽', strokeCount: 13),
+    RadkfileRadical(formalVariant: '鼎', strokeCount: 13),
+    RadkfileRadical(formalVariant: '鼓', strokeCount: 13),
+    RadkfileRadical(formalVariant: '鼠', strokeCount: 13),
+  ],
+  14: [
+    RadkfileRadical(formalVariant: '鼻', strokeCount: 14),
+    RadkfileRadical(formalVariant: '齊', strokeCount: 14),
+  ],
+  17: [RadkfileRadical(formalVariant: '龠', strokeCount: 17)],
+};
+
+final Map<String, RadkfileRadical> radicalsByFormalVariant = {
+  for (final r in radicals.values.flattened) r.formalVariant: r,
 };
@@ -1,5 +1,6 @@
 import 'package:jadb/table_names/jmdict.dart';
 import 'package:jadb/table_names/kanjidic.dart';
+import 'package:jadb/table_names/kanjivg.dart';
 import 'package:jadb/table_names/radkfile.dart';
 import 'package:jadb/table_names/tanos_jlpt.dart';
 import 'package:sqflite_common/sqlite_api.dart';
@@ -21,6 +22,7 @@ Future<void> verifyTablesWithDbConnection(DatabaseExecutor db) async {
    ...KANJIDICTableNames.allTables,
    ...RADKFILETableNames.allTables,
    ...TanosJLPTTableNames.allTables,
+    ...KanjiVGTableNames.allTables,
  };

  final missingTables = expectedTables.difference(tables);
@@ -33,6 +33,7 @@ class WordSearchMatchSpan {
  Map<String, Object?> toJson() => {
    'spanType': spanType.toString().split('.').last,
    'index': index,
+    'subIndex': subIndex,
    'start': start,
    'end': end,
  };
@@ -43,12 +44,13 @@ class WordSearchMatchSpan {
          (e) => e.toString().split('.').last == json['spanType'],
        ),
        index: json['index'] as int,
+        subIndex: json['subIndex'] as int? ?? 0,
        start: json['start'] as int,
        end: json['end'] as int,
      );

  @override
-  int get hashCode => Object.hash(spanType, index, start, end);
+  int get hashCode => Object.hash(spanType, index, subIndex, start, end);

  @override
  bool operator ==(Object other) {
@@ -56,6 +58,7 @@ class WordSearchMatchSpan {
    return other is WordSearchMatchSpan &&
        other.spanType == spanType &&
        other.index == index &&
+        other.subIndex == subIndex &&
        other.start == start &&
        other.end == end;
  }
@@ -81,7 +81,7 @@ class WordSearchSense {
    'dialects': dialects.map((e) => e.toJson()).toList(),
    'misc': misc.map((e) => e.toJson()).toList(),
    'info': info,
-    'languageSource': languageSource,
+    'languageSource': languageSource.map((e) => e.toJson()).toList(),
  };

  factory WordSearchSense.fromJson(Map<String, dynamic> json) =>
@@ -16,6 +16,8 @@ class WordSearchXrefEntry {
  final bool ambiguous;

  /// The result of the cross-reference, may or may not be included in the query.
+  ///
+  /// Be careful not to introduce circular references when using this field.
  final WordSearchResult? xrefResult;

  const WordSearchXrefEntry({
@@ -40,6 +42,10 @@ class WordSearchXrefEntry {
        ambiguous: json['ambiguous'] as bool,
        baseWord: json['baseWord'] as String,
        furigana: json['furigana'] as String?,
-        xrefResult: null,
+        xrefResult: json['xrefResult'] != null
+            ? WordSearchResult.fromJson(
+                Map<String, dynamic>.from(json['xrefResult'] as Map),
+              )
+            : null,
      );
 }
@@ -1,9 +1,11 @@
+import 'package:jadb/const_data/radicals.dart';
 import 'package:jadb/models/kanji_search/kanji_search_result.dart';
 import 'package:jadb/models/verify_tables.dart';
 import 'package:jadb/models/word_search/word_search_result.dart';
 import 'package:jadb/search/filter_kanji.dart';
 import 'package:jadb/search/kanji_search.dart';
 import 'package:jadb/search/radical_search.dart';
+import 'package:jadb/search/versions.dart';
 import 'package:jadb/search/word_search/word_search.dart';
 import 'package:sqflite_common/sqlite_api.dart';

@@ -18,12 +20,13 @@ extension JaDBConnection on DatabaseExecutor {
      searchKanjiWithDbConnection(this, kanji);

  /// Search for a kanji in the database.
-  Future<Map<String, KanjiSearchResult>> jadbGetManyKanji(Set<String> kanji) =>
-      searchManyKanjiWithDbConnection(this, kanji);
+  Future<Map<String, KanjiSearchResult>> jadbGetManyKanji(
+    Iterable<String> kanji,
+  ) => searchManyKanjiWithDbConnection(this, kanji);

  /// Filter a list of characters, and return the ones that are listed in the kanji dictionary.
  Future<List<String>> filterKanji(
-    List<String> kanji, {
+    Iterable<String> kanji, {
    bool deduplicate = false,
  }) => filterKanjiWithDbConnection(this, kanji, deduplicate);

@@ -61,11 +64,17 @@ extension JaDBConnection on DatabaseExecutor {
  /// of the radicals, find their other radicals, and return those.
  /// This is used to figure out which remaining combinations of radicals
  /// the user can search for without getting zero results.
-  Future<List<String>> jadbSearchRemainingRadicals(List<String> radicals) =>
-      searchRemainingRadicalsWithDbConnection(this, radicals);
+  Future<List<RadkfileRadical>> jadbSearchRemainingRadicals(
+    List<RadkfileRadical> radicals,
+  ) => searchRemainingRadicalsWithDbConnection(this, radicals);

  /// Given a list of radicals, search which kanji contains all
  /// of the radicals, and return those.
-  Future<List<String>> jadbSearchKanjiByRadicals(List<String> radicals) =>
-      searchKanjiByRadicalsWithDbConnection(this, radicals);
+  Future<List<String>> jadbSearchKanjiByRadicals(
+    List<RadkfileRadical> radicals,
+  ) => searchKanjiByRadicalsWithDbConnection(this, radicals);
+
+  /// Retrieve the version information for all datasources in the database.
+  Future<DatasourceVersions> jadbGetDatasourceVersions() =>
+      getDatasourceVersions(this);
 }
@@ -6,7 +6,7 @@ import 'package:sqflite_common/sqflite.dart';
 /// If [deduplicate] is true, the returned list will deduplicate the input kanji list before returning the filtered results.
 Future<List<String>> filterKanjiWithDbConnection(
  DatabaseExecutor connection,
-  List<String> kanji,
+  Iterable<String> kanji,
  bool deduplicate,
 ) async {
  final Set<String> filteredKanji = await connection
@@ -14,7 +14,7 @@ Future<List<String>> filterKanjiWithDbConnection(
      SELECT "literal"
      FROM "${KANJIDICTableNames.character}"
      WHERE "literal" IN (${kanji.map((_) => '?').join(',')})
-    ''', kanji)
+    ''', kanji.toList())
      .then((value) => value.map((e) => e['literal'] as String).toSet());

  if (deduplicate) {
@@ -274,7 +274,7 @@ Future<KanjiSearchResult?> searchKanjiWithDbConnection(
 /// Searches for multiple kanji at once, returning a map of kanji to their search results.
 Future<Map<String, KanjiSearchResult>> searchManyKanjiWithDbConnection(
  DatabaseExecutor connection,
-  Set<String> kanji,
+  Iterable<String> kanji,
 ) async {
  if (kanji.isEmpty) {
    return {};
@@ -1,3 +1,4 @@
+import 'package:jadb/const_data/radicals.dart';
 import 'package:jadb/table_names/radkfile.dart';
 import 'package:sqflite_common/sqlite_api.dart';

@@ -7,9 +8,9 @@ import 'package:sqflite_common/sqlite_api.dart';
 ///
 /// This can be used to limit the choices of additional radicals provided to a user,
 /// so that any choice they make will still yield at least one kanji.
-Future<List<String>> searchRemainingRadicalsWithDbConnection(
+Future<List<RadkfileRadical>> searchRemainingRadicalsWithDbConnection(
  DatabaseExecutor connection,
-  List<String> radicals,
+  List<RadkfileRadical> radicals,
 ) async {
  final distinctRadicals = radicals.toSet();

@@ -25,11 +26,14 @@ Future<List<String>> searchRemainingRadicalsWithDbConnection(
        HAVING COUNT(DISTINCT "radical") = ?
      )
    ''',
-    [...distinctRadicals, distinctRadicals.length],
+    [...distinctRadicals.map((r) => r.formalVariant), distinctRadicals.length],
  );

  final remainingRadicals = queryResult
      .map((row) => row['radical'] as String)
+      // TODO: maybe we should do some runtime checking here, just to throw a sensible error
+      //       if something goes horribly wrong?
+      .map((r) => radicalsByFormalVariant[r]!)
      .toList();

  return remainingRadicals;
@@ -38,7 +42,7 @@ Future<List<String>> searchRemainingRadicalsWithDbConnection(
 /// Returns a list of kanji that contain all of the input radicals.
 Future<List<String>> searchKanjiByRadicalsWithDbConnection(
  DatabaseExecutor connection,
-  List<String> radicals,
+  List<RadkfileRadical> radicals,
 ) async {
  final distinctRadicals = radicals.toSet();

@@ -50,7 +54,7 @@ Future<List<String>> searchKanjiByRadicalsWithDbConnection(
      GROUP BY "kanji"
      HAVING COUNT(DISTINCT "radical") = ?
    ''',
-    [...distinctRadicals, distinctRadicals.length],
+    [...distinctRadicals.map((r) => r.formalVariant), distinctRadicals.length],
  );

  final kanji = queryResult.map((row) => row['kanji'] as String).toList();
@@ -0,0 +1,84 @@
+import 'package:jadb/table_names/jmdict.dart';
+import 'package:jadb/table_names/kanjidic.dart';
+import 'package:jadb/table_names/radkfile.dart';
+import 'package:jadb/table_names/tanos_jlpt.dart';
+import 'package:sqflite_common/sqlite_api.dart';
+
+class DatasourceVersions {
+  final String jmdictVersion;
+  final DateTime jmdictDate;
+  final String jmdictHash;
+
+  final String kanjidic2Version;
+  final DateTime kanjidic2Date;
+  final String kanjidic2Hash;
+
+  final String radkfileVersion;
+  final DateTime radkfileDate;
+  final String radkfileHash;
+
+  final String tanosJlptVersion;
+  final DateTime tanosJlptDate;
+  final String tanosJlptHash;
+
+  const DatasourceVersions({
+    required this.jmdictVersion,
+    required this.jmdictDate,
+    required this.jmdictHash,
+    required this.kanjidic2Version,
+    required this.kanjidic2Date,
+    required this.kanjidic2Hash,
+    required this.radkfileVersion,
+    required this.radkfileDate,
+    required this.radkfileHash,
+    required this.tanosJlptVersion,
+    required this.tanosJlptDate,
+    required this.tanosJlptHash,
+  });
+}
+
+DateTime _parseDateTime(String dateString) {
+  try {
+    return DateTime.parse(dateString);
+  } catch (e) {
+    if (RegExp(r'^\d{4}-\d{2}$').hasMatch(dateString)) {
+      return DateTime.parse('$dateString-01');
+    } else if (RegExp(r'^\d{4}$').hasMatch(dateString)) {
+      return DateTime.parse('$dateString-01-01');
+    } else {
+      throw FormatException('Invalid date format: $dateString');
+    }
+  }
+}
+
+Future<DatasourceVersions> getDatasourceVersions(
+  final DatabaseExecutor connection,
+) async {
+  final jmdictVersion = await connection
+      .query(JMdictTableNames.version)
+      .then((rows) => rows.first);
+  final kanjidic2Version = await connection
+      .query(KANJIDICTableNames.version)
+      .then((rows) => rows.first);
+  final radkfileVersion = await connection
+      .query(RADKFILETableNames.version)
+      .then((rows) => rows.first);
+  final tanosJlptVersion = await connection
+      .query(TanosJLPTTableNames.version)
+      .then((rows) => rows.first);
+
+  return DatasourceVersions(
+    jmdictVersion: jmdictVersion['version'] as String,
+    jmdictDate: _parseDateTime(jmdictVersion['date'].toString()),
+    jmdictHash: jmdictVersion['hash'] as String,
+    kanjidic2Version: kanjidic2Version['version'] as String,
+    kanjidic2Date: _parseDateTime(kanjidic2Version['date'].toString()),
+    kanjidic2Hash: kanjidic2Version['hash'] as String,
+    radkfileVersion: radkfileVersion['version'] as String,
+    radkfileDate: _parseDateTime(radkfileVersion['date'].toString()),
+    radkfileHash: radkfileVersion['hash'] as String,
+    tanosJlptVersion: tanosJlptVersion['version'] as String,
+    tanosJlptDate: _parseDateTime(tanosJlptVersion['date'].toString()),
+    tanosJlptHash: tanosJlptVersion['hash'] as String,
+  );
+}
@@ -0,0 +1,9 @@
+abstract class KanjiVGTableNames {
+  static const String version = 'KanjiVG_Version';
+  static const String entry = 'KanjiVG_Entry';
+  static const String path = 'KanjiVG_Path';
+  static const String strokeNumber = 'KanjiVG_StrokeNumber';
+  static const String pathGroup = 'KanjiVG_PathGroup';
+
+  static Set<String> get allTables => {version, entry, path, strokeNumber, pathGroup};
+}
@@ -0,0 +1 @@
+const int jadbSchemaVersion = 1;
@@ -229,7 +229,7 @@ CREATE TABLE "JMdict_SenseGlossary" (
  PRIMARY KEY ("senseId", "phrase")
 ) WITHOUT ROWID;

-- CREATE INDEX "JMdict_SenseGlossary_byPhrase" ON JMdict_SenseGlossary("phrase");
+CREATE INDEX "JMdict_SenseGlossary_byPhrase" ON JMdict_SenseGlossary("phrase");

 CREATE TABLE "JMdict_SenseGlossaryType" (
  "senseId" INTEGER NOT NULL REFERENCES "JMdict_Sense"("senseId"),
@@ -1,6 +1,6 @@
 CREATE TABLE "JMdict_EntryScore" (
  "elementId" INTEGER PRIMARY KEY,
-  "score" INTEGER NOT NULL DEFAULT 0,
+  "score" INTEGER NOT NULL,
  "common" BOOLEAN NOT NULL DEFAULT FALSE,

  "entryId" INTEGER NOT NULL GENERATED ALWAYS AS (("elementId" / 100) % 10000000) STORED,
@@ -15,8 +15,7 @@ CREATE TABLE "JMdict_EntryScore" (
 CREATE INDEX "JMdict_EntryScore_byElementId_byScore" ON "JMdict_EntryScore"("elementId", "score");
 CREATE INDEX "JMdict_EntryScore_byScore" ON "JMdict_EntryScore"("score");

-CREATE INDEX "JMdict_EntryScore_byElementId_byCommon" ON "JMdict_EntryScore"("elementId", "common");
-CREATE INDEX "JMdict_EntryScore_byCommon" ON "JMdict_EntryScore"("common");
+CREATE INDEX "JMdict_EntryScore_byCommon" ON "JMdict_EntryScore"("common") WHERE "common" = TRUE;

 -- NOTE: these views are deduplicated in order not to perform an unnecessary
 --       UNION on every trigger
@@ -0,0 +1,45 @@
+CREATE TABLE "KanjiVG_Version" (
+  "version" VARCHAR(10) PRIMARY KEY NOT NULL,
+  "date" DATE NOT NULL,
+  "hash" VARCHAR(64) NOT NULL
+) WITHOUT ROWID;
+
+CREATE TRIGGER "KanjiVG_Version_SingleRow"
+BEFORE INSERT ON "KanjiVG_Version"
+WHEN (SELECT COUNT(*) FROM "KanjiVG_Version") >= 1
+BEGIN
+    SELECT RAISE(FAIL, 'Only one row allowed in KanjiVG_Version');
+END;
+
+CREATE TABLE "KanjiVG_Entry" (
+  "character" CHAR(1) PRIMARY KEY NOT NULL
+) WITHOUT ROWID;
+
+CREATE TABLE "KanjiVG_StrokeNumber" (
+  "character" CHAR(1) NOT NULL REFERENCES "KanjiVG_Entry"("character"),
+  "strokeNum" INTEGER NOT NULL,
+  "x" REAL NOT NULL,
+  "y" REAL NOT NULL,
+  PRIMARY KEY ("character", "strokeNum")
+) WITHOUT ROWID;
+
+CREATE TABLE "KanjiVG_Path" (
+  "character" CHAR(1) NOT NULL REFERENCES "KanjiVG_Entry"("character"),
+  "pathId" TEXT NOT NULL,
+  "type" VARCHAR(10) NOT NULL,
+  "svgPath" TEXT NOT NULL,
+  PRIMARY KEY ("character", "pathId")
+) WITHOUT ROWID;
+
+CREATE TABLE "KanjiVG_PathGroup" (
+  "character" CHAR(1) NOT NULL REFERENCES "KanjiVG_Entry"("character"),
+  "groupId" TEXT NOT NULL,
+  "parentGroupId" TEXT REFERENCES "KanjiVG_PathGroup"("groupId"),
+  "element" TEXT,
+  "original" TEXT,
+  "position" VARCHAR(10),
+  "radical" TEXT,
+  "part" INTEGER,
+  PRIMARY KEY ("character", "groupId"),
+  CHECK ("position" IN ('bottom', 'kamae', 'kamaec', 'left', 'middle', 'nyo', 'nyoc', 'right', 'tare', 'tarec', 'top') OR "position" IS NULL)
+) WITHOUT ROWID;
@@ -5,18 +5,18 @@ packages:
    dependency: transitive
    description:
      name: _fe_analyzer_shared
-      sha256: "8d718c5c58904f9937290fd5dbf2d6a0e02456867706bfb6cd7b81d394e738d5"
+      sha256: cd6add6f846f35fb79f3c315296703c1a24f3cfd7f4739d91a74961c1c7e9f1b
      url: "https://pub.dev"
    source: hosted
-    version: "98.0.0"
+    version: "100.0.0"
  analyzer:
    dependency: transitive
    description:
      name: analyzer
-      sha256: "6141ad5d092d1e1d13929c0504658bbeccc1703505830d7c26e859908f5efc88"
+      sha256: "6ba98576948803398b69e3a444df24eacdbe12ed699c7014e120ea38552debbf"
      url: "https://pub.dev"
    source: hosted
-    version: "12.0.0"
+    version: "13.0.0"
  args:
    dependency: "direct main"
    description:
@@ -61,10 +61,10 @@ packages:
    dependency: transitive
    description:
      name: code_assets
-      sha256: "83ccdaa064c980b5596c35dd64a8d3ecc68620174ab9b90b6343b753aa721687"
+      sha256: bf394f466ba9205f1812a0433b392d6af280f155f56651eda7c18cc32ed493b8
      url: "https://pub.dev"
    source: hosted
-    version: "1.0.0"
+    version: "1.2.1"
  collection:
    dependency: "direct main"
    description:
@@ -149,10 +149,10 @@ packages:
    dependency: transitive
    description:
      name: hooks
-      sha256: e79ed1e8e1929bc6ecb6ec85f0cb519c887aa5b423705ded0d0f2d9226def388
+      sha256: "62ae9bb76d02526c7c2110a19b6e6ad788fe28d35e553e35efb02a41a46ab43a"
      url: "https://pub.dev"
    source: hosted
-    version: "1.0.2"
+    version: "2.0.1"
  http_multi_server:
    dependency: transitive
    description:
@@ -197,10 +197,10 @@ packages:
    dependency: transitive
    description:
      name: matcher
-      sha256: dc0b7dc7651697ea4ff3e69ef44b0407ea32c487a39fff6a4004fa585e901861
+      sha256: "31bd099b47c10cd1aeb55146a2d46ce0277630ecef3f7dae54ad7873f36696cd"
      url: "https://pub.dev"
    source: hosted
-    version: "0.12.19"
+    version: "0.12.20"
  meta:
    dependency: transitive
    description:
@@ -221,10 +221,10 @@ packages:
    dependency: transitive
    description:
      name: native_toolchain_c
-      sha256: "6ba77bb18063eebe9de401f5e6437e95e1438af0a87a3a39084fbd37c90df572"
+      sha256: f59351d28f49520cd3a74eb1f41c5f19ae15e53c65a3231d14af672e46510a96
      url: "https://pub.dev"
    source: hosted
-    version: "0.17.6"
+    version: "0.19.1"
  node_preamble:
    dependency: transitive
    description:
@@ -273,6 +273,14 @@ packages:
      url: "https://pub.dev"
    source: hosted
    version: "2.2.0"
+  record_use:
+    dependency: transitive
+    description:
+      name: record_use
+      sha256: "2551bd8eecfe95d14ae75f6021ad0248be5c27f138c2ec12fcb52b500b3ba1ed"
+      url: "https://pub.dev"
+    source: hosted
+    version: "0.6.0"
  shelf:
    dependency: transitive
    description:
@@ -333,26 +341,26 @@ packages:
    dependency: "direct main"
    description:
      name: sqflite_common
-      sha256: "6ef422a4525ecc601db6c0a2233ff448c731307906e92cabc9ba292afaae16a6"
+      sha256: "1581ffbf7a0e333b380d6a30737d78516b826cb35beb7fb0bf8a3ea0c678b465"
      url: "https://pub.dev"
    source: hosted
-    version: "2.5.6"
+    version: "2.5.8"
  sqflite_common_ffi:
    dependency: "direct main"
    description:
      name: sqflite_common_ffi
-      sha256: c59fcdc143839a77581f7a7c4de018e53682408903a0a0800b95ef2dc4033eff
+      sha256: cd0c7f7de39a08f2d54ef144d9058c46eca8461879aaa648025643455c1e5a20
      url: "https://pub.dev"
    source: hosted
-    version: "2.4.0+2"
+    version: "2.4.0+3"
  sqlite3:
    dependency: "direct main"
    description:
      name: sqlite3
-      sha256: caa693ad15a587a2b4fde093b728131a1827903872171089dedb16f7665d3a91
+      sha256: "9488c7d2cdb1091c91cacf7e207cff81b28bff8e366f042bad3afe7d34afe189"
      url: "https://pub.dev"
    source: hosted
-    version: "3.2.0"
+    version: "3.3.2"
  stack_trace:
    dependency: transitive
    description:
@@ -381,10 +389,10 @@ packages:
    dependency: transitive
    description:
      name: synchronized
-      sha256: c254ade258ec8282947a0acbbc90b9575b4f19673533ee46f2f6e9b3aeefd7c0
+      sha256: "63896c27e81b28f8cb4e69ead0d3e8f03f1d1e5fc531a3e579cabed6a2c7c9e5"
      url: "https://pub.dev"
    source: hosted
-    version: "3.4.0"
+    version: "3.4.0+1"
  term_glyph:
    dependency: transitive
    description:
@@ -397,26 +405,26 @@ packages:
    dependency: "direct dev"
    description:
      name: test
-      sha256: "8d9ceddbab833f180fbefed08afa76d7c03513dfdba87ffcec2718b02bbcbf20"
+      sha256: ca578dc12bb8b2f40b67b7d3bd2fac4f31c01a6ff7130a14e2597b919934507f
      url: "https://pub.dev"
    source: hosted
-    version: "1.31.0"
+    version: "1.31.1"
  test_api:
    dependency: transitive
    description:
      name: test_api
-      sha256: "949a932224383300f01be9221c39180316445ecb8e7547f70a41a35bf421fb9e"
+      sha256: "2a122cbe059f8b610d3a5415f42e255b6c17b1f21eee1d960f31080237fb4f11"
      url: "https://pub.dev"
    source: hosted
-    version: "0.7.11"
+    version: "0.7.12"
  test_core:
    dependency: transitive
    description:
      name: test_core
-      sha256: "1991d4cfe85d5043241acac92962c3977c8d2f2add1ee73130c7b286417d1d34"
+      sha256: d2e98ec12998368dc59ddd47ab709f2cd55acd6b66dc7db764455a44082f4bc5
      url: "https://pub.dev"
    source: hosted
-    version: "0.6.17"
+    version: "0.6.18"
  typed_data:
    dependency: transitive
    description:
@@ -429,10 +437,10 @@ packages:
    dependency: transitive
    description:
      name: vm_service
-      sha256: "45caa6c5917fa127b5dbcfbd1fa60b14e583afdc08bfc96dda38886ca252eb60"
+      sha256: "0016aef94fc66495ac78af5859181e3f3bf2026bd8eecc72b9565601e19ab360"
      url: "https://pub.dev"
    source: hosted
-    version: "15.0.2"
+    version: "15.2.0"
  watcher:
    dependency: transitive
    description:
@@ -477,10 +485,10 @@ packages:
    dependency: "direct main"
    description:
      name: xml
-      sha256: "971043b3a0d3da28727e40ed3e0b5d18b742fa5a68665cca88e74b7876d5e025"
+      sha256: "67f0aff7be013d107995e9b75bf4e7f2c3ef2dfdb2c8e68024bba0a7fd5756a4"
      url: "https://pub.dev"
    source: hosted
-    version: "6.6.1"
+    version: "7.0.1"
  yaml:
    dependency: transitive
    description:
@@ -490,4 +498,4 @@ packages:
    source: hosted
    version: "3.1.3"
 sdks:
-  dart: ">=3.10.1 <4.0.0"
+  dart: ">=3.11.0 <4.0.0"
@@ -15,7 +15,7 @@ dependencies:
  sqflite_common: ^2.5.0
  sqflite_common_ffi: ^2.3.0
  sqlite3: ^3.1.6
-  xml: ^6.5.0
+  xml: '>=6.0.0 < 8.0.0'

 dev_dependencies:
  benchmark_harness: ^2.4.0
@@ -0,0 +1,93 @@
+import 'dart:convert';
+
+import 'package:jadb/models/kanji_search/kanji_search_radical.dart';
+import 'package:jadb/models/kanji_search/kanji_search_result.dart';
+import 'package:test/test.dart';
+
+Object? _roundTripJson(Object? value) => jsonDecode(jsonEncode(value));
+
+Map<String, dynamic> _roundTripMap(Object? json) =>
+    Map<String, dynamic>.from(_roundTripJson(json) as Map);
+
+void main() {
+  test('KanjiSearchRadical JSON serialization roundtrip', () {
+    const radical = KanjiSearchRadical(
+      symbol: '人',
+      names: ['ひと', 'にんべん'],
+      forms: ['亻'],
+      meanings: ['person', 'human'],
+    );
+
+    final restored = KanjiSearchRadical.fromJson(
+      _roundTripMap(radical.toJson()),
+    );
+
+    expect(restored, equals(radical));
+  });
+
+  test('KanjiSearchResult JSON serialization roundtrip', () {
+    const result = KanjiSearchResult(
+      kanji: '休',
+      taughtIn: 1,
+      jlptLevel: 'N5',
+      newspaperFrequencyRank: 1234,
+      strokeCount: 6,
+      meanings: ['rest', 'day off'],
+      kunyomi: ['やす.む', 'やす.まる'],
+      onyomi: ['キュウ'],
+      radical: KanjiSearchRadical(
+        symbol: '人',
+        names: ['ひと', 'にんべん'],
+        forms: ['亻'],
+        meanings: ['person', 'human'],
+      ),
+      parts: ['亻', '木'],
+      codepoints: {'ucs': '4F11', 'jis208': '1-22-57'},
+      nanori: ['やす'],
+      alternativeLanguageReadings: {
+        'korean': ['휴'],
+        'pinyin': ['xiū'],
+      },
+      strokeMiscounts: [5, 7],
+      queryCodes: {
+        'skip': ['1-2-4'],
+        'fourCorner': ['2429.0'],
+      },
+      dictionaryReferences: {'nelson_c': '122', 'heisig': '457'},
+    );
+
+    final restored = KanjiSearchResult.fromJson(_roundTripMap(result.toJson()));
+
+    expect(restored, equals(result));
+  });
+
+  test(
+    'KanjiSearchResult JSON serialization roundtrip - nullable and empty fields',
+    () {
+      const result = KanjiSearchResult(
+        kanji: '々',
+        taughtIn: null,
+        jlptLevel: null,
+        newspaperFrequencyRank: null,
+        strokeCount: 3,
+        meanings: ['iteration mark'],
+        kunyomi: [],
+        onyomi: [],
+        radical: null,
+        parts: [],
+        codepoints: {'ucs': '3005'},
+        nanori: [],
+        alternativeLanguageReadings: {},
+        strokeMiscounts: [],
+        queryCodes: {},
+        dictionaryReferences: {},
+      );
+
+      final restored = KanjiSearchResult.fromJson(
+        _roundTripMap(result.toJson()),
+      );
+
+      expect(restored, equals(result));
+    },
+  );
+}
@@ -0,0 +1,282 @@
+import 'dart:convert';
+
+import 'package:jadb/models/common/jlpt_level.dart';
+import 'package:jadb/models/jmdict/jmdict_dialect.dart';
+import 'package:jadb/models/jmdict/jmdict_field.dart';
+import 'package:jadb/models/jmdict/jmdict_kanji_info.dart';
+import 'package:jadb/models/jmdict/jmdict_misc.dart';
+import 'package:jadb/models/jmdict/jmdict_pos.dart';
+import 'package:jadb/models/jmdict/jmdict_reading_info.dart';
+import 'package:jadb/models/word_search/word_search_match_span.dart';
+import 'package:jadb/models/word_search/word_search_result.dart';
+import 'package:jadb/models/word_search/word_search_ruby.dart';
+import 'package:jadb/models/word_search/word_search_sense.dart';
+import 'package:jadb/models/word_search/word_search_sense_language_source.dart';
+import 'package:jadb/models/word_search/word_search_sources.dart';
+import 'package:jadb/models/word_search/word_search_xref_entry.dart';
+import 'package:test/test.dart';
+
+Object? _roundTripJson(Object? value) => jsonDecode(jsonEncode(value));
+
+Map<String, dynamic> _roundTripMap(Object? json) =>
+    Map<String, dynamic>.from(_roundTripJson(json) as Map);
+
+Map<String, Object?> _roundTripObjectMap(Object? json) =>
+    Map<String, Object?>.from(_roundTripJson(json) as Map);
+
+void _expectScalarRoundTrip<T>({
+  required Iterable<T> values,
+  required Object? Function(T value) toJson,
+  required T Function(Object? json) fromJson,
+}) {
+  for (final value in values) {
+    expect(
+      fromJson(_roundTripJson(toJson(value))),
+      equals(value),
+      reason: 'Roundtrip failed for $value',
+    );
+  }
+}
+
+void _expectMapRoundTrip<T>({
+  required Iterable<T> values,
+  required Map<String, Object?> Function(T value) toJson,
+  required T Function(Map<String, Object?> json) fromJson,
+}) {
+  for (final value in values) {
+    expect(
+      fromJson(_roundTripObjectMap(toJson(value))),
+      equals(value),
+      reason: 'Roundtrip failed for $value',
+    );
+  }
+}
+
+WordSearchResult _buildNestedXrefResult() => WordSearchResult(
+  score: 7,
+  entryId: 300,
+  isCommon: false,
+  japanese: [WordSearchRuby(base: '補助', furigana: 'ほじょ')],
+  kanjiInfo: const {},
+  readingInfo: const {},
+  senses: const [],
+  jlptLevel: JlptLevel.none,
+  sources: const WordSearchSources(jmdict: true, jmnedict: false),
+);
+
+WordSearchSense _buildSense() => WordSearchSense(
+  englishDefinitions: ['kana', 'syllabary'],
+  partsOfSpeech: [JMdictPOS.n],
+  seeAlso: [
+    WordSearchXrefEntry(
+      entryId: 300,
+      ambiguous: false,
+      baseWord: '仮名遣い',
+      furigana: 'かなづかい',
+      xrefResult: _buildNestedXrefResult(),
+    ),
+  ],
+  antonyms: const [
+    WordSearchXrefEntry(
+      entryId: 301,
+      ambiguous: true,
+      baseWord: '漢字',
+      furigana: 'かんじ',
+      xrefResult: null,
+    ),
+  ],
+  restrictedToReading: ['かな'],
+  restrictedToKanji: ['仮名'],
+  fields: [JMdictField.linguistics],
+  dialects: [JMdictDialect.kansai],
+  misc: [JMdictMisc.onlyKana, JMdictMisc.rare],
+  info: ['Typically written using kana alone.'],
+  languageSource: const [
+    WordSearchSenseLanguageSource(
+      language: 'por',
+      phrase: 'canoa',
+      fullyDescribesSense: false,
+      constructedFromSmallerWords: true,
+    ),
+  ],
+);
+
+WordSearchResult _buildWordSearchResult({
+  List<WordSearchMatchSpan>? matchSpans,
+}) => WordSearchResult(
+  score: 42,
+  entryId: 123,
+  isCommon: true,
+  japanese: [
+    WordSearchRuby(base: '仮名', furigana: 'かな'),
+    WordSearchRuby(base: 'かな'),
+  ],
+  kanjiInfo: {'仮名': JMdictKanjiInfo.rK, '仮': JMdictKanjiInfo.ateji},
+  readingInfo: {'かな': JMdictReadingInfo.gikun, 'カナ': JMdictReadingInfo.rk},
+  senses: [_buildSense()],
+  jlptLevel: JlptLevel.n5,
+  sources: const WordSearchSources(jmdict: true, jmnedict: true),
+  matchSpans: matchSpans,
+);
+
+void main() {
+  test('JlptLevel JSON serialization roundtrip', () {
+    _expectScalarRoundTrip<JlptLevel>(
+      values: JlptLevel.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JlptLevel.fromJson,
+    );
+  });
+
+  test('JMdictDialect JSON serialization roundtrip', () {
+    _expectMapRoundTrip<JMdictDialect>(
+      values: JMdictDialect.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JMdictDialect.fromJson,
+    );
+  });
+
+  test('JMdictField JSON serialization roundtrip', () {
+    _expectMapRoundTrip<JMdictField>(
+      values: JMdictField.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JMdictField.fromJson,
+    );
+  });
+
+  test('JMdictKanjiInfo JSON serialization roundtrip', () {
+    _expectMapRoundTrip<JMdictKanjiInfo>(
+      values: JMdictKanjiInfo.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JMdictKanjiInfo.fromJson,
+    );
+  });
+
+  test('JMdictMisc JSON serialization roundtrip', () {
+    _expectMapRoundTrip<JMdictMisc>(
+      values: JMdictMisc.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JMdictMisc.fromJson,
+    );
+  });
+
+  test('JMdictPOS JSON serialization roundtrip', () {
+    _expectMapRoundTrip<JMdictPOS>(
+      values: JMdictPOS.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JMdictPOS.fromJson,
+    );
+  });
+
+  test('JMdictReadingInfo JSON serialization roundtrip', () {
+    _expectMapRoundTrip<JMdictReadingInfo>(
+      values: JMdictReadingInfo.values,
+      toJson: (value) => value.toJson(),
+      fromJson: JMdictReadingInfo.fromJson,
+    );
+  });
+
+  test('WordSearchMatchSpan JSON serialization roundtrip', () {
+    final span = WordSearchMatchSpan(
+      spanType: WordSearchMatchSpanType.sense,
+      index: 2,
+      subIndex: 3,
+      start: 4,
+      end: 8,
+    );
+
+    final restored = WordSearchMatchSpan.fromJson(_roundTripMap(span.toJson()));
+
+    expect(restored, equals(span));
+    expect(restored.subIndex, equals(span.subIndex));
+  });
+
+  test('WordSearchRuby JSON serialization roundtrip', () {
+    final ruby = WordSearchRuby(base: '仮名', furigana: 'かな');
+
+    final restored = WordSearchRuby.fromJson(_roundTripMap(ruby.toJson()));
+
+    expect(restored.toJson(), equals(ruby.toJson()));
+  });
+
+  test('WordSearchSenseLanguageSource JSON serialization roundtrip', () {
+    const source = WordSearchSenseLanguageSource(
+      language: 'por',
+      phrase: 'canoa',
+      fullyDescribesSense: false,
+      constructedFromSmallerWords: true,
+    );
+
+    final restored = WordSearchSenseLanguageSource.fromJson(
+      _roundTripMap(source.toJson()),
+    );
+
+    expect(restored.toJson(), equals(source.toJson()));
+  });
+
+  test('WordSearchSources JSON serialization roundtrip', () {
+    const sources = WordSearchSources(jmdict: false, jmnedict: true);
+
+    final restored = WordSearchSources.fromJson(
+      _roundTripMap(sources.toJson()),
+    );
+
+    expect(restored.toJson(), equals(sources.toJson()));
+    expect(restored.sqlValue, equals(sources.sqlValue));
+  });
+
+  test('WordSearchXrefEntry JSON serialization roundtrip', () {
+    final entry = WordSearchXrefEntry(
+      entryId: 300,
+      ambiguous: false,
+      baseWord: '仮名遣い',
+      furigana: 'かなづかい',
+      xrefResult: _buildNestedXrefResult(),
+    );
+
+    final restored = WordSearchXrefEntry.fromJson(
+      _roundTripMap(entry.toJson()),
+    );
+
+    expect(restored.toJson(), equals(entry.toJson()));
+    expect(restored.xrefResult?.toJson(), equals(entry.xrefResult?.toJson()));
+  });
+
+  test('WordSearchSense JSON serialization roundtrip', () {
+    final sense = _buildSense();
+
+    final restored = WordSearchSense.fromJson(_roundTripMap(sense.toJson()));
+
+    expect(restored.toJson(), equals(sense.toJson()));
+  });
+
+  test('WordSearchResult JSON serialization roundtrip', () {
+    final result = _buildWordSearchResult();
+
+    final restored = WordSearchResult.fromJson(_roundTripMap(result.toJson()));
+
+    expect(restored.toJson(), equals(result.toJson()));
+    expect(restored.matchSpans, isNull);
+  });
+
+  test('WordSearchResult leaves matchSpans out of JSON', () {
+    final result = _buildWordSearchResult(
+      matchSpans: [
+        WordSearchMatchSpan(
+          spanType: WordSearchMatchSpanType.sense,
+          index: 0,
+          subIndex: 1,
+          start: 0,
+          end: 4,
+        ),
+      ],
+    );
+
+    final json = _roundTripMap(result.toJson());
+    final restored = WordSearchResult.fromJson(json);
+
+    expect(json.containsKey('matchSpans'), isFalse);
+    expect(restored.matchSpans, isNull);
+    expect(restored.toJson(), equals(result.toJson()));
+  });
+}
@@ -0,0 +1,12 @@
+import 'package:jadb/search.dart';
+import 'package:test/test.dart';
+
+import 'setup_database_connection.dart';
+
+void main() {
+  test('Retrieve datasource versions', () async {
+    final connection = await setupDatabaseConnection();
+    final result = await connection.jadbGetDatasourceVersions();
+    expect(result, isNotNull);
+  });
+}
@@ -7,32 +7,29 @@ import 'package:test/test.dart';
 import 'setup_database_connection.dart';

 void main() {
-  test(
-    'All constant radicals should exist in the database',
-    () async {
-      final connection = await setupDatabaseConnection();
-      final allRadicalsInDb = await connection.query(
-        RADKFILETableNames.radkfile,
-        columns: ['radical'],
-        distinct: true,
-      );
+  test('All constant radicals should exist in the database', () async {
+    final connection = await setupDatabaseConnection();
+    final allRadicalsInDb = await connection.query(
+      RADKFILETableNames.radkfile,
+      columns: ['radical'],
+      distinct: true,
+    );

-      final radicalsInDb = allRadicalsInDb
-          .map((e) => e['radical'] as String)
-          .toSet();
+    final radicalsInDb = allRadicalsInDb
+        .map((e) => e['radical'] as String)
+        .toSet();

-      final missingRadicals = radicals.values.flattenedToSet.difference(
-        radicalsInDb,
-      );
+    final missingRadicals = radicals.values.flattenedToSet
+        .map((e) => e.formalVariant)
+        .toSet()
+        .difference(radicalsInDb);

-      expect(
-        missingRadicals,
-        isEmpty,
-        reason: 'Missing radicals in database: $missingRadicals',
-      );
-    },
-    skip: 'Test is valid, code is broken, fix me',
-  );
+    expect(
+      missingRadicals,
+      isEmpty,
+      reason: 'Missing radicals in database: $missingRadicals',
+    );
+  });

  test(
    'All radicals in database should be in the constant radical list',
@@ -49,7 +46,7 @@ void main() {
          .toSet();

      final extraRadicals = radicalsInDb.difference(
-        radicals.values.flattenedToSet,
+        radicals.values.flattenedToSet.map((e) => e.formalVariant).toSet(),
      );

      expect(
@@ -59,21 +56,30 @@ void main() {
            'Extra radicals in database missing in the constant list: $extraRadicals',
      );
    },
-    skip: 'Test is valid, code is broken, fix me',
  );

-  group(
-    'All radicals should return results',
-    () {
-      for (final radical in radicals.values.flattened) {
-        test(' - $radical', () async {
-          final connection = await setupDatabaseConnection();
-          final result = await connection.jadbSearchKanjiByRadicals([radical]);
-          expect(result, isNotEmpty);
-        });
+  test('All constant radicals are located in the correct stroke count group', () {
+    for (final mapEntry in radicals.entries) {
+      final strokeCount = mapEntry.key;
+      final radicalsInGroup = mapEntry.value;
+      for (final radical in radicalsInGroup) {
+        expect(
+          strokeCount,
+          radical.strokeCount,
+          reason:
+              'Radical ${radical.formalVariant} should have stroke count $strokeCount but has ${radical.strokeCount}',
+        );
      }
-    },
-    skip:
-        'These will be automatically fixed once the other radical tests are passing',
-  );
+    }
+  });
+
+  group('All radicals should return results', () {
+    for (final radical in radicals.values.flattened) {
+      test(' - $radical', () async {
+        final connection = await setupDatabaseConnection();
+        final result = await connection.jadbSearchKanjiByRadicals([radical]);
+        expect(result, isNotEmpty);
+      });
+    }
+  });
 }
@@ -8,7 +8,7 @@ void main() {
      expect(result, 'かたまり');
    });

-    test('Basic test with diacritics', () {
+    test('Basic test with dakuten', () {
      final result = transliterateLatinToHiragana('gadamari');
      expect(result, 'がだまり');
    });
@@ -54,7 +54,7 @@ void main() {

    test('Basic test', expectSpans('katamari', ['か', 'た', 'ま', 'り']));
    test(
-      'Basic test with diacritics',
+      'Basic test with dakuten',
      expectSpans('gadamari', ['が', 'だ', 'ま', 'り']),
    );
    test('wi and we', expectSpans('wiwe', ['うぃ', 'うぇ']));
@@ -72,7 +72,7 @@ void main() {
      expect(result, 'katamari');
    });

-    test('Basic test with diacritics', () {
+    test('Basic test with dakuten', () {
      final result = transliterateHiraganaToLatin('がだまり');
      expect(result, 'gadamari');
    });
@@ -91,6 +91,21 @@ void main() {
      final result = transliterateHiraganaToLatin('かっぱ');
      expect(result, 'kappa');
    });
+
+    test('Iteration mark', () {
+      final result = transliterateHiraganaToLatin('さゝき');
+      expect(result, 'sasaki');
+    }, skip: 'Not yet implemented');
+
+    test('Iteration mark with dakuten', () {
+      final result = transliterateHiraganaToLatin('あひゞき');
+      expect(result, 'ahibiki');
+    }, skip: 'Not yet implemented');
+
+    test('Yori', () {
+      final result = transliterateHiraganaToLatin('ゟ');
+      expect(result, 'yori');
+    }, skip: 'Not yet implemented');
  });

  group('Hiragana -> Romaji Spans', () {
@@ -110,7 +125,7 @@ void main() {

    test('Basic test', expectSpans('かたまり', ['ka', 'ta', 'ma', 'ri']));
    test(
-      'Basic test with diacritics',
+      'Basic test with dakuten',
      expectSpans('がだまり', ['ga', 'da', 'ma', 'ri']),
    );
    test('wi and we', expectSpans('うぃうぇ', ['whi', 'whe']));
@@ -118,5 +133,17 @@ void main() {

    // TODO: fix the implementation
    // test('Double consonant', expectSpans('かっぱ', ['ka', 'ppa']));
+
+    test(
+      'Iteration mark',
+      expectSpans('さゝき', ['sa', 'sa', 'ki']),
+      skip: 'Not yet implemented',
+    );
+    test(
+      'Iteration mark with dakuten',
+      expectSpans('あひゞき', ['a', 'hi', 'bi', 'ki']),
+      skip: 'Not yet implemented',
+    );
+    test('Yori', expectSpans('ゟ', ['yori']), skip: 'Not yet implemented');
  });
 }
Author	SHA1	Message	Date
oysteikt	1aea3c5874	WIP: add kanjivg data Build and test / build (push) Successful in 5m47s Details	2026-06-04 00:54:03 +09:00
oysteikt	a46d61c335	search: let radical search API take `RadkfileRadical` inputs Build and test / build (push) Successful in 5m11s Details	2026-06-03 20:18:02 +09:00
oysteikt	f3dba3ffab	const_data/radicals: store both formal and informal variant of radical Build and test / build (push) Successful in 5m47s Details	2026-06-03 17:42:09 +09:00
oysteikt	5556a73899	test/models: test json serialization roundtrips for all models Build and test / build (push) Successful in 5m17s Details	2026-06-03 17:01:53 +09:00
oysteikt	541680b95d	models/word_search: clean up json de/serialization	2026-06-03 17:01:17 +09:00
oysteikt	505380d2cc	lib/_data_ingestion: add xref exception list for jmdict Build and test / build (push) Successful in 7m30s Details	2026-06-02 03:19:52 +09:00
oysteikt	c4abcabca2	search: add api for fetching datasource versions Build and test / build (push) Failing after 3m45s Details	2026-06-02 02:56:51 +09:00
oysteikt	5629198539	version: add const schema version number Build and test / build (push) Failing after 3m55s Details	2026-06-02 02:36:59 +09:00
oysteikt	7360144136	flake.lock: bump, pubspec.{yaml,lock}: bump	2026-06-02 02:36:59 +09:00
oysteikt	c43285c78a	.gitea/workflows/build-and-test: don't autoupdate datasources Build and test / build (push) Successful in 5m6s Details	2026-05-06 00:19:59 +09:00
oysteikt	c22e96b7f9	flake.lock: bump nixpkgs Build and test / build (push) Failing after 3m54s Details	2026-05-06 00:13:32 +09:00
oysteikt	d3516495ab	data_ingestion/jmdict: throw proper errors on invalid xrefs	2026-05-05 23:48:05 +09:00
oysteikt	be4dd72dcb	migrations/JMdict: no default value for `EntryScore.score` Build and test / build (push) Failing after 4m4s Details	2026-05-05 23:23:01 +09:00
oysteikt	32ec34a150	migrations/JMdict: only index common entries in `EntryScore_byCommon`	2026-05-05 23:22:37 +09:00
oysteikt	bd0822b740	.gitea/workflows/build-and-test: remove duplicate nix config Build and test / build (push) Successful in 5m32s Details	2026-04-29 08:39:52 +09:00
oysteikt	8827893101	search: prefer `Iterable` over `List` in some public APIs Build and test / build (push) Successful in 7m51s Details	2026-04-14 17:55:40 +09:00
oysteikt	28c4403e2d	test/romaji_transliteration: add tests for iteration marks and yori ligature Build and test / build (push) Successful in 6m56s Details	2026-04-13 22:06:03 +09:00