Feature/improve text analysis

lib/services/dictionary_service.dart

@@ -11,6 +11,7 @@ import 'package:archive/archive_io.dart' as archive;
 import 'package:sagase/datamodels/user_backup.dart';
 import 'package:sagase/services/isar_service.dart';
 import 'package:sagase/utils/constants.dart' as constants;
+import 'package:sagase/utils/enum_utils.dart';
 import 'package:sagase_dictionary/sagase_dictionary.dart';
 import 'package:path/path.dart' as path;
 
@@ -188,7 +189,7 @@ class DictionaryService {
 
       if (list.length == 1) {
         return list;
-      } else if (list.length > 1) {
+      } else if (list.isNotEmpty) {
         results = list;
       }
     }

lib/services/mecab_service.dart

@@ -133,24 +133,39 @@ class MecabService {
       PartOfSpeech? pos;
       if (tokens[i].features[0] == '助詞') {
         pos = PartOfSpeech.particle;
-      } else if (tokens[i].features[4] == 'サ変・スル') {
-        pos = PartOfSpeech.verbSuruIncluded;
+      } else if (tokens[i].features[0] == '動詞') {
+        pos = _identifyVerb(tokens[i].features);
       } else if (tokens[i].features[1] == '固有名詞' &&
           tokens[i].features[2] == '人名') {
         pos = PartOfSpeech.nounProper;
       }
 
-      // Handle corner cases where the base form (index 6) is different than the real base form
+      // Handle corner cases where the base or base reading form is different than the real form
       String base = tokens[i].features[6];
       if (tokens[i].surface == 'なら' && tokens[i].features[6] == 'だ') {
         base = 'なら';
       }
+      String baseReading = tokens[i].features[7];
+      if (baseReading.endsWith('ッ') &&
+          tokens[i].features[5] == '連用タ接続' &&
+          tokens.length > i + 1) {
+        switch (tokens[i + 1].features[6]) {
+          case 'て':
+            baseReading =
+                '${baseReading.substring(0, baseReading.length - 1)}ク';
+            break;
+          case 'で':
+            baseReading =
+                '${baseReading.substring(0, baseReading.length - 1)}グ';
+            break;
+        }
+      }
 
       // Created token to add to list or trailing of previous token
       final current = JapaneseTextToken(
         original: tokens[i].surface,
         base: base,
-        baseReading: tokens[i].features[7],
+        baseReading: baseReading,
         rubyTextPairs: createRubyTextPairs(
           tokens[i].surface,
           tokens[i].features[7],
@@ -159,7 +174,12 @@ class MecabService {
       );
 
       // Check if the current token should be trailing of previous token
-      if (list.isNotEmpty && list.last.pos != PartOfSpeech.particle) {
+      if (list.isNotEmpty) {
+        if (tokens[i - 1].features[5] == '連用タ接続') {
+          list.last.trailing ??= [];
+          list.last.trailing!.add(current);
+          continue;
+        } else if (list.last.pos != PartOfSpeech.particle) {
         if (tokens[i].features[1] == '接続助詞' && tokens[i].features[6] == 'て') {
           list.last.trailing ??= [];
           list.last.trailing!.add(current);
@@ -180,6 +200,7 @@ class MecabService {
           list.last.trailing ??= [];
           list.last.trailing!.add(current);
           continue;
+          }
         }
       }
 
@@ -189,6 +210,49 @@ class MecabService {
     return list;
   }
 
+  PartOfSpeech? _identifyVerb(List<String> features) {
+    switch (features[4]) {
+      case 'サ変・スル':
+        return PartOfSpeech.verbSuruIncluded;
+      case '一段':
+        return PartOfSpeech.verbIchidan;
+      case '一段・クレル':
+        return PartOfSpeech.verbIchidanS;
+      // case '':
+      //   return PartOfSpeech.verbGodanAru;
+      case '五段・ラ行':
+        if (features[6] == 'ある' || features[6] == '有る' || features[6] == 'アル') {
+          return PartOfSpeech.verbGodanRI;
+        } else {
+          return PartOfSpeech.verbGodanR;
+        }
+      case '五段・バ行':
+        return PartOfSpeech.verbGodanB;
+      case '五段・ガ行':
+        return PartOfSpeech.verbGodanG;
+      case '五段・カ行促音便':
+        return PartOfSpeech.verbGodanKS;
+      case '五段・カ行イ音便':
+        return PartOfSpeech.verbGodanK;
+      case '五段・マ行':
+        return PartOfSpeech.verbGodanM;
+      case '五段・ナ行':
+        return PartOfSpeech.verbGodanN;
+      case '五段・サ行':
+        return PartOfSpeech.verbGodanS;
+      case '五段・タ行':
+        return PartOfSpeech.verbGodanT;
+      case '五段・ワ行促音便':
+        return PartOfSpeech.verbGodanU;
+      // case '':
+      //   return PartOfSpeech.verbGodanUS;
+      // case '':
+      //   return PartOfSpeech.verbGodanUru;
+      default:
+        return null;
+    }
+  }
+
   List<RubyTextPair> createRubyTextPairs(String writing, String reading) {
     // First check if only kana
     if (_kanaKit.isKana(writing)) return [RubyTextPair(writing: writing)];