Add: ユーザー辞書機能で対応する品詞の種類を拡充

「固有名詞」の中のどういうカテゴリなのかも含めて辞書登録できるようにする
Aivis-Project · Jan 6, 2025 · e2d1053 · e2d1053
1 parent cd110d6
commit e2d1053
Show file tree

Hide file tree

Showing 4 changed files with 68 additions and 8 deletions.
diff --git a/test/e2e/__snapshots__/test_openapi/test_OpenAPIの形が変わっていないことを確認.json b/test/e2e/__snapshots__/test_openapi/test_OpenAPIの形が変わっていないことを確認.json
diff --git a/voicevox_engine/app/routers/user_dict.py b/voicevox_engine/app/routers/user_dict.py
@@ -58,7 +58,7 @@ def add_user_dict_word(
         word_type: Annotated[
             WordTypes | SkipJsonSchema[None],
             Query(
-                description="PROPER_NOUN（固有名詞）、COMMON_NOUN（普通名詞）、VERB（動詞）、ADJECTIVE（形容詞）、SUFFIX（語尾）のいずれか"
+                description="PROPER_NOUN（固有名詞）、LOCATION_NAME（地名）、ORGANIZATION_NAME（組織・施設名）、PERSON_NAME（人名）、PERSON_FAMILY_NAME（姓）、PERSON_GIVEN_NAME（名）、COMMON_NOUN（普通名詞）、VERB（動詞）、ADJECTIVE（形容詞）、SUFFIX（語尾）のいずれか"
             ),
         ] = None,
         priority: Annotated[

diff --git a/voicevox_engine/user_dict/model.py b/voicevox_engine/user_dict/model.py
@@ -15,11 +15,16 @@
 class WordTypes(str, Enum):
     """品詞"""
 
-    PROPER_NOUN = "PROPER_NOUN"
-    COMMON_NOUN = "COMMON_NOUN"
-    VERB = "VERB"
-    ADJECTIVE = "ADJECTIVE"
-    SUFFIX = "SUFFIX"
+    PROPER_NOUN = "PROPER_NOUN"  # 固有名詞
+    LOCATION_NAME = "LOCATION_NAME"  # 地名
+    ORGANIZATION_NAME = "ORGANIZATION_NAME"  # 組織・施設名
+    PERSON_NAME = "PERSON_NAME"  # 人名
+    PERSON_FAMILY_NAME = "PERSON_FAMILY_NAME"  # 姓
+    PERSON_GIVEN_NAME = "PERSON_GIVEN_NAME"  # 名
+    COMMON_NOUN = "COMMON_NOUN"  # 一般名詞
+    VERB = "VERB"  # 動詞
+    ADJECTIVE = "ADJECTIVE"  # 形容詞
+    SUFFIX = "SUFFIX"  # 接尾辞
 
 
 USER_DICT_MIN_PRIORITY = 0

diff --git a/voicevox_engine/user_dict/user_dict_word.py b/voicevox_engine/user_dict/user_dict_word.py
@@ -26,6 +26,11 @@ class _PartOfSpeechDetail:
 
 
 _costs_proper_noun = [-988, 3488, 4768, 6048, 7328, 8609, 8734, 8859, 8984, 9110, 14176]
+_costs_location_name = [147, 4350, 5403, 6456, 7509, 8562, 8778, 8994, 9210, 9427, 13960]  # fmt: skip # noqa
+_costs_organization_name = [1691, 4087, 4808, 5530, 6251, 6973, 7420, 7867, 8314, 8762, 13887]  # fmt: skip # noqa
+_costs_person_name = [515, 516, 2253, 3990, 5727, 7464, 8098, 8732, 9366, 10000, 10001]
+_costs_person_family_name = [1991, 4126, 4974, 5823, 6672, 7521, 8001, 8481, 8961, 9442, 12808]  # fmt: skip # noqa
+_costs_person_given_name = [2209, 3772, 4905, 6038, 7171, 8304, 8728, 9152, 9576, 10000, 13842]  # fmt: skip # noqa
 _costs_common_noun = [-4445, 49, 1473, 2897, 4321, 5746, 6554, 7362, 8170, 8979, 15001]
 _costs_verb = [3100, 6160, 6360, 6561, 6761, 6962, 7414, 7866, 8318, 8771, 13433]
 _costs_adjective = [1527, 3266, 3561, 3857, 4153, 4449, 5149, 5849, 6549, 7250, 10001]
@@ -42,6 +47,51 @@ class _PartOfSpeechDetail:
         cost_candidates=_costs_proper_noun,
         accent_associative_rules=["*", "C1", "C2", "C3", "C4", "C5"],
     ),
+    WordTypes.LOCATION_NAME: _PartOfSpeechDetail(
+        part_of_speech="名詞",
+        part_of_speech_detail_1="固有名詞",
+        part_of_speech_detail_2="地域",
+        part_of_speech_detail_3="一般",
+        context_id=1353,
+        cost_candidates=_costs_location_name,
+        accent_associative_rules=["*", "C1", "C2", "C3", "C4", "C5"],
+    ),
+    WordTypes.ORGANIZATION_NAME: _PartOfSpeechDetail(
+        part_of_speech="名詞",
+        part_of_speech_detail_1="固有名詞",
+        part_of_speech_detail_2="組織",
+        part_of_speech_detail_3="*",
+        context_id=1352,
+        cost_candidates=_costs_organization_name,
+        accent_associative_rules=["*", "C1", "C2", "C3", "C4", "C5"],
+    ),
+    WordTypes.PERSON_NAME: _PartOfSpeechDetail(
+        part_of_speech="名詞",
+        part_of_speech_detail_1="固有名詞",
+        part_of_speech_detail_2="人名",
+        part_of_speech_detail_3="一般",
+        context_id=1349,
+        cost_candidates=_costs_person_name,
+        accent_associative_rules=["*", "C1", "C2", "C3", "C4", "C5"],
+    ),
+    WordTypes.PERSON_FAMILY_NAME: _PartOfSpeechDetail(
+        part_of_speech="名詞",
+        part_of_speech_detail_1="固有名詞",
+        part_of_speech_detail_2="人名",
+        part_of_speech_detail_3="姓",
+        context_id=1350,
+        cost_candidates=_costs_person_family_name,
+        accent_associative_rules=["*", "C1", "C2", "C3", "C4", "C5"],
+    ),
+    WordTypes.PERSON_GIVEN_NAME: _PartOfSpeechDetail(
+        part_of_speech="名詞",
+        part_of_speech_detail_1="固有名詞",
+        part_of_speech_detail_2="人名",
+        part_of_speech_detail_3="名",
+        context_id=1351,
+        cost_candidates=_costs_person_given_name,
+        accent_associative_rules=["*", "C1", "C2", "C3", "C4", "C5"],
+    ),
     WordTypes.COMMON_NOUN: _PartOfSpeechDetail(
         part_of_speech="名詞",
         part_of_speech_detail_1="一般",