fix: Refactored Word Count Method

astrbot/core/pipeline/respond/stage.py

@@ -10,6 +10,7 @@
 from astrbot.core.platform.astr_message_event import AstrMessageEvent
 from astrbot.core.star.star_handler import EventType
 from astrbot.core.utils.path_util import path_Mapping
+from astrbot.core.utils.text_utils import calculate_word_count
 
 from ..context import PipelineContext, call_event_hook
 from ..stage import Stage, register_stage
@@ -88,12 +89,7 @@ async def initialize(self, ctx: PipelineContext) -> None:
             logger.info(f"分段回复间隔时间：{self.interval}")
 
     async def _word_cnt(self, text: str) -> int:
-        """分段回复 统计字数"""
-        if all(ord(c) < 128 for c in text):
-            word_count = len(text.split())
-        else:
-            word_count = len([c for c in text if c.isalnum()])
-        return word_count
+        return calculate_word_count(text)
 
     async def _calc_comp_interval(self, comp: BaseMessageComponent) -> float:
         """分段回复 计算间隔时间"""

astrbot/core/pipeline/result_decorate/stage.py

@@ -13,6 +13,7 @@
 from astrbot.core.star.session_llm_manager import SessionServiceManager
 from astrbot.core.star.star import star_map
 from astrbot.core.star.star_handler import EventType, star_handlers_registry
+from astrbot.core.utils.text_utils import calculate_word_count
 
 from ..context import PipelineContext
 from ..stage import Stage, register_stage, registered_stages
@@ -213,7 +214,8 @@ async def process(
                     new_chain = []
                     for comp in result.chain:
                         if isinstance(comp, Plain):
-                            if len(comp.text) > self.words_count_threshold:
+                            word_count = calculate_word_count(comp.text)
+                            if word_count > self.words_count_threshold:
                                 # 不分段回复
                                 new_chain.append(comp)
                                 continue
@@ -296,7 +298,7 @@ async def process(
             if should_tts and tts_provider:
                 new_chain = []
                 for comp in result.chain:
-                    if isinstance(comp, Plain) and len(comp.text) > 1:
+                    if isinstance(comp, Plain) and calculate_word_count(comp.text) > 1:
                         try:
                             logger.info(f"TTS 请求: {comp.text}")
                             audio_path = await tts_provider.get_audio(comp.text)
@@ -356,7 +358,8 @@ async def process(
                     else:
                         break
                 plain_str = "".join(parts)
-                if plain_str and len(plain_str) > self.t2i_word_threshold:
+                word_count = calculate_word_count(plain_str)
+                if plain_str and word_count > self.t2i_word_threshold:
                     render_start = time.time()
                     try:
                         url = await html_renderer.render_t2i(
@@ -391,7 +394,7 @@ async def process(
                 word_cnt = 0
                 for comp in result.chain:
                     if isinstance(comp, Plain):
-                        word_cnt += len(comp.text)
+                        word_cnt += calculate_word_count(comp.text)
                 if word_cnt > self.forward_threshold:
                     node = Node(
                         uin=event.get_self_id(),

astrbot/core/utils/text_utils.py

@@ -0,0 +1,21 @@
+import re
+
+
+def calculate_word_count(text: str) -> int:
+    """
+    将不同语言分开计算
+    - 中文/日语: 按字数计算
+    - 其他语言: 按空格分割计算
+    """
+    # \u4e00-\u9fff : CJK Unified Ideographs (Chinese Hanzi & Japanese Kanji)
+    # \u3040-\u309f : Japanese Hiragana
+    # \u30a0-\u30ff : Japanese Katakana
+    no_space_pattern = r"[\u4e00-\u9fff\u3040-\u309f\u30a0-\u30ff]"
+
+    char_count = len(re.findall(no_space_pattern, text))
+
+    text_remaining = re.sub(no_space_pattern, " ", text)
+
+    spaced_words = len(text_remaining.split())
+
+    return char_count + spaced_words