better:normal模式表达提取和切换优化

src/chat/focus_chat/info_processors/expression_selector_processor.py

@@ -285,75 +285,6 @@ class ExpressionSelectorProcessor(BaseProcessor):
             logger.error(f"{self.log_prefix} LLM处理表达方式选择时出错: {e}")
             return []
 
-    async def _select_suitable_expressions_random(self, chat_info: str) -> List[Dict[str, str]]:
-        """随机选择表达方式（原replyer逻辑）"""
-
-        # 获取所有表达方式
-        expression_learner = get_expression_learner()
-        (
-            learnt_style_expressions,
-            learnt_grammar_expressions,
-            personality_expressions,
-        ) = await expression_learner.get_expression_by_chat_id(self.subheartflow_id)
-
-        selected_expressions = []
-
-        # 1. learnt_style_expressions相似度匹配选择3条
-        if learnt_style_expressions:
-            similar_exprs = self._find_similar_expressions(chat_info, learnt_style_expressions, 3)
-            for expr in similar_exprs:
-                if isinstance(expr, dict) and "situation" in expr and "style" in expr:
-                    expr_copy = expr.copy()
-                    expr_copy["type"] = "style"
-                    selected_expressions.append(expr_copy)
-
-        # 2. learnt_grammar_expressions加权随机选2条
-        if learnt_grammar_expressions:
-            weights = [expr.get("count", 1) for expr in learnt_grammar_expressions]
-            selected_learnt = weighted_sample_no_replacement(learnt_grammar_expressions, weights, 2)
-            for expr in selected_learnt:
-                if isinstance(expr, dict) and "situation" in expr and "style" in expr:
-                    expr_copy = expr.copy()
-                    expr_copy["type"] = "grammar"
-                    selected_expressions.append(expr_copy)
-
-        # 3. personality_expressions随机选1条
-        if personality_expressions:
-            expr = random.choice(personality_expressions)
-            if isinstance(expr, dict) and "situation" in expr and "style" in expr:
-                expr_copy = expr.copy()
-                expr_copy["type"] = "personality"
-                selected_expressions.append(expr_copy)
-
-        logger.info(f"{self.log_prefix} 随机模式选择了{len(selected_expressions)}个表达方式")
-        return selected_expressions
-
-    def _find_similar_expressions(self, input_text: str, expressions: List[Dict], top_k: int = 3) -> List[Dict]:
-        """使用简单的文本匹配找出相似的表达方式（简化版，避免依赖sklearn）"""
-        if not expressions or not input_text:
-            return random.sample(expressions, min(top_k, len(expressions))) if expressions else []
-
-        # 简单的关键词匹配
-        scored_expressions = []
-        input_words = set(input_text.lower().split())
-
-        for expr in expressions:
-            situation = expr.get("situation", "").lower()
-            situation_words = set(situation.split())
-
-            # 计算交集大小作为相似度
-            similarity = len(input_words & situation_words)
-            scored_expressions.append((similarity, expr))
-
-        # 按相似度排序
-        scored_expressions.sort(key=lambda x: x[0], reverse=True)
-
-        # 如果没有匹配的，随机选择
-        if all(score == 0 for score, _ in scored_expressions):
-            return random.sample(expressions, min(top_k, len(expressions)))
-
-        # 返回top_k个最相似的
-        return [expr for _, expr in scored_expressions[:top_k]]
 
 
 init_prompt()

src/chat/focus_chat/replyer/default_generator.py

@@ -710,35 +710,5 @@ def weighted_sample_no_replacement(items, weights, k) -> list:
     return selected
 
 
-def find_similar_expressions(input_text: str, expressions: List[Dict], top_k: int = 3) -> List[Dict]:
-    """使用TF-IDF和余弦相似度找出与输入文本最相似的top_k个表达方式"""
-    if not expressions:
-        return []
-
-    # 准备文本数据
-    texts = [expr["situation"] for expr in expressions]
-    texts.append(input_text)  # 添加输入文本
-
-    # 使用TF-IDF向量化
-    vectorizer = TfidfVectorizer()
-    tfidf_matrix = vectorizer.fit_transform(texts)
-
-    # 计算余弦相似度
-    similarity_matrix = cosine_similarity(tfidf_matrix)
-
-    # 获取输入文本的相似度分数（最后一行）
-    scores = similarity_matrix[-1][:-1]  # 排除与自身的相似度
-
-    # 获取top_k的索引
-    top_indices = np.argsort(scores)[::-1][:top_k]
-
-    # 获取相似表达
-    similar_exprs = []
-    for idx in top_indices:
-        if scores[idx] > 0:  # 只保留有相似度的
-            similar_exprs.append(expressions[idx])
-
-    return similar_exprs
-
 
 init_prompt()