feat(chat): 优化提及检测并精简兴趣度评分日志
增强了机器人提及检测逻辑,使其能够识别配置文件中设置的所有别名(alias_names),而不仅仅是主昵称。这提高了交互的灵活性和准确性。 此外,还对兴趣度评分和匹配系统的日志输出进行了大幅重构: - 将多条评分计算日志合并为一条包含核心指标的摘要日志,使输出更简洁。 - 调整了部分日志级别,将非关键信息移至 DEBUG 级别,以减少日志噪音。 - 在关键日志中增加了消息内容预览,以便于快速上下文定位和调试。
This commit is contained in:
@@ -124,7 +124,10 @@ class ActionPlanner:
|
||||
# 3. 根据兴趣度调整可用动作
|
||||
if interest_scores:
|
||||
latest_score = max(interest_scores, key=lambda s: s.total_score)
|
||||
should_reply, score = self.interest_scoring.should_reply(latest_score)
|
||||
latest_message = next(
|
||||
(msg for msg in unread_messages if msg.message_id == latest_score.message_id), None
|
||||
)
|
||||
should_reply, score = self.interest_scoring.should_reply(latest_score, latest_message)
|
||||
|
||||
reply_not_available = False
|
||||
if not should_reply and "reply" in initial_plan.available_actions:
|
||||
|
||||
Reference in New Issue
Block a user