better：微调关系prompt的构建
This commit is contained in:
SengokuCola
767
src/chat/replyer/default_generator.py
Normal file
767
src/chat/replyer/default_generator.py
Normal file
@@ -0,0 +1,767 @@
|
||||
import traceback
|
||||
from typing import List, Optional, Dict, Any, Tuple
|
||||
|
||||
from src.chat.message_receive.message import MessageRecv, MessageThinking, MessageSending
|
||||
from src.chat.message_receive.message import Seg # Local import needed after move
|
||||
from src.chat.message_receive.message import UserInfo
|
||||
from src.chat.message_receive.chat_stream import get_chat_manager
|
||||
from src.common.logger import get_logger
|
||||
from src.llm_models.utils_model import LLMRequest
|
||||
from src.config.config import global_config
|
||||
from src.chat.utils.timer_calculator import Timer # <--- Import Timer
|
||||
from src.chat.focus_chat.heartFC_sender import HeartFCSender
|
||||
from src.chat.utils.utils import process_llm_response
|
||||
from src.chat.heart_flow.utils_chat import get_chat_type_and_target_info
|
||||
from src.chat.message_receive.chat_stream import ChatStream
|
||||
from src.chat.focus_chat.hfc_utils import parse_thinking_id_to_timestamp
|
||||
from src.chat.utils.prompt_builder import Prompt, global_prompt_manager
|
||||
from src.chat.utils.chat_message_builder import build_readable_messages, get_raw_msg_before_timestamp_with_chat
|
||||
from src.chat.express.exprssion_learner import get_expression_learner
|
||||
import time
|
||||
import random
|
||||
import ast
|
||||
from src.person_info.person_info import get_person_info_manager
|
||||
from datetime import datetime
|
||||
import re
|
||||
|
||||
logger = get_logger("replyer")
|
||||
|
||||
|
||||
def init_prompt():
|
||||
Prompt(
|
||||
"""
|
||||
{expression_habits_block}
|
||||
{structured_info_block}
|
||||
{memory_block}
|
||||
{relation_info_block}
|
||||
{extra_info_block}
|
||||
{time_block}
|
||||
{chat_target}
|
||||
{chat_info}
|
||||
{reply_target_block}
|
||||
{identity}
|
||||
|
||||
你需要使用合适的语言习惯和句法，参考聊天内容，组织一条日常且口语化的回复。注意不要复读你说过的话。
|
||||
{config_expression_style}。回复不要浮夸，不要用夸张修辞，平淡一些。
|
||||
{keywords_reaction_prompt}
|
||||
请不要输出违法违规内容，不要输出色情，暴力，政治相关内容，如有敏感内容，请规避。
|
||||
不要浮夸，不要夸张修辞，请注意不要输出多余内容(包括前后缀，冒号和引号，括号()，表情包，at或 @等 )。只输出一条回复就好。
|
||||
现在，你说：
|
||||
""",
|
||||
"default_generator_prompt",
|
||||
)
|
||||
|
||||
Prompt(
|
||||
"""
|
||||
{expression_habits_block}
|
||||
{structured_info_block}
|
||||
{memory_block}
|
||||
{relation_info_block}
|
||||
{extra_info_block}
|
||||
{time_block}
|
||||
{chat_target}
|
||||
{chat_info}
|
||||
现在"{sender_name}"说:{target_message}。你想要回复对方的这条消息。
|
||||
{identity}，
|
||||
你需要使用合适的语法和句法，参考聊天内容，组织一条日常且口语化的回复。注意不要复读你说过的话。
|
||||
|
||||
{config_expression_style}。回复不要浮夸，不要用夸张修辞，平淡一些。
|
||||
{keywords_reaction_prompt}
|
||||
请不要输出违法违规内容，不要输出色情，暴力，政治相关内容，如有敏感内容，请规避。
|
||||
不要浮夸，不要夸张修辞，请注意不要输出多余内容(包括前后缀，冒号和引号，括号()，表情包，at或 @等 )。只输出一条回复就好。
|
||||
现在，你说：
|
||||
""",
|
||||
"default_generator_private_prompt",
|
||||
)
|
||||
|
||||
Prompt(
|
||||
"""
|
||||
你可以参考你的以下的语言习惯，如果情景合适就使用，不要盲目使用,不要生硬使用，而是结合到表达中：
|
||||
{style_habbits}
|
||||
|
||||
你现在正在群里聊天，以下是群里正在进行的聊天内容：
|
||||
{chat_info}
|
||||
|
||||
以上是聊天内容，你需要了解聊天记录中的内容
|
||||
|
||||
{chat_target}
|
||||
你的名字是{bot_name}，{prompt_personality}，在这聊天中，"{sender_name}"说的"{target_message}"引起了你的注意，对这句话，你想表达：{raw_reply},原因是：{reason}。你现在要思考怎么回复
|
||||
你需要使用合适的语法和句法，参考聊天内容，组织一条日常且口语化的回复。请你修改你想表达的原句，符合你的表达风格和语言习惯
|
||||
请你根据情景使用以下句法：
|
||||
{grammar_habbits}
|
||||
{config_expression_style}，你可以完全重组回复，保留最基本的表达含义就好，但重组后保持语意通顺。
|
||||
不要浮夸，不要夸张修辞，平淡且不要输出多余内容(包括前后缀，冒号和引号，括号，表情包，at或 @等 )，只输出一条回复就好。
|
||||
现在，你说：
|
||||
""",
|
||||
"default_expressor_prompt",
|
||||
)
|
||||
|
||||
Prompt(
|
||||
"""
|
||||
你可以参考以下的语言习惯，如果情景合适就使用，不要盲目使用,不要生硬使用，而是结合到表达中：
|
||||
{style_habbits}
|
||||
|
||||
你现在正在群里聊天，以下是群里正在进行的聊天内容：
|
||||
{chat_info}
|
||||
|
||||
以上是聊天内容，你需要了解聊天记录中的内容
|
||||
|
||||
{chat_target}
|
||||
你的名字是{bot_name}，{prompt_personality}，在这聊天中，"{sender_name}"说的"{target_message}"引起了你的注意，对这句话，你想表达：{raw_reply},原因是：{reason}。你现在要思考怎么回复
|
||||
你需要使用合适的语法和句法，参考聊天内容，组织一条日常且口语化的回复。
|
||||
请你根据情景使用以下句法：
|
||||
{grammar_habbits}
|
||||
{config_expression_style}，你可以完全重组回复，保留最基本的表达含义就好，但重组后保持语意通顺。
|
||||
不要浮夸，不要夸张修辞，平淡且不要输出多余内容(包括前后缀，冒号和引号，括号，表情包，at或 @等 )，只输出一条回复就好。
|
||||
现在，你说：
|
||||
""",
|
||||
"default_expressor_private_prompt", # New template for private FOCUSED chat
|
||||
)
|
||||
|
||||
|
||||
class DefaultReplyer:
|
||||
def __init__(self, chat_stream: ChatStream):
|
||||
self.log_prefix = "replyer"
|
||||
# TODO: API-Adapter修改标记
|
||||
self.express_model = LLMRequest(
|
||||
model=global_config.model.replyer_1,
|
||||
request_type="focus.replyer",
|
||||
)
|
||||
self.heart_fc_sender = HeartFCSender()
|
||||
|
||||
self.chat_stream = chat_stream
|
||||
self.is_group_chat, self.chat_target_info = get_chat_type_and_target_info(self.chat_stream.stream_id)
|
||||
|
||||
async def _create_thinking_message(self, anchor_message: Optional[MessageRecv], thinking_id: str):
|
||||
"""创建思考消息 (尝试锚定到 anchor_message)"""
|
||||
if not anchor_message or not anchor_message.chat_stream:
|
||||
logger.error(f"{self.log_prefix} 无法创建思考消息，缺少有效的锚点消息或聊天流。")
|
||||
return None
|
||||
|
||||
chat = anchor_message.chat_stream
|
||||
messageinfo = anchor_message.message_info
|
||||
thinking_time_point = parse_thinking_id_to_timestamp(thinking_id)
|
||||
bot_user_info = UserInfo(
|
||||
user_id=global_config.bot.qq_account,
|
||||
user_nickname=global_config.bot.nickname,
|
||||
platform=messageinfo.platform,
|
||||
)
|
||||
|
||||
thinking_message = MessageThinking(
|
||||
message_id=thinking_id,
|
||||
chat_stream=chat,
|
||||
bot_user_info=bot_user_info,
|
||||
reply=anchor_message, # 回复的是锚点消息
|
||||
thinking_start_time=thinking_time_point,
|
||||
)
|
||||
# logger.debug(f"创建思考消息thinking_message：{thinking_message}")
|
||||
|
||||
await self.heart_fc_sender.register_thinking(thinking_message)
|
||||
return None
|
||||
|
||||
async def generate_reply_with_context(
|
||||
self,
|
||||
reply_data: Dict[str, Any],
|
||||
) -> Tuple[bool, Optional[List[str]]]:
|
||||
"""
|
||||
回复器 (Replier): 核心逻辑，负责生成回复文本。
|
||||
(已整合原 HeartFCGenerator 的功能)
|
||||
"""
|
||||
try:
|
||||
# 3. 构建 Prompt
|
||||
with Timer("构建Prompt", {}): # 内部计时器，可选保留
|
||||
prompt = await self.build_prompt_reply_context(
|
||||
reply_data=reply_data, # 传递action_data
|
||||
)
|
||||
|
||||
# 4. 调用 LLM 生成回复
|
||||
content = None
|
||||
reasoning_content = None
|
||||
model_name = "unknown_model"
|
||||
|
||||
try:
|
||||
with Timer("LLM生成", {}): # 内部计时器，可选保留
|
||||
logger.info(f"{self.log_prefix}Prompt:\n{prompt}\n")
|
||||
content, (reasoning_content, model_name) = await self.express_model.generate_response_async(prompt)
|
||||
|
||||
logger.info(f"最终回复: {content}")
|
||||
|
||||
except Exception as llm_e:
|
||||
# 精简报错信息
|
||||
logger.error(f"{self.log_prefix}LLM 生成失败: {llm_e}")
|
||||
return False, None # LLM 调用失败则无法生成回复
|
||||
|
||||
processed_response = process_llm_response(content)
|
||||
|
||||
# 5. 处理 LLM 响应
|
||||
if not content:
|
||||
logger.warning(f"{self.log_prefix}LLM 生成了空内容。")
|
||||
return False, None
|
||||
if not processed_response:
|
||||
logger.warning(f"{self.log_prefix}处理后的回复为空。")
|
||||
return False, None
|
||||
|
||||
reply_set = []
|
||||
for str in processed_response:
|
||||
reply_seg = ("text", str)
|
||||
reply_set.append(reply_seg)
|
||||
|
||||
return True, reply_set
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"{self.log_prefix}回复生成意外失败: {e}")
|
||||
traceback.print_exc()
|
||||
return False, None
|
||||
|
||||
async def rewrite_reply_with_context(
|
||||
self,
|
||||
reply_data: Dict[str, Any],
|
||||
) -> Tuple[bool, Optional[List[str]]]:
|
||||
"""
|
||||
表达器 (Expressor): 核心逻辑，负责生成回复文本。
|
||||
"""
|
||||
try:
|
||||
reply_to = reply_data.get("reply_to", "")
|
||||
raw_reply = reply_data.get("raw_reply", "")
|
||||
reason = reply_data.get("reason", "")
|
||||
|
||||
with Timer("构建Prompt", {}): # 内部计时器，可选保留
|
||||
prompt = await self.build_prompt_rewrite_context(
|
||||
raw_reply=raw_reply,
|
||||
reason=reason,
|
||||
reply_to=reply_to,
|
||||
)
|
||||
|
||||
content = None
|
||||
reasoning_content = None
|
||||
model_name = "unknown_model"
|
||||
if not prompt:
|
||||
logger.error(f"{self.log_prefix}Prompt 构建失败，无法生成回复。")
|
||||
return False, None
|
||||
|
||||
try:
|
||||
with Timer("LLM生成", {}): # 内部计时器，可选保留
|
||||
# TODO: API-Adapter修改标记
|
||||
content, (reasoning_content, model_name) = await self.express_model.generate_response_async(prompt)
|
||||
|
||||
logger.info(f"想要表达：{raw_reply}||理由：{reason}")
|
||||
logger.info(f"最终回复: {content}\n")
|
||||
|
||||
except Exception as llm_e:
|
||||
# 精简报错信息
|
||||
logger.error(f"{self.log_prefix}LLM 生成失败: {llm_e}")
|
||||
return False, None # LLM 调用失败则无法生成回复
|
||||
|
||||
processed_response = process_llm_response(content)
|
||||
|
||||
# 5. 处理 LLM 响应
|
||||
if not content:
|
||||
logger.warning(f"{self.log_prefix}LLM 生成了空内容。")
|
||||
return False, None
|
||||
if not processed_response:
|
||||
logger.warning(f"{self.log_prefix}处理后的回复为空。")
|
||||
return False, None
|
||||
|
||||
reply_set = []
|
||||
for str in processed_response:
|
||||
reply_seg = ("text", str)
|
||||
reply_set.append(reply_seg)
|
||||
|
||||
return True, reply_set
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"{self.log_prefix}回复生成意外失败: {e}")
|
||||
traceback.print_exc()
|
||||
return False, None
|
||||
|
||||
async def build_prompt_reply_context(
|
||||
self,
|
||||
reply_data=None,
|
||||
) -> str:
|
||||
chat_stream = self.chat_stream
|
||||
person_info_manager = get_person_info_manager()
|
||||
bot_person_id = person_info_manager.get_person_id("system", "bot_id")
|
||||
|
||||
is_group_chat = bool(chat_stream.group_info)
|
||||
|
||||
self_info_block = reply_data.get("self_info_block", "")
|
||||
structured_info = reply_data.get("structured_info", "")
|
||||
relation_info_block = reply_data.get("relation_info_block", "")
|
||||
reply_to = reply_data.get("reply_to", "none")
|
||||
memory_block = reply_data.get("memory_block", "")
|
||||
|
||||
# 优先使用 extra_info_block，没有则用 extra_info
|
||||
extra_info_block = reply_data.get("extra_info_block", "") or reply_data.get("extra_info", "")
|
||||
|
||||
sender = ""
|
||||
target = ""
|
||||
if ":" in reply_to or "：" in reply_to:
|
||||
# 使用正则表达式匹配中文或英文冒号
|
||||
parts = re.split(pattern=r"[:：]", string=reply_to, maxsplit=1)
|
||||
if len(parts) == 2:
|
||||
sender = parts[0].strip()
|
||||
target = parts[1].strip()
|
||||
|
||||
message_list_before_now = get_raw_msg_before_timestamp_with_chat(
|
||||
chat_id=chat_stream.stream_id,
|
||||
timestamp=time.time(),
|
||||
limit=global_config.focus_chat.observation_context_size,
|
||||
)
|
||||
# print(f"message_list_before_now: {message_list_before_now}")
|
||||
chat_talking_prompt = build_readable_messages(
|
||||
message_list_before_now,
|
||||
replace_bot_name=True,
|
||||
merge_messages=False,
|
||||
timestamp_mode="normal_no_YMD",
|
||||
read_mark=0.0,
|
||||
truncate=True,
|
||||
show_actions=True,
|
||||
)
|
||||
# print(f"chat_talking_prompt: {chat_talking_prompt}")
|
||||
|
||||
style_habbits = []
|
||||
grammar_habbits = []
|
||||
|
||||
# 使用从处理器传来的选中表达方式
|
||||
selected_expressions = reply_data.get("selected_expressions", []) if reply_data else []
|
||||
|
||||
if selected_expressions:
|
||||
logger.info(f"{self.log_prefix} 使用处理器选中的{len(selected_expressions)}个表达方式")
|
||||
for expr in selected_expressions:
|
||||
if isinstance(expr, dict) and "situation" in expr and "style" in expr:
|
||||
expr_type = expr.get("type", "style")
|
||||
if expr_type == "grammar":
|
||||
grammar_habbits.append(f"当{expr['situation']}时，使用 {expr['style']}")
|
||||
else:
|
||||
style_habbits.append(f"当{expr['situation']}时，使用 {expr['style']}")
|
||||
else:
|
||||
logger.debug(f"{self.log_prefix} 没有从处理器获得表达方式，将使用空的表达方式")
|
||||
# 不再在replyer中进行随机选择，全部交给处理器处理
|
||||
|
||||
style_habbits_str = "\n".join(style_habbits)
|
||||
grammar_habbits_str = "\n".join(grammar_habbits)
|
||||
|
||||
# 动态构建expression habits块
|
||||
expression_habits_block = ""
|
||||
if style_habbits_str.strip():
|
||||
expression_habits_block += f"你可以参考以下的语言习惯，如果情景合适就使用，不要盲目使用,不要生硬使用，而是结合到表达中：\n{style_habbits_str}\n\n"
|
||||
if grammar_habbits_str.strip():
|
||||
expression_habits_block += f"请你根据情景使用以下句法：\n{grammar_habbits_str}\n"
|
||||
|
||||
if structured_info:
|
||||
structured_info_block = f"以下是一些额外的信息，现在请你阅读以下内容，进行决策\n{structured_info}\n以上是一些额外的信息，现在请你阅读以下内容，进行决策"
|
||||
else:
|
||||
structured_info_block = ""
|
||||
|
||||
if extra_info_block:
|
||||
extra_info_block = f"以下是你在回复时需要参考的信息，现在请你阅读以下内容，进行决策\n{extra_info_block}\n以上是你在回复时需要参考的信息，现在请你阅读以下内容，进行决策"
|
||||
else:
|
||||
extra_info_block = ""
|
||||
|
||||
# 关键词检测与反应
|
||||
keywords_reaction_prompt = ""
|
||||
try:
|
||||
# 处理关键词规则
|
||||
for rule in global_config.keyword_reaction.keyword_rules:
|
||||
if any(keyword in target for keyword in rule.keywords):
|
||||
logger.info(f"检测到关键词规则：{rule.keywords}，触发反应：{rule.reaction}")
|
||||
keywords_reaction_prompt += f"{rule.reaction}，"
|
||||
|
||||
# 处理正则表达式规则
|
||||
for rule in global_config.keyword_reaction.regex_rules:
|
||||
for pattern_str in rule.regex:
|
||||
try:
|
||||
pattern = re.compile(pattern_str)
|
||||
if result := pattern.search(target):
|
||||
reaction = rule.reaction
|
||||
for name, content in result.groupdict().items():
|
||||
reaction = reaction.replace(f"[{name}]", content)
|
||||
logger.info(f"匹配到正则表达式：{pattern_str}，触发反应：{reaction}")
|
||||
keywords_reaction_prompt += reaction + "，"
|
||||
break
|
||||
except re.error as e:
|
||||
logger.error(f"正则表达式编译错误: {pattern_str}, 错误信息: {str(e)}")
|
||||
continue
|
||||
except Exception as e:
|
||||
logger.error(f"关键词检测与反应时发生异常: {str(e)}", exc_info=True)
|
||||
|
||||
time_block = f"当前时间：{datetime.now().strftime('%Y-%m-%d %H:%M:%S')}"
|
||||
|
||||
# logger.debug("开始构建 focus prompt")
|
||||
bot_name = global_config.bot.nickname
|
||||
if global_config.bot.alias_names:
|
||||
bot_nickname = f",也有人叫你{','.join(global_config.bot.alias_names)}"
|
||||
else:
|
||||
bot_nickname = ""
|
||||
short_impression = await person_info_manager.get_value(bot_person_id, "short_impression")
|
||||
# 解析字符串形式的Python列表
|
||||
try:
|
||||
if isinstance(short_impression, str) and short_impression.strip():
|
||||
short_impression = ast.literal_eval(short_impression)
|
||||
elif not short_impression:
|
||||
logger.warning("short_impression为空，使用默认值")
|
||||
short_impression = ["友好活泼", "人类"]
|
||||
except (ValueError, SyntaxError) as e:
|
||||
logger.error(f"解析short_impression失败: {e}, 原始值: {short_impression}")
|
||||
short_impression = ["友好活泼", "人类"]
|
||||
|
||||
# 确保short_impression是列表格式且有足够的元素
|
||||
if not isinstance(short_impression, list) or len(short_impression) < 2:
|
||||
logger.warning(f"short_impression格式不正确: {short_impression}, 使用默认值")
|
||||
short_impression = ["友好活泼", "人类"]
|
||||
personality = short_impression[0]
|
||||
identity = short_impression[1]
|
||||
prompt_personality = personality + "，" + identity
|
||||
indentify_block = f"你的名字是{bot_name}{bot_nickname}，你{prompt_personality}："
|
||||
|
||||
if sender:
|
||||
reply_target_block = f"现在{sender}说的:{target}。引起了你的注意，你想要在群里发言或者回复这条消息。"
|
||||
elif target:
|
||||
reply_target_block = f"现在{target}引起了你的注意，你想要在群里发言或者回复这条消息。"
|
||||
else:
|
||||
reply_target_block = "现在，你想要在群里发言或者回复消息。"
|
||||
|
||||
# --- Choose template based on chat type ---
|
||||
if is_group_chat:
|
||||
template_name = "default_generator_prompt"
|
||||
# Group specific formatting variables (already fetched or default)
|
||||
chat_target_1 = await global_prompt_manager.get_prompt_async("chat_target_group1")
|
||||
# chat_target_2 = await global_prompt_manager.get_prompt_async("chat_target_group2")
|
||||
|
||||
prompt = await global_prompt_manager.format_prompt(
|
||||
template_name,
|
||||
expression_habits_block=expression_habits_block,
|
||||
chat_target=chat_target_1,
|
||||
chat_info=chat_talking_prompt,
|
||||
memory_block=memory_block,
|
||||
structured_info_block=structured_info_block,
|
||||
extra_info_block=extra_info_block,
|
||||
relation_info_block=relation_info_block,
|
||||
self_info_block=self_info_block,
|
||||
time_block=time_block,
|
||||
reply_target_block=reply_target_block,
|
||||
keywords_reaction_prompt=keywords_reaction_prompt,
|
||||
identity=indentify_block,
|
||||
target_message=target,
|
||||
sender_name=sender,
|
||||
config_expression_style=global_config.expression.expression_style,
|
||||
)
|
||||
else: # Private chat
|
||||
template_name = "default_generator_private_prompt"
|
||||
# 在私聊时获取对方的昵称信息
|
||||
chat_target_name = "对方"
|
||||
if self.chat_target_info:
|
||||
chat_target_name = (
|
||||
self.chat_target_info.get("person_name") or self.chat_target_info.get("user_nickname") or "对方"
|
||||
)
|
||||
chat_target_1 = f"你正在和 {chat_target_name} 聊天"
|
||||
prompt = await global_prompt_manager.format_prompt(
|
||||
template_name,
|
||||
expression_habits_block=expression_habits_block,
|
||||
chat_target=chat_target_1,
|
||||
chat_info=chat_talking_prompt,
|
||||
memory_block=memory_block,
|
||||
structured_info_block=structured_info_block,
|
||||
relation_info_block=relation_info_block,
|
||||
extra_info_block=extra_info_block,
|
||||
time_block=time_block,
|
||||
keywords_reaction_prompt=keywords_reaction_prompt,
|
||||
identity=indentify_block,
|
||||
target_message=target,
|
||||
sender_name=sender,
|
||||
config_expression_style=global_config.expression.expression_style,
|
||||
)
|
||||
|
||||
return prompt
|
||||
|
||||
async def build_prompt_rewrite_context(
|
||||
self,
|
||||
reason,
|
||||
raw_reply,
|
||||
reply_to,
|
||||
) -> str:
|
||||
sender = ""
|
||||
target = ""
|
||||
if ":" in reply_to or "：" in reply_to:
|
||||
# 使用正则表达式匹配中文或英文冒号
|
||||
parts = re.split(pattern=r"[:：]", string=reply_to, maxsplit=1)
|
||||
if len(parts) == 2:
|
||||
sender = parts[0].strip()
|
||||
target = parts[1].strip()
|
||||
|
||||
chat_stream = self.chat_stream
|
||||
|
||||
is_group_chat = bool(chat_stream.group_info)
|
||||
|
||||
message_list_before_now = get_raw_msg_before_timestamp_with_chat(
|
||||
chat_id=chat_stream.stream_id,
|
||||
timestamp=time.time(),
|
||||
limit=global_config.focus_chat.observation_context_size,
|
||||
)
|
||||
chat_talking_prompt = build_readable_messages(
|
||||
message_list_before_now,
|
||||
replace_bot_name=True,
|
||||
merge_messages=True,
|
||||
timestamp_mode="relative",
|
||||
read_mark=0.0,
|
||||
truncate=True,
|
||||
)
|
||||
|
||||
expression_learner = get_expression_learner()
|
||||
(
|
||||
learnt_style_expressions,
|
||||
learnt_grammar_expressions,
|
||||
personality_expressions,
|
||||
) = expression_learner.get_expression_by_chat_id(chat_stream.stream_id)
|
||||
|
||||
style_habbits = []
|
||||
grammar_habbits = []
|
||||
# 1. learnt_expressions加权随机选3条
|
||||
if learnt_style_expressions:
|
||||
weights = [expr["count"] for expr in learnt_style_expressions]
|
||||
selected_learnt = weighted_sample_no_replacement(learnt_style_expressions, weights, 3)
|
||||
for expr in selected_learnt:
|
||||
if isinstance(expr, dict) and "situation" in expr and "style" in expr:
|
||||
style_habbits.append(f"当{expr['situation']}时，使用 {expr['style']}")
|
||||
# 2. learnt_grammar_expressions加权随机选3条
|
||||
if learnt_grammar_expressions:
|
||||
weights = [expr["count"] for expr in learnt_grammar_expressions]
|
||||
selected_learnt = weighted_sample_no_replacement(learnt_grammar_expressions, weights, 3)
|
||||
for expr in selected_learnt:
|
||||
if isinstance(expr, dict) and "situation" in expr and "style" in expr:
|
||||
grammar_habbits.append(f"当{expr['situation']}时，使用 {expr['style']}")
|
||||
# 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:
|
||||
style_habbits.append(f"当{expr['situation']}时，使用 {expr['style']}")
|
||||
|
||||
style_habbits_str = "\n".join(style_habbits)
|
||||
grammar_habbits_str = "\n".join(grammar_habbits)
|
||||
|
||||
logger.debug("开始构建 focus prompt")
|
||||
|
||||
# --- Choose template based on chat type ---
|
||||
if is_group_chat:
|
||||
template_name = "default_expressor_prompt"
|
||||
# Group specific formatting variables (already fetched or default)
|
||||
chat_target_1 = await global_prompt_manager.get_prompt_async("chat_target_group1")
|
||||
# chat_target_2 = await global_prompt_manager.get_prompt_async("chat_target_group2")
|
||||
|
||||
prompt = await global_prompt_manager.format_prompt(
|
||||
template_name,
|
||||
style_habbits=style_habbits_str,
|
||||
grammar_habbits=grammar_habbits_str,
|
||||
chat_target=chat_target_1,
|
||||
chat_info=chat_talking_prompt,
|
||||
bot_name=global_config.bot.nickname,
|
||||
prompt_personality="",
|
||||
reason=reason,
|
||||
raw_reply=raw_reply,
|
||||
sender_name=sender,
|
||||
target_message=target,
|
||||
config_expression_style=global_config.expression.expression_style,
|
||||
)
|
||||
else: # Private chat
|
||||
template_name = "default_expressor_private_prompt"
|
||||
# 在私聊时获取对方的昵称信息
|
||||
chat_target_name = "对方"
|
||||
if self.chat_target_info:
|
||||
chat_target_name = (
|
||||
self.chat_target_info.get("person_name") or self.chat_target_info.get("user_nickname") or "对方"
|
||||
)
|
||||
chat_target_1 = f"你正在和 {chat_target_name} 聊天"
|
||||
prompt = await global_prompt_manager.format_prompt(
|
||||
template_name,
|
||||
style_habbits=style_habbits_str,
|
||||
grammar_habbits=grammar_habbits_str,
|
||||
chat_target=chat_target_1,
|
||||
chat_info=chat_talking_prompt,
|
||||
bot_name=global_config.bot.nickname,
|
||||
prompt_personality="",
|
||||
reason=reason,
|
||||
raw_reply=raw_reply,
|
||||
sender_name=sender,
|
||||
target_message=target,
|
||||
config_expression_style=global_config.expression.expression_style,
|
||||
)
|
||||
|
||||
return prompt
|
||||
|
||||
async def send_response_messages(
|
||||
self,
|
||||
anchor_message: Optional[MessageRecv],
|
||||
response_set: List[Tuple[str, str]],
|
||||
thinking_id: str = "",
|
||||
display_message: str = "",
|
||||
) -> Optional[MessageSending]:
|
||||
"""发送回复消息 (尝试锚定到 anchor_message)，使用 HeartFCSender"""
|
||||
chat = self.chat_stream
|
||||
chat_id = self.chat_stream.stream_id
|
||||
if chat is None:
|
||||
logger.error(f"{self.log_prefix} 无法发送回复，chat_stream 为空。")
|
||||
return None
|
||||
if not anchor_message:
|
||||
logger.error(f"{self.log_prefix} 无法发送回复，anchor_message 为空。")
|
||||
return None
|
||||
|
||||
stream_name = get_chat_manager().get_stream_name(chat_id) or chat_id # 获取流名称用于日志
|
||||
|
||||
# 检查思考过程是否仍在进行，并获取开始时间
|
||||
if thinking_id:
|
||||
# print(f"thinking_id: {thinking_id}")
|
||||
thinking_start_time = await self.heart_fc_sender.get_thinking_start_time(chat_id, thinking_id)
|
||||
else:
|
||||
print("thinking_id is None")
|
||||
# thinking_id = "ds" + str(round(time.time(), 2))
|
||||
thinking_start_time = time.time()
|
||||
|
||||
if thinking_start_time is None:
|
||||
logger.error(f"[{stream_name}]replyer思考过程未找到或已结束，无法发送回复。")
|
||||
return None
|
||||
|
||||
mark_head = False
|
||||
# first_bot_msg: Optional[MessageSending] = None
|
||||
reply_message_ids = [] # 记录实际发送的消息ID
|
||||
|
||||
sent_msg_list = []
|
||||
|
||||
for i, msg_text in enumerate(response_set):
|
||||
# 为每个消息片段生成唯一ID
|
||||
type = msg_text[0]
|
||||
data = msg_text[1]
|
||||
|
||||
if global_config.experimental.debug_show_chat_mode and type == "text":
|
||||
data += "ᶠ"
|
||||
|
||||
part_message_id = f"{thinking_id}_{i}"
|
||||
message_segment = Seg(type=type, data=data)
|
||||
|
||||
if type == "emoji":
|
||||
is_emoji = True
|
||||
else:
|
||||
is_emoji = False
|
||||
reply_to = not mark_head
|
||||
|
||||
bot_message: MessageSending = await self._build_single_sending_message(
|
||||
anchor_message=anchor_message,
|
||||
message_id=part_message_id,
|
||||
message_segment=message_segment,
|
||||
display_message=display_message,
|
||||
reply_to=reply_to,
|
||||
is_emoji=is_emoji,
|
||||
thinking_id=thinking_id,
|
||||
thinking_start_time=thinking_start_time,
|
||||
)
|
||||
|
||||
try:
|
||||
if (
|
||||
bot_message.is_private_message()
|
||||
or bot_message.reply.processed_plain_text != "[System Trigger Context]"
|
||||
or mark_head
|
||||
):
|
||||
set_reply = False
|
||||
else:
|
||||
set_reply = True
|
||||
|
||||
if not mark_head:
|
||||
mark_head = True
|
||||
typing = False
|
||||
else:
|
||||
typing = True
|
||||
|
||||
sent_msg = await self.heart_fc_sender.send_message(bot_message, typing=typing, set_reply=set_reply)
|
||||
|
||||
reply_message_ids.append(part_message_id) # 记录我们生成的ID
|
||||
|
||||
sent_msg_list.append((type, sent_msg))
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"{self.log_prefix}发送回复片段 {i} ({part_message_id}) 时失败: {e}")
|
||||
traceback.print_exc()
|
||||
# 这里可以选择是继续发送下一个片段还是中止
|
||||
|
||||
# 在尝试发送完所有片段后，完成原始的 thinking_id 状态
|
||||
try:
|
||||
await self.heart_fc_sender.complete_thinking(chat_id, thinking_id)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"{self.log_prefix}完成思考状态 {thinking_id} 时出错: {e}")
|
||||
|
||||
return sent_msg_list
|
||||
|
||||
async def _build_single_sending_message(
|
||||
self,
|
||||
message_id: str,
|
||||
message_segment: Seg,
|
||||
reply_to: bool,
|
||||
is_emoji: bool,
|
||||
thinking_start_time: float,
|
||||
display_message: str,
|
||||
anchor_message: MessageRecv = None,
|
||||
) -> MessageSending:
|
||||
"""构建单个发送消息"""
|
||||
|
||||
bot_user_info = UserInfo(
|
||||
user_id=global_config.bot.qq_account,
|
||||
user_nickname=global_config.bot.nickname,
|
||||
platform=self.chat_stream.platform,
|
||||
)
|
||||
|
||||
# await anchor_message.process()
|
||||
if anchor_message:
|
||||
sender_info = anchor_message.message_info.user_info
|
||||
else:
|
||||
sender_info = None
|
||||
|
||||
bot_message = MessageSending(
|
||||
message_id=message_id, # 使用片段的唯一ID
|
||||
chat_stream=self.chat_stream,
|
||||
bot_user_info=bot_user_info,
|
||||
sender_info=sender_info,
|
||||
message_segment=message_segment,
|
||||
reply=anchor_message, # 回复原始锚点
|
||||
is_head=reply_to,
|
||||
is_emoji=is_emoji,
|
||||
thinking_start_time=thinking_start_time, # 传递原始思考开始时间
|
||||
display_message=display_message,
|
||||
)
|
||||
|
||||
return bot_message
|
||||
|
||||
|
||||
def weighted_sample_no_replacement(items, weights, k) -> list:
|
||||
"""
|
||||
加权且不放回地随机抽取k个元素。
|
||||
|
||||
参数：
|
||||
items: 待抽取的元素列表
|
||||
weights: 每个元素对应的权重（与items等长，且为正数）
|
||||
k: 需要抽取的元素个数
|
||||
返回：
|
||||
selected: 按权重加权且不重复抽取的k个元素组成的列表
|
||||
|
||||
如果 items 中的元素不足 k 个，就只会返回所有可用的元素
|
||||
|
||||
实现思路：
|
||||
每次从当前池中按权重加权随机选出一个元素，选中后将其从池中移除，重复k次。
|
||||
这样保证了：
|
||||
1. count越大被选中概率越高
|
||||
2. 不会重复选中同一个元素
|
||||
"""
|
||||
selected = []
|
||||
pool = list(zip(items, weights))
|
||||
for _ in range(min(k, len(pool))):
|
||||
total = sum(w for _, w in pool)
|
||||
r = random.uniform(0, total)
|
||||
upto = 0
|
||||
for idx, (item, weight) in enumerate(pool):
|
||||
upto += weight
|
||||
if upto >= r:
|
||||
selected.append(item)
|
||||
pool.pop(idx)
|
||||
break
|
||||
return selected
|
||||
|
||||
|
||||
init_prompt()
|
||||
Reference in New Issue
Block a user