Losita/mai-bot
0
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat：优化重复思考，表情现在一次选所有且可配置

Browse Source
This commit is contained in:
SengokuCola
2026-04-04 01:55:05 +08:00
parent 97fb4cb36f
commit eda8ce66f0
4 changed files with 141 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

49
src/chat/replyer/maisaka_generator.py
View File

@@ -17,6 +17,7 @@ from src.common.data_models.reply_generation_data_models import (
)
from src.common.logger import get_logger
from src.common.prompt_i18n import load_prompt
from src.common.utils.utils_session import SessionUtils
from src.config.config import global_config
from src.core.types import ActionInfo
from src.services.llm_service import LLMServiceClient
@@ -285,18 +286,52 @@ class MaisakaReplyGenerator:
)
return block, selected_ids
def _get_related_session_ids(self, session_id: str) -> List[str]:
"""根据表达互通组配置，解析当前会话可共享的会话 ID。"""
related_session_ids = {session_id}
expression_groups = global_config.expression.expression_groups
for expression_group in expression_groups:
target_items = expression_group.expression_groups
group_session_ids: set[str] = set()
contains_current_session = False
for target_item in target_items:
platform = target_item.platform.strip()
item_id = target_item.item_id.strip()
if not platform or not item_id:
continue
rule_type = target_item.rule_type
target_session_id = SessionUtils.calculate_session_id(
platform,
group_id=item_id if rule_type == "group" else None,
user_id=None if rule_type == "group" else item_id,
)
group_session_ids.add(target_session_id)
if target_session_id == session_id:
contains_current_session = True
if contains_current_session:
related_session_ids.update(group_session_ids)
return list(related_session_ids)
def _load_expression_records(self, session_id: str) -> List[_ExpressionRecord]:
"""提取表达方式静态数据，避免 detached ORM 对象。"""
with get_db_session(auto_commit=False) as session:
query = select(Expression).where(Expression.rejected.is_(False)) # type: ignore[attr-defined]
if global_config.expression.expression_checked_only:
query = query.where(Expression.checked.is_(True)) # type: ignore[attr-defined]
related_session_ids = self._get_related_session_ids(session_id)
query = query.where(
(Expression.session_id == session_id) | (Expression.session_id.is_(None)) # type: ignore[attr-defined]
with get_db_session(auto_commit=False) as session:
base_query = select(Expression).where(Expression.rejected.is_(False)) # type: ignore[attr-defined]
scoped_query = base_query.where(
(Expression.session_id.in_(related_session_ids)) | (Expression.session_id.is_(None)) # type: ignore[attr-defined]
).order_by(Expression.count.desc(), Expression.last_active_time.desc()) # type: ignore[attr-defined]
expressions = session.exec(query.limit(5)).all()
if global_config.expression.expression_checked_only:
scoped_query = scoped_query.where(Expression.checked.is_(True)) # type: ignore[attr-defined]
expressions = session.exec(scoped_query.limit(5)).all()
return [
_ExpressionRecord(
expression_id=expression.id,

71
src/chat/replyer/maisaka_generator_multi.py
View File

@@ -17,6 +17,7 @@ from src.common.data_models.reply_generation_data_models import (
)
from src.common.logger import get_logger
from src.common.prompt_i18n import load_prompt
from src.common.utils.utils_session import SessionUtils
from src.config.config import global_config
from src.core.types import ActionInfo
from src.llm_models.payload_content.message import ImageMessagePart, Message, MessageBuilder, RoleType, TextMessagePart
@@ -172,24 +173,15 @@ class MaisakaReplyGenerator:
extra_sections.append(expression_habits.strip())
if target_message_block:
extra_sections.append(target_message_block)
if reply_reason.strip():
extra_sections.append(f"【回复信息参考】\n{reply_reason}")
if not extra_sections:
return system_prompt
return f"{system_prompt}\n\n" + "\n\n".join(extra_sections)
def _build_reply_instruction(
self,
reply_message: Optional[SessionMessage],
reply_reason: str,
) -> str:
def _build_reply_instruction(self) -> str:
"""构建追加在上下文末尾的回复指令。"""
sections: List[str] = []
target_message_block = self._build_target_message_block(reply_message)
if target_message_block:
sections.append(target_message_block)
if reply_reason.strip():
sections.append(f"【回复信息参考】\n{reply_reason}")
sections.append("请基于以上逐条对话消息，自然地继续回复。直接输出你要说的话，不要额外解释。")
return "\n\n".join(sections)
return "请基于以上逐条对话消息，自然地继续回复。直接输出你要说的话，不要额外解释。"
def _build_multimodal_user_message(
self,
@@ -281,10 +273,7 @@ class MaisakaReplyGenerator:
reply_reason=reply_reason,
expression_habits=expression_habits,
)
instruction = self._build_reply_instruction(
reply_message=reply_message,
reply_reason=reply_reason,
)
instruction = self._build_reply_instruction()
messages.append(MessageBuilder().set_role(RoleType.System).add_text_content(system_prompt).build())
messages.extend(self._build_history_messages(chat_history))
@@ -369,18 +358,52 @@ class MaisakaReplyGenerator:
)
return block, selected_ids
def _get_related_session_ids(self, session_id: str) -> List[str]:
"""根据表达互通组配置，解析当前会话可共享的会话 ID。"""
related_session_ids = {session_id}
expression_groups = global_config.expression.expression_groups
for expression_group in expression_groups:
target_items = expression_group.expression_groups
group_session_ids: set[str] = set()
contains_current_session = False
for target_item in target_items:
platform = target_item.platform.strip()
item_id = target_item.item_id.strip()
if not platform or not item_id:
continue
rule_type = target_item.rule_type
target_session_id = SessionUtils.calculate_session_id(
platform,
group_id=item_id if rule_type == "group" else None,
user_id=None if rule_type == "group" else item_id,
)
group_session_ids.add(target_session_id)
if target_session_id == session_id:
contains_current_session = True
if contains_current_session:
related_session_ids.update(group_session_ids)
return list(related_session_ids)
def _load_expression_records(self, session_id: str) -> List[_ExpressionRecord]:
"""提取表达方式静态数据，避免 detached ORM 对象。"""
with get_db_session(auto_commit=False) as session:
query = select(Expression).where(Expression.rejected.is_(False)) # type: ignore[attr-defined]
if global_config.expression.expression_checked_only:
query = query.where(Expression.checked.is_(True)) # type: ignore[attr-defined]
related_session_ids = self._get_related_session_ids(session_id)
query = query.where(
(Expression.session_id == session_id) | (Expression.session_id.is_(None)) # type: ignore[attr-defined]
with get_db_session(auto_commit=False) as session:
base_query = select(Expression).where(Expression.rejected.is_(False)) # type: ignore[attr-defined]
scoped_query = base_query.where(
(Expression.session_id.in_(related_session_ids)) | (Expression.session_id.is_(None)) # type: ignore[attr-defined]
).order_by(Expression.count.desc(), Expression.last_active_time.desc()) # type: ignore[attr-defined]
expressions = session.exec(query.limit(5)).all()
if global_config.expression.expression_checked_only:
scoped_query = scoped_query.where(Expression.checked.is_(True)) # type: ignore[attr-defined]
expressions = session.exec(scoped_query.limit(5)).all()
return [
_ExpressionRecord(
expression_id=expression.id,

15
src/config/official_configs.py
View File

@@ -716,17 +716,16 @@ class EmojiConfig(ConfigBase):
__ui_label__ = "功能"
__ui_icon__ = "puzzle"
emoji_chance: float = Field(
default=0.4,
ge=0,
le=1,
emoji_send_num: int = Field(
default=25,
ge=1,
le=64,
json_schema_extra={
"x-widget": "slider",
"x-icon": "smile",
"step": 0.1,
"x-widget": "input",
"x-icon": "grid",
},
)
"""发送表情包的基础概率"""
"""一次从多少个表情包中选择发送，最大为 64"""
max_reg_num: int = Field(
default=100,

54
src/maisaka/builtin_tool/send_emoji.py
View File

@@ -2,6 +2,7 @@
from datetime import datetime
from io import BytesIO
import math
from random import sample
from typing import Any, Dict, Optional
@@ -16,6 +17,7 @@ from src.chat.emoji_system.maisaka_tool import send_emoji_for_maisaka
from src.common.data_models.image_data_model import MaiEmoji
from src.common.data_models.message_component_data_model import ImageComponent, MessageSequence, TextComponent
from src.common.logger import get_logger
from src.config.config import global_config
from src.core.tooling import ToolExecutionContext, ToolExecutionResult, ToolInvocation, ToolSpec
from src.llm_models.payload_content.resp_format import RespFormat, RespFormatType
from src.maisaka.context_messages import (
@@ -31,7 +33,7 @@ logger = get_logger("maisaka_builtin_send_emoji")
_EMOJI_SUB_AGENT_CONTEXT_LIMIT = 12
_EMOJI_SUB_AGENT_MAX_TOKENS = 240
_EMOJI_CANDIDATE_COUNT = 25
_EMOJI_MAX_CANDIDATE_COUNT = 64
_EMOJI_CANDIDATE_TILE_SIZE = 256
_EMOJI_SUCCESS_MESSAGE = "表情包发送成功"
@@ -70,6 +72,36 @@ async def _load_emoji_bytes(emoji: MaiEmoji) -> bytes:
return await asyncio.to_thread(emoji.full_path.read_bytes)
def _get_emoji_candidate_count() -> int:
"""获取本次表情包候选数量配置。"""
configured_count = int(getattr(global_config.emoji, "emoji_send_num", 25))
return max(1, min(configured_count, _EMOJI_MAX_CANDIDATE_COUNT))
def _calculate_grid_shape(candidate_count: int) -> tuple[int, int]:
"""根据候选数量计算尽量接近矩形的拼图行列数。"""
if candidate_count <= 0:
return 1, 1
best_columns = candidate_count
best_rows = 1
best_score: tuple[int, int] | None = None
for columns in range(1, candidate_count + 1):
rows = math.ceil(candidate_count / columns)
empty_slots = rows * columns - candidate_count
aspect_gap = abs(columns - rows)
score = (aspect_gap, empty_slots)
if best_score is None or score < best_score:
best_score = score
best_columns = columns
best_rows = rows
return best_rows, best_columns
def _build_placeholder_tile(label: str, tile_size: int) -> PILImage.Image:
"""构建图片读取失败时使用的占位图。"""
@@ -134,12 +166,12 @@ def _build_labeled_tile(image_bytes: bytes, index: int, tile_size: int) -> PILIm
def _merge_emoji_tiles(image_bytes_list: list[bytes]) -> bytes:
"""将候选表情图拼接成一张 5x5 网格图片。"""
"""将候选表情图拼接成一张尽量接近矩形的网格图片。"""
tile_size = _EMOJI_CANDIDATE_TILE_SIZE
gap = 12
grid_columns = 5
grid_rows = 5
candidate_count = len(image_bytes_list)
grid_rows, grid_columns = _calculate_grid_shape(candidate_count)
tiles = [
_build_labeled_tile(image_bytes=image_bytes, index=index, tile_size=tile_size)
for index, image_bytes in enumerate(image_bytes_list, start=1)
@@ -161,7 +193,7 @@ def _merge_emoji_tiles(image_bytes_list: list[bytes]) -> bytes:
async def _build_emoji_candidate_message(emojis: list[MaiEmoji]) -> SessionBackedMessage:
"""构建供子代理挑选的 5x5 拼图候选消息。"""
"""构建供子代理挑选的拼图候选消息。"""
image_bytes_list = await asyncio.gather(*[_load_emoji_bytes(emoji) for emoji in emojis])
merged_image_bytes = await asyncio.to_thread(_merge_emoji_tiles, list(image_bytes_list))
@@ -195,15 +227,17 @@ async def _select_emoji_with_sub_agent(
if not available_emojis:
return None, ""
total_candidate_count = min(len(available_emojis), _EMOJI_CANDIDATE_COUNT)
total_candidate_count = min(len(available_emojis), _get_emoji_candidate_count())
sampled_emojis = sample(available_emojis, total_candidate_count)
candidate_message = await _build_emoji_candidate_message(sampled_emojis)
grid_rows, grid_columns = _calculate_grid_shape(len(sampled_emojis))
system_prompt = (
"你是 Maisaka 的临时表情包选择子代理。\n"
"你会收到群聊上下文，以及 1 条额外候选消息，其中包含一张 5x5 的表情包拼图，一共 25 个位置。\n"
"每张小图左上角都有一个较大的序号，范围是 1 到 25。\n"
"你的任务是根据上下文和当前语气，从这 25 张图里选出最合适的一张表情包。\n"
f"你会收到群聊上下文，以及 1 条额外候选消息，其中包含一张 {grid_rows}x{grid_columns} 的表情包拼图，"
f"一共 {len(sampled_emojis)} 个位置。\n"
f"每张小图左上角都有一个较大的序号，范围是 1 到 {len(sampled_emojis)}。\n"
f"你的任务是根据上下文和当前语气，从这 {len(sampled_emojis)} 张图里选出最合适的一张表情包。\n"
"如果提供了 requested_emotion，请优先考虑与其接近的候选；如果没有完全匹配，则选择最符合上下文语气的候选。\n"
"你必须返回一个 JSON 对象（json object），不要输出任何 JSON 之外的内容。\n"
'返回格式固定为：{"emoji_index":1,"reason":"简短理由"}'
@@ -212,6 +246,8 @@ async def _select_emoji_with_sub_agent(
content=(
f"[选择任务]\n"
f"requested_emotion: {requested_emotion or '未指定'}\n"
f"候选总数: {len(sampled_emojis)}\n"
f"拼图布局: {grid_rows}x{grid_columns}\n"
"请只输出 JSON。"
),
timestamp=datetime.now(),