Files
life-echo/api/app/agents/memoir/classification_agent.py
Kevin 786ebf8ae6 refactor(api,expo): 多智能体与会话收敛、回忆录兼容层移除、后端测试集大幅删减
- 对齐「多智能体收敛」与「回忆录 stories-first / markdown-first」方向:收紧运行时契约、
  删除过渡兼容路径与双轨逻辑,并同步更新客户端与文档。

- Chat:以 ChatOrchestrator 为实时编排入口;删除独立 conversation_agent,精简 prompts。
- Memoir:删除 memory_agent;MemoirOrchestrator、classification / story_route 与 prompts 收敛到
  prepare_batches + run_story_pipeline_for_category_batch 主链路。
- 将 agents 侧 processor 迁入 feature 层为 background_runner,并移除 features 下重复/过时
  processor 封装。

- 新增 history_store,强化「conversation_messages 为 DB 真源、Redis 为缓存」模型。
- 调整 models、repo、service、session_history;精简 WS message_types,重构 pipeline 与 router。

- 移除章节占位、整章再生等旧路径;章节列表与封面逻辑要求 story 关联;收紧 cover 资格与
  enqueue。
- helpers、repo、service、router、reading_segment_materialize、story_pipeline_sync、pdf_service
  等按 canonical markdown / cover_asset_id 收缩;删除 memoir_images/provider 等冗余。
- tasks:memoir_tasks、chapter_cover_tasks 等大幅瘦身;story_image_tasks 等与当前图片任务对齐。

- core:config、logging、redis、task_tracker 小幅调整。
- auth / user / payment / quota:路由或服务侧删减过时接口或逻辑(如 payment router 行数减少)。

- pyproject.toml、development.sh、.env.example / .env.production、README 等同步说明或变量。

- Alembic 0001_initial_schema 微调(与当前 schema 叙事一致的小改动)。

- 回忆录:types / mappers / api、章节页与 memoir 页与后端契约对齐;markdown-renderer 调整。
- 语音:删除 voice/player,voice-segment-store 相应精简。

- api/tests:删除 conftest 及绝大部分既有测试文件(websocket_baseline、conversation、memoir
  图片、PDF、SMS 等),属有意收缩/待按 backend-test-system 重建的信号。
- docs:新增多智能体收敛与移除兼容层计划摘要;更新 story-first 设计、backend-test-system、
  multi-agent-refactor-plan、实施总结等。

BREAKING CHANGE: 后端对外契约、回忆录章节字段与若干路由/任务行为已变更;大量 API 测试被移除,
  CI 若依赖这些用例需按新策略补测或调整流水线。
2026-03-22 18:10:28 +08:00

82 lines
2.7 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
"""
ClassificationAgent将内容分类到 8 个章节类别,或判定无价值返回 None。
对应现有逻辑_classify_chapter_category
"""
from __future__ import annotations
from typing import Any, Optional
from app.agents.memoir.prompts import (
CHAPTER_CATEGORIES,
get_chapter_classification_prompt,
)
from app.core.logging import get_logger
logger = get_logger(__name__)
# 5-stage 关键词(用于 LLM 失败时的兜底)
STAGE_KEYWORDS = {
"childhood": ["童年", "小时候", "出生", "家乡", "小镇"],
"education": ["上学", "学校", "老师", "同学", "教育", "大学"],
"career": ["工作", "职业", "事业", "公司", "同事", "创业"],
"family": ["伴侣", "孩子", "家庭", "家人", "结婚", "父母"],
"belief": ["信念", "价值观", "座右铭", "坚持", "原则"],
}
# 5-stage → 默认 8-category 映射LLM 分类失败时的兜底)
_STAGE_TO_DEFAULT_CATEGORY = {
"childhood": "childhood",
"education": "education",
"career": "career_early",
"family": "family",
"belief": "beliefs",
}
def _detect_stage(text: str, fallback_stage: str) -> str:
"""根据关键词检测消息所属的 5-stage 阶段"""
message = (text or "").lower()
for stage, keywords in STAGE_KEYWORDS.items():
if any(word in message for word in keywords):
return stage
return fallback_stage
class ClassificationAgent:
"""将内容分类到 8 个章节类别之一,或判定无价值返回 None"""
def classify(
self,
text: str,
fallback_stage: str,
llm: Any,
) -> Optional[str]:
"""
分类到 8 个章节类别之一。
若 LLM 判定内容无实质回忆录价值,返回 None。
llm 需支持 .invoke(prompt) 同步调用。
"""
if llm:
try:
prompt = get_chapter_classification_prompt(text)
response = llm.invoke(prompt)
category = (response.content or "").strip().lower()
if category == "none":
logger.debug(
"LLM 判定内容无回忆录价值,跳过: text_len=%s text=%s",
len(text or ""),
text or "",
)
return None
if category in CHAPTER_CATEGORIES:
return category
except Exception as e:
logger.warning("ClassificationAgent LLM 章节分类失败: %s", e)
stage = _detect_stage(text, fallback_stage)
return _STAGE_TO_DEFAULT_CATEGORY.get(
stage,
_STAGE_TO_DEFAULT_CATEGORY.get(fallback_stage, "childhood"),
)