2026-07-09 11:47:16 +08:00
|
|
|
|
# =============================================================================
|
|
|
|
|
|
# 企微IT智能服务台 — H5 员工端 AI 回复后台任务
|
|
|
|
|
|
# =============================================================================
|
|
|
|
|
|
# 背景:原 h5_send_message 在同步 HTTP 请求内 await AI 推理(Dify 3~15s),
|
|
|
|
|
|
# 整条请求被阻塞,前端表现为"发送中"长时间卡顿。
|
|
|
|
|
|
# 本模块将 AI 推理移出请求,改为 asyncio 后台任务,结果经 WebSocket
|
|
|
|
|
|
# 流式推回(ai_reply_chunk / ai_reply),发送瞬时完成。
|
|
|
|
|
|
#
|
|
|
|
|
|
# 关键约束(详见 docs/02-需求分析/技术架构演进/员工端消息发送延时改造方案.md):
|
|
|
|
|
|
# 1. 必须单 worker 运行(docker-compose --workers 1):
|
|
|
|
|
|
# ws_manager 是进程内单例,多 worker 时后台任务与员工 WS 连接可能不在
|
|
|
|
|
|
# 同进程,broadcast 会静默丢失(约 50%)。
|
|
|
|
|
|
# 2. 使用独立 DB session(_get_session_factory),不可复用请求的 db
|
|
|
|
|
|
# (请求返回后该 session 会被关闭)。
|
|
|
|
|
|
# =============================================================================
|
|
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
import asyncio
|
2026-07-09 11:47:16 +08:00
|
|
|
|
import logging
|
2026-07-13 02:17:03 +08:00
|
|
|
|
import os
|
|
|
|
|
|
from datetime import datetime, timedelta
|
|
|
|
|
|
from pathlib import Path
|
|
|
|
|
|
|
|
|
|
|
|
from sqlalchemy import select
|
2026-07-09 11:47:16 +08:00
|
|
|
|
|
2026-07-11 23:13:10 +08:00
|
|
|
|
from app.api.byod import _byod_keyword_prefilter
|
2026-07-09 11:47:16 +08:00
|
|
|
|
from app.database import _get_session_factory
|
|
|
|
|
|
from app.dependencies import get_shared_ai_handler
|
|
|
|
|
|
from app.models.conversation import Conversation
|
|
|
|
|
|
from app.models.message import Message
|
2026-07-11 23:13:10 +08:00
|
|
|
|
from app.services.routing_service import (
|
|
|
|
|
|
routing_keyword_prefilter,
|
|
|
|
|
|
detect_routing_intent,
|
|
|
|
|
|
get_contact_by_category,
|
|
|
|
|
|
send_contact_card,
|
|
|
|
|
|
record_routing_event,
|
|
|
|
|
|
_keyword_fallback_category,
|
|
|
|
|
|
)
|
2026-07-13 02:17:03 +08:00
|
|
|
|
from app.services.vision_service import VisionService
|
2026-07-09 11:47:16 +08:00
|
|
|
|
from app.services.ws_manager import manager as ws_manager
|
|
|
|
|
|
|
|
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
# =============================================================================
|
|
|
|
|
|
# Phase 4A: VisionService 接入 — 图片消息视觉理解
|
|
|
|
|
|
# =============================================================================
|
|
|
|
|
|
|
|
|
|
|
|
# 图片文件本地存储根目录(与 upload.py 中 UPLOAD_DIR 一致)
|
|
|
|
|
|
_UPLOAD_DIR = Path(os.getenv("UPLOAD_DIR", "./uploads"))
|
|
|
|
|
|
|
|
|
|
|
|
# 视觉理解置信度阈值:低于此值不注入描述(避免错误描述误导 AI)
|
|
|
|
|
|
_VISION_CONFIDENCE_THRESHOLD = 0.6
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def _media_url_to_local_path(media_url: str) -> Path:
|
|
|
|
|
|
"""将媒体 URL 路径转换为本地文件系统路径。
|
|
|
|
|
|
|
|
|
|
|
|
做什么:把 "/api/media/2026/07/13/abc.png" 转换为
|
|
|
|
|
|
"./uploads/2026/07/13/abc.png"
|
|
|
|
|
|
为什么:VisionService 需要读取原始图片字节流,
|
|
|
|
|
|
而媒体 URL 是 HTTP 访问路径,不是文件系统路径。
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
media_url: 媒体文件 URL(如 /api/media/2026/07/13/abc.png)
|
|
|
|
|
|
|
|
|
|
|
|
Returns:
|
|
|
|
|
|
Path: 本地文件路径对象
|
|
|
|
|
|
"""
|
|
|
|
|
|
# 去掉 URL 前缀 /api/media/,拼接到 UPLOAD_DIR
|
|
|
|
|
|
# 例: "/api/media/2026/07/13/abc.png" → "2026/07/13/abc.png"
|
|
|
|
|
|
relative = media_url.replace("/api/media/", "", 1)
|
|
|
|
|
|
return _UPLOAD_DIR / relative
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
async def _fetch_recent_employee_text(
|
|
|
|
|
|
db, conversation_id: str, employee_id: str, within_seconds: int = 5
|
|
|
|
|
|
) -> str:
|
|
|
|
|
|
"""获取最近 N 秒内员工的文字消息(Phase 4B 消息融合)。
|
|
|
|
|
|
|
|
|
|
|
|
做什么:查询同一会话中,当前图片消息之前 within_seconds 秒内,
|
|
|
|
|
|
员工发送的文本消息内容。
|
|
|
|
|
|
为什么:用户经常先打字描述问题再发截图,或先发截图再补充文字。
|
|
|
|
|
|
将文字与图片视觉描述融合后一次性传给 Dify,
|
|
|
|
|
|
避免 AI 分别处理两条消息导致上下文割裂。
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
db: 异步 DB session
|
|
|
|
|
|
conversation_id: 会话 ID
|
|
|
|
|
|
employee_id: 员工企微 UserID
|
|
|
|
|
|
within_seconds: 时间窗口(秒),默认 5 秒
|
|
|
|
|
|
|
|
|
|
|
|
Returns:
|
|
|
|
|
|
str: 最近的员工文字消息内容(多条用换行拼接),无则返回空字符串
|
|
|
|
|
|
"""
|
|
|
|
|
|
cutoff = datetime.now() - timedelta(seconds=within_seconds)
|
|
|
|
|
|
stmt = (
|
|
|
|
|
|
select(Message)
|
|
|
|
|
|
.where(
|
|
|
|
|
|
Message.conversation_id == conversation_id,
|
|
|
|
|
|
Message.sender_type == "employee",
|
|
|
|
|
|
Message.sender_id == employee_id,
|
|
|
|
|
|
Message.msg_type == "text",
|
|
|
|
|
|
Message.created_at >= cutoff,
|
|
|
|
|
|
)
|
|
|
|
|
|
.order_by(Message.created_at.desc())
|
|
|
|
|
|
.limit(3) # 最多取 3 条,避免内容过长
|
|
|
|
|
|
)
|
|
|
|
|
|
result = await db.execute(stmt)
|
|
|
|
|
|
messages = result.scalars().all()
|
|
|
|
|
|
|
|
|
|
|
|
if not messages:
|
|
|
|
|
|
return ""
|
|
|
|
|
|
|
|
|
|
|
|
# 按时间正序拼接(先发的在前)
|
|
|
|
|
|
texts = [m.content for m in reversed(messages) if m.content]
|
|
|
|
|
|
return "\n".join(texts)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
async def _enrich_image_content(
|
|
|
|
|
|
db,
|
|
|
|
|
|
media_url: str,
|
|
|
|
|
|
original_content: str,
|
|
|
|
|
|
conversation_id: str,
|
|
|
|
|
|
employee_id: str,
|
|
|
|
|
|
) -> str:
|
|
|
|
|
|
"""用 VisionService 分析图片,生成增强后的消息内容。
|
|
|
|
|
|
|
|
|
|
|
|
做什么:
|
|
|
|
|
|
1. 从本地文件系统读取图片
|
|
|
|
|
|
2. 调用 VisionService.analyze_screenshot() 获取视觉描述
|
|
|
|
|
|
3. 查询最近 5 秒内的员工文字消息(消息融合)
|
|
|
|
|
|
4. 拼接视觉描述 + 用户文字 → 传给 Dify
|
|
|
|
|
|
|
|
|
|
|
|
为什么:Dify 文本模型无法直接"看"图片,需要先将图片转为
|
|
|
|
|
|
文字描述,再与用户输入融合后传给 Dify 推理。
|
|
|
|
|
|
|
|
|
|
|
|
降级策略:
|
|
|
|
|
|
- 图片文件不存在 → 返回原始 content
|
|
|
|
|
|
- VisionService 调用失败 → 返回 "我收到了您的截图,但暂时无法识别内容"
|
|
|
|
|
|
- 置信度 < 0.6 → 不注入视觉描述,仅使用用户文字
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
db: 异步 DB session
|
|
|
|
|
|
media_url: 图片 URL(如 /api/media/2026/07/13/abc.png)
|
|
|
|
|
|
original_content: 原始消息内容(如 "[图片] 截图")
|
|
|
|
|
|
conversation_id: 会话 ID
|
|
|
|
|
|
employee_id: 员工企微 UserID
|
|
|
|
|
|
|
|
|
|
|
|
Returns:
|
|
|
|
|
|
str: 增强后的消息内容(视觉描述 + 用户文字)
|
|
|
|
|
|
"""
|
|
|
|
|
|
# 1. 读取本地图片文件
|
|
|
|
|
|
local_path = _media_url_to_local_path(media_url)
|
|
|
|
|
|
if not local_path.exists():
|
|
|
|
|
|
logger.warning(f"图片文件不存在: {local_path} (media_url={media_url})")
|
|
|
|
|
|
return original_content
|
|
|
|
|
|
|
|
|
|
|
|
try:
|
|
|
|
|
|
image_bytes = local_path.read_bytes()
|
|
|
|
|
|
except Exception as e:
|
|
|
|
|
|
logger.error(f"读取图片文件失败: {local_path} - {e}")
|
|
|
|
|
|
return original_content
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 调用 VisionService 分析截图
|
|
|
|
|
|
vision_service = VisionService()
|
|
|
|
|
|
try:
|
|
|
|
|
|
result = await vision_service.analyze_screenshot(
|
|
|
|
|
|
image_bytes, conversation_id
|
|
|
|
|
|
)
|
|
|
|
|
|
description = result.get("description", "")
|
|
|
|
|
|
confidence = result.get("confidence", 0.0)
|
|
|
|
|
|
|
|
|
|
|
|
logger.info(
|
|
|
|
|
|
f"VisionService 分析完成: conversation={conversation_id}, "
|
|
|
|
|
|
f"confidence={confidence:.2f}, desc_len={len(description)}"
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 注入视觉描述到会话上下文(供后续多轮对话使用)
|
|
|
|
|
|
if description and confidence >= _VISION_CONFIDENCE_THRESHOLD:
|
|
|
|
|
|
await vision_service.inject_to_conversation_context(
|
|
|
|
|
|
description, conversation_id
|
|
|
|
|
|
)
|
|
|
|
|
|
except Exception as e:
|
|
|
|
|
|
logger.error(f"VisionService 调用异常: {e}")
|
|
|
|
|
|
description = ""
|
|
|
|
|
|
confidence = 0.0
|
|
|
|
|
|
finally:
|
|
|
|
|
|
await vision_service.close()
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 消息融合:查询最近 5 秒内员工的文字消息
|
|
|
|
|
|
recent_text = await _fetch_recent_employee_text(
|
|
|
|
|
|
db, conversation_id, employee_id, within_seconds=5
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 5. 拼接增强内容
|
|
|
|
|
|
# 格式:[视觉描述] + [用户最近文字] + [原始消息内容]
|
|
|
|
|
|
parts = []
|
|
|
|
|
|
|
|
|
|
|
|
if description and confidence >= _VISION_CONFIDENCE_THRESHOLD:
|
|
|
|
|
|
parts.append(f"[用户发送了截图,视觉理解结果] {description}")
|
|
|
|
|
|
|
|
|
|
|
|
if recent_text:
|
|
|
|
|
|
parts.append(f"[用户最近的文字描述] {recent_text}")
|
|
|
|
|
|
|
|
|
|
|
|
# 原始内容如果不是纯占位符(如"[图片] 截图"),也加入
|
|
|
|
|
|
if original_content and not original_content.startswith("[图片]"):
|
|
|
|
|
|
parts.append(original_content)
|
|
|
|
|
|
|
|
|
|
|
|
if not parts:
|
|
|
|
|
|
# 降级:视觉分析失败且无文字补充
|
|
|
|
|
|
return "我收到了您的截图,但暂时无法识别内容,请描述一下您遇到的问题。"
|
|
|
|
|
|
|
|
|
|
|
|
return "\n".join(parts)
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-07-09 11:47:16 +08:00
|
|
|
|
async def _persist_and_push(
|
|
|
|
|
|
db,
|
|
|
|
|
|
conversation: Conversation,
|
|
|
|
|
|
employee_id: str,
|
|
|
|
|
|
content: str,
|
|
|
|
|
|
is_guidance: bool,
|
|
|
|
|
|
should_count: bool,
|
|
|
|
|
|
should_transfer: bool,
|
|
|
|
|
|
dify_conversation_id,
|
|
|
|
|
|
):
|
|
|
|
|
|
"""持久化 AI 回复并推送给员工端 + 广播坐席端。
|
|
|
|
|
|
|
|
|
|
|
|
做什么:
|
|
|
|
|
|
1. 存 AI 消息到 DB
|
|
|
|
|
|
2. 更新会话状态(dify 上下文 / 计数 / 转人工)
|
|
|
|
|
|
3. 经 WS 向员工推 ai_reply 终态(前端据此替换打字机气泡)
|
|
|
|
|
|
4. 经 WS 向坐席端广播 new_message + conversation_updated
|
|
|
|
|
|
为什么:把"落库 + 推送"封装为单点,供同步路径与流式路径复用。
|
|
|
|
|
|
"""
|
|
|
|
|
|
# 1. 存 AI 消息
|
|
|
|
|
|
ai_message = Message(
|
|
|
|
|
|
conversation_id=conversation.id,
|
|
|
|
|
|
sender_type="ai",
|
|
|
|
|
|
sender_id="ai_bot",
|
|
|
|
|
|
sender_name="Duckula(达寇拉)",
|
|
|
|
|
|
content=content,
|
|
|
|
|
|
msg_type="text",
|
|
|
|
|
|
is_read=True,
|
|
|
|
|
|
)
|
|
|
|
|
|
db.add(ai_message)
|
|
|
|
|
|
await db.flush()
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 更新会话状态
|
|
|
|
|
|
if dify_conversation_id:
|
|
|
|
|
|
conversation.dify_conversation_id = dify_conversation_id
|
|
|
|
|
|
if should_count:
|
|
|
|
|
|
conversation.ai_substantive_reply_count += 1
|
|
|
|
|
|
if should_transfer:
|
|
|
|
|
|
conversation.status = "queued"
|
|
|
|
|
|
conversation.updated_at = datetime.now()
|
|
|
|
|
|
db.add(conversation)
|
|
|
|
|
|
await db.flush()
|
|
|
|
|
|
await db.commit()
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 推 ai_reply 终态给员工(前端替换打字机气泡)
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_reply",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"sender_type": "ai",
|
|
|
|
|
|
"sender_id": "ai_bot",
|
|
|
|
|
|
"sender_name": "Duckula(达寇拉)",
|
|
|
|
|
|
"content": content,
|
|
|
|
|
|
"msg_type": "text",
|
|
|
|
|
|
"is_guidance": is_guidance,
|
|
|
|
|
|
"ai_reply_count": conversation.ai_substantive_reply_count,
|
|
|
|
|
|
"can_call_agent": conversation.ai_substantive_reply_count >= 3,
|
|
|
|
|
|
"conversation_status": conversation.status,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 广播坐席端(new_message + conversation_updated)
|
|
|
|
|
|
try:
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "new_message",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"sender_type": "ai",
|
|
|
|
|
|
"sender_id": "ai_bot",
|
|
|
|
|
|
"sender_name": "Duckula(达寇拉)",
|
|
|
|
|
|
"content": content,
|
|
|
|
|
|
"msg_type": "text",
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "conversation_updated",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"status": conversation.status,
|
|
|
|
|
|
"assigned_agent_id": str(conversation.assigned_agent_id) if conversation.assigned_agent_id else None,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
except Exception as ws_err:
|
|
|
|
|
|
# WS 广播失败不阻塞消息存储,只记录 warning
|
|
|
|
|
|
logger.warning(f"WS 广播 AI 回复给坐席失败(消息已存储): {ws_err}")
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
async def _persist_and_push_structured(
|
|
|
|
|
|
db,
|
|
|
|
|
|
conversation: Conversation,
|
|
|
|
|
|
employee_id: str,
|
|
|
|
|
|
result: dict,
|
|
|
|
|
|
):
|
|
|
|
|
|
"""持久化结构化 AI 回复并推送给员工端 + 广播坐席端(v2.0 双 WS 通道)。
|
|
|
|
|
|
|
|
|
|
|
|
改造后的核心变化(2026-07-13):
|
|
|
|
|
|
- Dify 返回 JSON {text, action, options},后端解析后同时发两条 WS:
|
|
|
|
|
|
① ai_reply → 聊天气泡(text + options)
|
|
|
|
|
|
② dynamic_recommend → 侧边栏推荐(action 卡片)
|
|
|
|
|
|
- 两条消息同一时刻发出,零时间差到达
|
|
|
|
|
|
- 文字明确引用侧边栏内容(如"右侧已为您准备好入口"),语义强关联
|
|
|
|
|
|
|
|
|
|
|
|
命中判断规则:
|
|
|
|
|
|
- 结构化回复且有 action 或 options → 视为命中(AI 在主动引导)
|
|
|
|
|
|
- 纯文本回复 → 走原有 _check_knowledge_hit 判断
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
db: 异步 DB session
|
|
|
|
|
|
conversation: 当前会话对象
|
|
|
|
|
|
employee_id: 员工企微 UserID
|
|
|
|
|
|
result: get_structured_reply() 返回的结构化结果
|
|
|
|
|
|
"""
|
|
|
|
|
|
text = result.get("text", "")
|
|
|
|
|
|
action = result.get("action")
|
|
|
|
|
|
options = result.get("options")
|
|
|
|
|
|
hit = result.get("hit", False)
|
|
|
|
|
|
is_structured = result.get("is_structured", False)
|
|
|
|
|
|
dify_conv_id = result.get("conversation_id")
|
|
|
|
|
|
# Phase 6A: 提取诊断阶段
|
|
|
|
|
|
diagnosis_stage = result.get("diagnosis_stage")
|
|
|
|
|
|
|
|
|
|
|
|
# 结构化回复且有 action 或 options → 视为命中(AI 在主动引导/推荐)
|
|
|
|
|
|
if is_structured and (action or options):
|
|
|
|
|
|
hit = True
|
|
|
|
|
|
|
|
|
|
|
|
# Phase 6A: 基于 diagnosis_stage 调整会话状态
|
|
|
|
|
|
# escalating → AI 建议转人工
|
|
|
|
|
|
# resolved → AI 认为问题已解决
|
|
|
|
|
|
if diagnosis_stage == "escalating":
|
|
|
|
|
|
hit = False # 不计为有效回复,触发转人工
|
|
|
|
|
|
elif diagnosis_stage == "resolved":
|
|
|
|
|
|
hit = True # 计为有效回复
|
|
|
|
|
|
|
|
|
|
|
|
should_count = hit
|
|
|
|
|
|
should_transfer = not hit
|
|
|
|
|
|
|
|
|
|
|
|
# 确定消息类型
|
|
|
|
|
|
if is_structured and (options or action):
|
|
|
|
|
|
msg_type = "ai_structured"
|
|
|
|
|
|
else:
|
|
|
|
|
|
msg_type = "text"
|
|
|
|
|
|
|
|
|
|
|
|
# 构建 extra_data(存储 options 和 action 供前端渲染)
|
|
|
|
|
|
extra_data = {}
|
|
|
|
|
|
if options:
|
|
|
|
|
|
extra_data["options"] = options
|
|
|
|
|
|
if action:
|
|
|
|
|
|
extra_data["action"] = action
|
|
|
|
|
|
|
|
|
|
|
|
# 1. 存 AI 消息
|
|
|
|
|
|
ai_message = Message(
|
|
|
|
|
|
conversation_id=conversation.id,
|
|
|
|
|
|
sender_type="ai",
|
|
|
|
|
|
sender_id="ai_bot",
|
|
|
|
|
|
sender_name="Duckula(达寇拉)",
|
|
|
|
|
|
content=text,
|
|
|
|
|
|
msg_type=msg_type,
|
|
|
|
|
|
extra_data=extra_data if extra_data else None,
|
|
|
|
|
|
is_read=True,
|
|
|
|
|
|
)
|
|
|
|
|
|
db.add(ai_message)
|
|
|
|
|
|
await db.flush()
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 更新会话状态
|
|
|
|
|
|
if dify_conv_id:
|
|
|
|
|
|
conversation.dify_conversation_id = dify_conv_id
|
|
|
|
|
|
if should_count:
|
|
|
|
|
|
conversation.ai_substantive_reply_count += 1
|
|
|
|
|
|
if should_transfer:
|
|
|
|
|
|
conversation.status = "queued"
|
|
|
|
|
|
# Phase 6A: 将 diagnosis_stage 存入 tags(无需迁移,利用现有 JSON 字段)
|
|
|
|
|
|
if diagnosis_stage:
|
|
|
|
|
|
tags = conversation.tags or {}
|
|
|
|
|
|
tags["diagnosis_stage"] = diagnosis_stage
|
|
|
|
|
|
tags["diagnosis_updated_at"] = datetime.now().isoformat()
|
|
|
|
|
|
conversation.tags = tags
|
|
|
|
|
|
conversation.updated_at = datetime.now()
|
|
|
|
|
|
db.add(conversation)
|
|
|
|
|
|
await db.flush()
|
|
|
|
|
|
await db.commit()
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 推 ai_reply 给员工端(聊天气泡:text + options)
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_reply",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"sender_type": "ai",
|
|
|
|
|
|
"sender_id": "ai_bot",
|
|
|
|
|
|
"sender_name": "Duckula(达寇拉)",
|
|
|
|
|
|
"content": text,
|
|
|
|
|
|
"msg_type": msg_type,
|
|
|
|
|
|
"extra_data": extra_data if extra_data else None,
|
|
|
|
|
|
"is_guidance": False,
|
|
|
|
|
|
"ai_reply_count": conversation.ai_substantive_reply_count,
|
|
|
|
|
|
"can_call_agent": conversation.ai_substantive_reply_count >= 3,
|
|
|
|
|
|
"conversation_status": conversation.status,
|
|
|
|
|
|
# Phase 6A: 诊断阶段(前端可据此调整 UI/提示)
|
|
|
|
|
|
"diagnosis_stage": diagnosis_stage,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 推 dynamic_recommend 给员工端侧边栏(仅当 action 非空时)
|
|
|
|
|
|
# 与 ai_reply 同一时刻发出 → 零时间差到达
|
|
|
|
|
|
if action:
|
|
|
|
|
|
recommend_data = {
|
|
|
|
|
|
"recommend_id": f"rec_{ai_message.id}",
|
|
|
|
|
|
"card_type": action.get("type", "approval_card"),
|
|
|
|
|
|
"title": action.get("title", ""),
|
|
|
|
|
|
"description": action.get("description", ""),
|
|
|
|
|
|
"approval_type": action.get("approval_type"),
|
|
|
|
|
|
"confidence": action.get("confidence", 0.85),
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
}
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "dynamic_recommend",
|
|
|
|
|
|
"data": recommend_data,
|
|
|
|
|
|
})
|
|
|
|
|
|
logger.info(
|
|
|
|
|
|
f"动态推荐已推送: employee={employee_id}, "
|
|
|
|
|
|
f"card_type={recommend_data['card_type']}, "
|
|
|
|
|
|
f"title={recommend_data['title']}"
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 5. 广播坐席端(new_message + conversation_updated)
|
|
|
|
|
|
try:
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "new_message",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"sender_type": "ai",
|
|
|
|
|
|
"sender_id": "ai_bot",
|
|
|
|
|
|
"sender_name": "Duckula(达寇拉)",
|
|
|
|
|
|
"content": text,
|
|
|
|
|
|
"msg_type": msg_type,
|
|
|
|
|
|
"extra_data": extra_data if extra_data else None,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "conversation_updated",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"status": conversation.status,
|
|
|
|
|
|
"assigned_agent_id": (
|
|
|
|
|
|
str(conversation.assigned_agent_id)
|
|
|
|
|
|
if conversation.assigned_agent_id else None
|
|
|
|
|
|
),
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
except Exception as ws_err:
|
|
|
|
|
|
logger.warning(f"WS 广播结构化 AI 回复给坐席失败(消息已存储): {ws_err}")
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-07-11 23:13:10 +08:00
|
|
|
|
async def _handle_byod_query(db, conversation, employee_id, content):
|
|
|
|
|
|
"""处理 BYOD 自备电脑补贴查询。
|
|
|
|
|
|
|
|
|
|
|
|
在 H5 聊天消息流中拦截 BYOD 关键词后执行资格检查,并以 byod_card
|
|
|
|
|
|
卡片消息形式推送给员工端(前端 MessageBubble 据 msg_type 渲染
|
|
|
|
|
|
ByodSubsidyCard)。
|
|
|
|
|
|
|
|
|
|
|
|
流程:
|
|
|
|
|
|
1. 通过 WecomService 获取员工岗位(position)
|
|
|
|
|
|
2. 与 BYOD 资格清单匹配(_match_position)
|
|
|
|
|
|
3. 创建 byod_card 类型 AI 消息并落库
|
|
|
|
|
|
4. 经 WS 推送 ai_reply 给员工端(携带 extra_data.byod_result)
|
|
|
|
|
|
5. 广播 new_message + conversation_updated 给坐席端(与 _persist_and_push 一致)
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
db: 异步 DB session(process_h5_ai_reply 的 factory session)
|
|
|
|
|
|
conversation: 当前会话对象(Conversation)
|
|
|
|
|
|
employee_id: 员工企微 UserID
|
|
|
|
|
|
content: 用户消息原文(用于日志)
|
|
|
|
|
|
"""
|
|
|
|
|
|
# 延迟导入避免循环依赖(byod 模块注册路由时可能引用 app.main)
|
|
|
|
|
|
from app.api.byod import _match_position, BYOD_APPLICATION_URL, BYOD_NOTES, BYOD_REGISTER_NOTES
|
|
|
|
|
|
from app.services.wecom_service import WecomService
|
|
|
|
|
|
|
|
|
|
|
|
# 1. 获取员工岗位(企微通讯录 API)
|
|
|
|
|
|
position = ""
|
|
|
|
|
|
try:
|
|
|
|
|
|
wecom_service = WecomService()
|
|
|
|
|
|
try:
|
|
|
|
|
|
user_info = await wecom_service.get_user_info(employee_id)
|
|
|
|
|
|
position = user_info.get("position", "")
|
|
|
|
|
|
finally:
|
|
|
|
|
|
await wecom_service.close()
|
|
|
|
|
|
except Exception as e:
|
|
|
|
|
|
logger.error(f"BYOD: 获取员工岗位失败: {e}")
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 岗位匹配(返回: 是否匹配, 匹配岗位, 匹配类别)
|
|
|
|
|
|
eligible, matched_pos, matched_category = _match_position(position)
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 构建 BYOD 结果数据
|
|
|
|
|
|
# 字段与前端 ByodSubsidyCard.vue props 完全一致:
|
|
|
|
|
|
# eligible / position / matched_category / application_url / notes / reason
|
|
|
|
|
|
byod_result = {
|
|
|
|
|
|
"eligible": eligible,
|
|
|
|
|
|
"has_subsidy": eligible,
|
|
|
|
|
|
"position": position,
|
|
|
|
|
|
"matched_category": matched_category,
|
|
|
|
|
|
"application_url": BYOD_APPLICATION_URL, # 所有岗位都提供链接
|
|
|
|
|
|
"notes": BYOD_NOTES if eligible else BYOD_REGISTER_NOTES,
|
|
|
|
|
|
"reason": (
|
|
|
|
|
|
"" if eligible
|
|
|
|
|
|
else f"您的岗位「{position}」不在自备电脑补贴资格清单中,可进行自备电脑登记(无补贴)"
|
|
|
|
|
|
),
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 展示文本(AI 气泡的 content,卡片下方不直接展示,但会话列表/坐席端可见)
|
|
|
|
|
|
if eligible:
|
|
|
|
|
|
display_text = f"您岗位为「{position}」,符合自备电脑补贴申请资格"
|
|
|
|
|
|
else:
|
|
|
|
|
|
display_text = f"您岗位为「{position}」,可进行自备电脑登记(无补贴)"
|
|
|
|
|
|
|
|
|
|
|
|
# 5. 创建 AI 消息(byod_card 类型,携带 byod_result)
|
|
|
|
|
|
ai_message = Message(
|
|
|
|
|
|
conversation_id=conversation.id,
|
|
|
|
|
|
sender_type="ai",
|
|
|
|
|
|
sender_id="ai_bot",
|
|
|
|
|
|
sender_name="Duckula(达寇拉)",
|
|
|
|
|
|
content=display_text,
|
|
|
|
|
|
msg_type="byod_card",
|
|
|
|
|
|
extra_data={"byod_result": byod_result},
|
|
|
|
|
|
is_read=True,
|
|
|
|
|
|
)
|
|
|
|
|
|
db.add(ai_message)
|
|
|
|
|
|
await db.flush()
|
|
|
|
|
|
|
|
|
|
|
|
# 6. 更新会话状态(计数 + 时间,BYOD 视为一次实质性 AI 回复)
|
|
|
|
|
|
conversation.ai_substantive_reply_count += 1
|
|
|
|
|
|
conversation.updated_at = datetime.now()
|
|
|
|
|
|
db.add(conversation)
|
|
|
|
|
|
await db.flush()
|
|
|
|
|
|
await db.commit()
|
|
|
|
|
|
|
|
|
|
|
|
# 7. 推送 ai_reply 给员工端(前端据 msg_type="byod_card" 渲染卡片)
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_reply",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"sender_type": "ai",
|
|
|
|
|
|
"sender_id": "ai_bot",
|
|
|
|
|
|
"sender_name": "Duckula(达寇拉)",
|
|
|
|
|
|
"content": display_text,
|
|
|
|
|
|
"msg_type": "byod_card",
|
|
|
|
|
|
"extra_data": {"byod_result": byod_result},
|
|
|
|
|
|
"is_guidance": False,
|
|
|
|
|
|
"ai_reply_count": conversation.ai_substantive_reply_count,
|
|
|
|
|
|
"can_call_agent": conversation.ai_substantive_reply_count >= 3,
|
|
|
|
|
|
"conversation_status": conversation.status,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
|
|
# 8. 广播坐席端(new_message + conversation_updated,与 _persist_and_push 一致)
|
|
|
|
|
|
try:
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "new_message",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"message_id": str(ai_message.id),
|
|
|
|
|
|
"sender_type": "ai",
|
|
|
|
|
|
"sender_id": "ai_bot",
|
|
|
|
|
|
"sender_name": "Duckula(达寇拉)",
|
|
|
|
|
|
"content": display_text,
|
|
|
|
|
|
"msg_type": "byod_card",
|
|
|
|
|
|
"extra_data": {"byod_result": byod_result},
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "conversation_updated",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": str(conversation.id),
|
|
|
|
|
|
"status": conversation.status,
|
|
|
|
|
|
"assigned_agent_id": (
|
|
|
|
|
|
str(conversation.assigned_agent_id)
|
|
|
|
|
|
if conversation.assigned_agent_id else None
|
|
|
|
|
|
),
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
except Exception as ws_err:
|
|
|
|
|
|
logger.warning(f"BYOD: WS 广播给坐席失败: {ws_err}")
|
|
|
|
|
|
|
|
|
|
|
|
logger.info(
|
|
|
|
|
|
f"BYOD 查询完成: employee_id={employee_id}, position={position}, "
|
|
|
|
|
|
f"eligible={eligible}, matched_category={matched_category}"
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
async def _handle_routing(
|
|
|
|
|
|
db,
|
|
|
|
|
|
conversation: Conversation,
|
|
|
|
|
|
employee_id: str,
|
|
|
|
|
|
content: str,
|
|
|
|
|
|
) -> bool:
|
|
|
|
|
|
"""处理非IT业务路由推荐。
|
|
|
|
|
|
|
|
|
|
|
|
在 H5 聊天消息流中拦截路由关键词后调用 Dify 统一意图识别,
|
|
|
|
|
|
判定为 non_it_routing 且 routing_confidence ≥ 阈值时发送名片三段式消息。
|
|
|
|
|
|
|
|
|
|
|
|
流程:
|
|
|
|
|
|
1. 调用 Dify 统一意图识别(detect_routing_intent)
|
|
|
|
|
|
2. 检查 intent_type == "non_it_routing" && routing_confidence ≥ 阈值
|
|
|
|
|
|
→ YES: 查询联系人 → 发送名片三段式消息 → 记录路由事件 → 返回 True
|
|
|
|
|
|
→ NO: 返回 False(继续走正常 AI 流程)
|
|
|
|
|
|
3. Dify 调用失败 → 关键词降级兜底(按 ROUTING_KEYWORD_TO_CATEGORY 映射)
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
db: 异步 DB session(process_h5_ai_reply 的 factory session)
|
|
|
|
|
|
conversation: 当前会话对象(Conversation)
|
|
|
|
|
|
employee_id: 员工企微 UserID
|
|
|
|
|
|
content: 用户消息原文
|
|
|
|
|
|
|
|
|
|
|
|
Returns:
|
|
|
|
|
|
bool: True 表示已发送路由名片(应 return 中断后续流程),
|
|
|
|
|
|
False 表示未触发路由(继续走正常 AI 流程)
|
|
|
|
|
|
"""
|
|
|
|
|
|
from app.config import settings
|
|
|
|
|
|
|
|
|
|
|
|
# 1. 调用 Dify 统一意图识别
|
|
|
|
|
|
try:
|
|
|
|
|
|
result = await detect_routing_intent(content, employee_id)
|
|
|
|
|
|
intent_type = result.get("intent_type", "chitchat")
|
|
|
|
|
|
business_category = result.get("business_category")
|
|
|
|
|
|
routing_confidence = result.get("routing_confidence", 0.0)
|
|
|
|
|
|
|
|
|
|
|
|
# 如果是审批意图,不拦截(让审批流程处理)
|
|
|
|
|
|
if intent_type == "approval":
|
|
|
|
|
|
return False
|
|
|
|
|
|
|
|
|
|
|
|
logger.info(
|
|
|
|
|
|
f"路由意图检测(Dify): intent_type={intent_type}, "
|
|
|
|
|
|
f"business_category={business_category}, "
|
|
|
|
|
|
f"routing_confidence={routing_confidence}"
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 检查是否触发路由推荐
|
|
|
|
|
|
threshold = settings.routing_confidence_threshold
|
|
|
|
|
|
if intent_type != "non_it_routing" or routing_confidence < threshold:
|
|
|
|
|
|
# 置信度不足或非路由意图,走正常 AI 流程
|
|
|
|
|
|
return False
|
|
|
|
|
|
|
|
|
|
|
|
if not business_category:
|
|
|
|
|
|
logger.warning("路由意图为 non_it_routing 但 business_category 为空,跳过")
|
|
|
|
|
|
return False
|
|
|
|
|
|
|
|
|
|
|
|
except Exception as e:
|
|
|
|
|
|
logger.warning(f"Dify 路由意图识别失败,降级为关键词匹配: {e}")
|
|
|
|
|
|
# 3. 降级为关键词匹配
|
|
|
|
|
|
business_category = _keyword_fallback_category(content)
|
|
|
|
|
|
if not business_category:
|
|
|
|
|
|
# 关键词也未命中,走正常 AI 流程
|
|
|
|
|
|
return False
|
|
|
|
|
|
routing_confidence = 0.75 # 降级兜底给一个略高于阈值的置信度
|
|
|
|
|
|
logger.info(f"路由意图检测(兜底): business_category={business_category}")
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 查询联系人
|
|
|
|
|
|
contact = await get_contact_by_category(db, business_category)
|
|
|
|
|
|
if not contact:
|
|
|
|
|
|
logger.warning(f"未找到 {business_category} 类别的联系人,跳过路由推荐")
|
|
|
|
|
|
return False
|
|
|
|
|
|
|
|
|
|
|
|
# 5. 构建路由说明文本
|
|
|
|
|
|
category_display = business_category.replace("行政-物业", "物业")
|
|
|
|
|
|
reason = (
|
|
|
|
|
|
f"您的问题属于{category_display}业务范畴,不在IT服务台服务范围内 😊\n\n"
|
|
|
|
|
|
f"为您推荐{category_display}服务相关联系人,您可以直接点击名片联系TA:"
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 6. 发送名片三段式消息
|
|
|
|
|
|
await send_contact_card(
|
|
|
|
|
|
db=db,
|
|
|
|
|
|
conversation=conversation,
|
|
|
|
|
|
employee_id=employee_id,
|
|
|
|
|
|
contact=contact,
|
|
|
|
|
|
reason=reason,
|
|
|
|
|
|
business_category=business_category,
|
|
|
|
|
|
routing_confidence=routing_confidence,
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
# 7. 记录路由事件(P1)
|
|
|
|
|
|
await record_routing_event(
|
|
|
|
|
|
db=db,
|
|
|
|
|
|
conversation_id=str(conversation.id),
|
|
|
|
|
|
employee_id=employee_id,
|
|
|
|
|
|
message_content=content,
|
|
|
|
|
|
business_category=business_category,
|
|
|
|
|
|
routing_confidence=routing_confidence,
|
|
|
|
|
|
contact=contact,
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
return True
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-07-09 11:47:16 +08:00
|
|
|
|
async def process_h5_ai_reply(
|
|
|
|
|
|
conversation_id: str,
|
|
|
|
|
|
employee_id: str,
|
|
|
|
|
|
content: str,
|
|
|
|
|
|
dify_conversation_id=None,
|
2026-07-13 02:17:03 +08:00
|
|
|
|
msg_type: str = "text",
|
|
|
|
|
|
media_url: str = None,
|
2026-07-09 11:47:16 +08:00
|
|
|
|
):
|
|
|
|
|
|
"""H5 发送消息后的 AI 回复处理(asyncio.create_task 入口)。
|
|
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
v2.0 改造(2026-07-13):
|
|
|
|
|
|
- AI 回复从流式 SSE 改为 blocking + JSON 结构化输出
|
|
|
|
|
|
- Dify 返回 {text, action, options} JSON → 后端解析 → 双 WS 推送
|
|
|
|
|
|
- 聊天气泡收到 ai_reply(text + options),侧边栏收到 dynamic_recommend(action)
|
|
|
|
|
|
- 新增 ai_thinking 指示器,用户发送后立即看到"正在思考..."
|
|
|
|
|
|
|
|
|
|
|
|
v2.1 改造(2026-07-13 Phase 4):
|
|
|
|
|
|
- 新增图片消息处理分支(msg_type=image)
|
|
|
|
|
|
- 图片 → VisionService.analyze_screenshot() → 视觉描述 → 融合到用户文字
|
|
|
|
|
|
- 消息融合:查询最近 5 秒内员工的文字消息,与图片描述合并后传给 Dify
|
|
|
|
|
|
- 降级:VisionService 失败/低置信度 → 使用原始文字或提示用户描述问题
|
|
|
|
|
|
|
2026-07-09 11:47:16 +08:00
|
|
|
|
流程:
|
2026-07-13 02:17:03 +08:00
|
|
|
|
1. BYOD 关键词拦截 → byod_card 卡片
|
|
|
|
|
|
2. 路由关键词拦截 → 名片推荐
|
|
|
|
|
|
3. 本地快判断(打招呼/呼叫人工)→ 同步引导
|
|
|
|
|
|
4. ★ 图片消息处理(Phase 4A)→ VisionService 分析 → 内容增强
|
|
|
|
|
|
5. ★ 结构化 AI 回复(blocking + JSON 解析 + 双 WS 推送)
|
|
|
|
|
|
6. 任意异常 → 推 ai_reply_failed
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
conversation_id: 会话 ID
|
|
|
|
|
|
employee_id: 员工企微 UserID
|
|
|
|
|
|
content: 消息文本内容
|
|
|
|
|
|
dify_conversation_id: Dify 会话 ID(用于多轮上下文)
|
|
|
|
|
|
msg_type: 消息类型(text/image/file),默认 text
|
|
|
|
|
|
media_url: 媒体文件 URL(图片消息时使用)
|
2026-07-09 11:47:16 +08:00
|
|
|
|
"""
|
|
|
|
|
|
ai_handler = get_shared_ai_handler()
|
|
|
|
|
|
factory = _get_session_factory()
|
|
|
|
|
|
async with factory() as db:
|
|
|
|
|
|
try:
|
|
|
|
|
|
conversation = await db.get(Conversation, conversation_id)
|
|
|
|
|
|
if not conversation:
|
|
|
|
|
|
logger.warning(f"后台 AI 任务:会话不存在 {conversation_id}")
|
|
|
|
|
|
return
|
|
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
# === BYOD 关键词拦截(仅文本消息)===
|
|
|
|
|
|
# 图片消息的 content 是占位符(如 "[图片] 截图"),跳过关键词拦截
|
|
|
|
|
|
if msg_type == "text" and _byod_keyword_prefilter(content):
|
2026-07-11 23:13:10 +08:00
|
|
|
|
await _handle_byod_query(db, conversation, employee_id, content)
|
2026-07-13 02:17:03 +08:00
|
|
|
|
return
|
2026-07-11 23:13:10 +08:00
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
# === 业务路由检测(仅文本消息)===
|
|
|
|
|
|
if msg_type == "text" and routing_keyword_prefilter(content):
|
2026-07-11 23:13:10 +08:00
|
|
|
|
routed = await _handle_routing(db, conversation, employee_id, content)
|
|
|
|
|
|
if routed:
|
2026-07-13 02:17:03 +08:00
|
|
|
|
return
|
2026-07-11 23:13:10 +08:00
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
# === 本地快判断:打招呼 / 呼叫人工(仅文本消息)===
|
|
|
|
|
|
if msg_type == "text" and (ai_handler.is_greeting(content) or ai_handler.is_call_human(content)):
|
2026-07-09 11:47:16 +08:00
|
|
|
|
result = await ai_handler.handle_message(
|
|
|
|
|
|
content=content,
|
|
|
|
|
|
dify_conversation_id=dify_conversation_id,
|
|
|
|
|
|
user_id=employee_id,
|
|
|
|
|
|
)
|
|
|
|
|
|
await _persist_and_push(
|
|
|
|
|
|
db, conversation, employee_id, result.content,
|
|
|
|
|
|
result.is_guidance, result.should_count,
|
|
|
|
|
|
result.should_transfer, result.dify_conversation_id,
|
|
|
|
|
|
)
|
|
|
|
|
|
return
|
|
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
# === ★ v2.1 图片消息处理(Phase 4A/4B)===
|
|
|
|
|
|
# 做什么:检测到图片消息 → 调用 VisionService 分析截图 →
|
|
|
|
|
|
# 将视觉描述与用户文字融合 → 传给 Dify 推理
|
|
|
|
|
|
# 为什么:Dify 文本模型无法"看"图片,需要先将图片转为文字描述
|
|
|
|
|
|
# 降级:VisionService 失败 → 使用原始 content 继续流程
|
|
|
|
|
|
enriched_content = content # 默认使用原始内容
|
|
|
|
|
|
if msg_type == "image" and media_url:
|
|
|
|
|
|
logger.info(
|
|
|
|
|
|
f"图片消息检测: conversation={conversation_id}, "
|
|
|
|
|
|
f"media_url={media_url}"
|
|
|
|
|
|
)
|
|
|
|
|
|
try:
|
|
|
|
|
|
enriched_content = await _enrich_image_content(
|
|
|
|
|
|
db=db,
|
|
|
|
|
|
media_url=media_url,
|
|
|
|
|
|
original_content=content,
|
|
|
|
|
|
conversation_id=conversation_id,
|
|
|
|
|
|
employee_id=employee_id,
|
|
|
|
|
|
)
|
|
|
|
|
|
logger.info(
|
|
|
|
|
|
f"图片内容增强完成: original_len={len(content)}, "
|
|
|
|
|
|
f"enriched_len={len(enriched_content)}"
|
|
|
|
|
|
)
|
|
|
|
|
|
except Exception as vision_err:
|
|
|
|
|
|
logger.error(
|
|
|
|
|
|
f"VisionService 处理失败,降级为纯文本: {vision_err}"
|
|
|
|
|
|
)
|
|
|
|
|
|
# 降级:使用原始 content,AI 会收到 "[图片] 截图" 这样的占位符
|
|
|
|
|
|
# Dify 会回复"我收到了您的截图,请描述一下问题"
|
2026-07-09 11:47:16 +08:00
|
|
|
|
|
2026-07-13 02:17:03 +08:00
|
|
|
|
# === ★ v2.0 结构化 AI 回复(替代流式)===
|
|
|
|
|
|
# 1. 立即推送 "正在思考..." 指示器
|
|
|
|
|
|
# 同时推给员工(气泡动画)和坐席(状态指示)
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_thinking",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": conversation_id,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
# 坐席端也通知:AI 正在处理此会话的消息
|
|
|
|
|
|
try:
|
|
|
|
|
|
await ws_manager.broadcast({
|
|
|
|
|
|
"type": "ai_thinking",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": conversation_id,
|
|
|
|
|
|
"employee_id": employee_id,
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
except Exception:
|
|
|
|
|
|
pass # 坐席端通知失败不影响主流程
|
|
|
|
|
|
|
|
|
|
|
|
# 2. 启动延迟 "仍在思考" 后台任务(15 秒后触发)
|
|
|
|
|
|
# 如果 Dify 在 15 秒内返回,此任务会被取消
|
|
|
|
|
|
async def _push_still_thinking():
|
|
|
|
|
|
"""15 秒后推送 "仍在思考" 提示,缓解用户等待焦虑。"""
|
|
|
|
|
|
await asyncio.sleep(15)
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_thinking",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": conversation_id,
|
|
|
|
|
|
"status": "still_thinking",
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
|
|
thinking_task = asyncio.create_task(_push_still_thinking())
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 调用 Dify(blocking 模式 + JSON 解析 + 30 秒硬超时)
|
|
|
|
|
|
# get_structured_reply 内部处理 HTTP 错误和 JSON 解析失败
|
|
|
|
|
|
# asyncio.wait_for 处理 30 秒硬超时 → 建议转人工
|
|
|
|
|
|
# 注意:图片消息使用 enriched_content(视觉描述+用户文字融合)
|
|
|
|
|
|
try:
|
|
|
|
|
|
result = await asyncio.wait_for(
|
|
|
|
|
|
ai_handler.ai_service.get_structured_reply(
|
|
|
|
|
|
message=enriched_content,
|
|
|
|
|
|
conversation_id=dify_conversation_id,
|
|
|
|
|
|
user_id=employee_id,
|
|
|
|
|
|
),
|
|
|
|
|
|
timeout=30,
|
|
|
|
|
|
)
|
|
|
|
|
|
except asyncio.TimeoutError:
|
|
|
|
|
|
# 30 秒硬超时 → 建议转人工
|
|
|
|
|
|
thinking_task.cancel()
|
|
|
|
|
|
logger.warning(f"Dify 30 秒超时: conversation={conversation_id}")
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_reply_failed",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": conversation_id,
|
|
|
|
|
|
"message": "AI 响应时间较长,建议转人工坐席处理。",
|
|
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
return
|
|
|
|
|
|
|
|
|
|
|
|
# 4. 取消 "仍在思考" 任务(Dify 已返回)
|
|
|
|
|
|
thinking_task.cancel()
|
|
|
|
|
|
try:
|
|
|
|
|
|
await thinking_task # 等待 task 真正取消,避免 warning
|
|
|
|
|
|
except asyncio.CancelledError:
|
|
|
|
|
|
pass
|
|
|
|
|
|
|
|
|
|
|
|
# 5. 持久化 + 双 WS 推送(ai_reply + dynamic_recommend)
|
|
|
|
|
|
await _persist_and_push_structured(
|
|
|
|
|
|
db, conversation, employee_id, result,
|
2026-07-09 11:47:16 +08:00
|
|
|
|
)
|
2026-07-13 02:17:03 +08:00
|
|
|
|
|
2026-07-09 11:47:16 +08:00
|
|
|
|
except Exception as e:
|
|
|
|
|
|
logger.error(f"后台 AI 任务异常: {e}", exc_info=True)
|
|
|
|
|
|
try:
|
|
|
|
|
|
await ws_manager.broadcast_to_employees([employee_id], {
|
|
|
|
|
|
"type": "ai_reply_failed",
|
|
|
|
|
|
"data": {
|
|
|
|
|
|
"conversation_id": conversation_id,
|
2026-07-13 02:17:03 +08:00
|
|
|
|
"message": "AI 服务异常,请转人工坐席或稍后重试。",
|
2026-07-09 11:47:16 +08:00
|
|
|
|
},
|
|
|
|
|
|
})
|
|
|
|
|
|
except Exception:
|
|
|
|
|
|
# 推送失败也无所谓,员工端 3 秒轮询兜底
|
|
|
|
|
|
pass
|