hrz
4140981a8a
评审提的几条一起改：观测时间和数据更新时间分开标注，网页那条路也把观测时间解析出来；「今天」按北京时间算，不跟服务器时区走；失败路径不再传那个读不到的参数。 凭据缺失时改成直接回复，服务端日志也留一条带文档路径的告警，开发者看日志能定位。
553 lines
20 KiB
Python
553 lines
20 KiB
Python
import asyncio
|
|
import datetime
|
|
import re
|
|
import httpx
|
|
from bs4 import BeautifulSoup
|
|
from config.logger import setup_logging
|
|
from plugins_func.register import register_function, ToolType, ActionResponse, Action
|
|
from core.utils.util import get_ip_info
|
|
from typing import TYPE_CHECKING
|
|
|
|
if TYPE_CHECKING:
|
|
from core.connection import ConnectionHandler
|
|
|
|
TAG = __name__
|
|
logger = setup_logging()
|
|
|
|
GET_WEATHER_FUNCTION_DESC = {
|
|
"type": "function",
|
|
"function": {
|
|
"name": "get_weather",
|
|
"description": (
|
|
"获取某个地点的天气，用户应提供一个位置，比如用户说杭州天气，参数为：杭州。"
|
|
"如果用户说的是省份，默认用省会城市。如果用户说的不是省份或城市而是一个地名，默认用该地所在省份的省会城市。"
|
|
"重要：本地未来7天天气已在上下文中提供，用户未指明其他城市时绝对不要调用此工具。"
|
|
),
|
|
"parameters": {
|
|
"type": "object",
|
|
"properties": {
|
|
"location": {
|
|
"type": "string",
|
|
"description": "地点名，例如杭州。可选参数，如果不提供则不传",
|
|
},
|
|
"lang": {
|
|
"type": "string",
|
|
"description": "返回用户使用的语言code，例如zh_CN/zh_HK/en_US/ja_JP等，默认zh_CN",
|
|
},
|
|
},
|
|
"required": ["lang"],
|
|
},
|
|
},
|
|
}
|
|
|
|
HEADERS = {
|
|
"User-Agent": (
|
|
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 "
|
|
"(KHTML, like Gecko) Chrome/92.0.4515.107 Safari/537.36"
|
|
)
|
|
}
|
|
|
|
# 仓库里不再预置和风天气的凭据，自己部署的人得自己去申请一个
|
|
WEATHER_CREDENTIAL_HINT = (
|
|
"尚未配置和风天气 api_key，请参考 docs/weather-integration.md 申请后填入"
|
|
)
|
|
WEATHER_HOST_HINT = (
|
|
"尚未配置和风天气 api_host，请参考 docs/weather-integration.md 申请后填入"
|
|
)
|
|
|
|
# 上游挂着的时候每次提问都要把整条链路重走一遍，用户要等二十几秒。
|
|
# 失败结果短时间内不会变，先记下来直接回
|
|
WEATHER_FAILURE_TTL = 300
|
|
|
|
WEEKDAY_NAMES = ["周一", "周二", "周三", "周四", "周五", "周六", "周日"]
|
|
|
|
# 报告里的「今天」按北京时间算。默认部署的服务器时区是 UTC，国内用户北京时间
|
|
# 凌晨到早上八点这段时间服务器还在昨天，「今天」会说错一整天。
|
|
# 时区自己造，不依赖 zoneinfo：容器和 Windows 上不一定有 tzdata
|
|
BEIJING_TZ = datetime.timezone(datetime.timedelta(hours=8))
|
|
|
|
# 网页上写的是符号单位，报告里统一用中文，两条路径读出来的东西才对得上
|
|
UNIT_NAMES = {
|
|
"km/h": "公里/小时",
|
|
"km": "公里",
|
|
"mm": "毫米",
|
|
"hpa": "百帕",
|
|
"°": "℃",
|
|
"°c": "℃",
|
|
}
|
|
|
|
# 网页上的参数名和接口返回的不一样，报告里用接口那套
|
|
ITEM_NAMES = {
|
|
"大气压": "气压",
|
|
"相对湿度": "湿度",
|
|
}
|
|
|
|
# 天气代码 https://dev.qweather.com/docs/resource/icons/#weather-icons
|
|
WEATHER_CODE_MAP = {
|
|
"100": "晴",
|
|
"101": "多云",
|
|
"102": "少云",
|
|
"103": "晴间多云",
|
|
"104": "阴",
|
|
"150": "晴",
|
|
"151": "多云",
|
|
"152": "少云",
|
|
"153": "晴间多云",
|
|
"300": "阵雨",
|
|
"301": "强阵雨",
|
|
"302": "雷阵雨",
|
|
"303": "强雷阵雨",
|
|
"304": "雷阵雨伴有冰雹",
|
|
"305": "小雨",
|
|
"306": "中雨",
|
|
"307": "大雨",
|
|
"308": "极端降雨",
|
|
"309": "毛毛雨/细雨",
|
|
"310": "暴雨",
|
|
"311": "大暴雨",
|
|
"312": "特大暴雨",
|
|
"313": "冻雨",
|
|
"314": "小到中雨",
|
|
"315": "中到大雨",
|
|
"316": "大到暴雨",
|
|
"317": "暴雨到大暴雨",
|
|
"318": "大暴雨到特大暴雨",
|
|
"350": "阵雨",
|
|
"351": "强阵雨",
|
|
"399": "雨",
|
|
"400": "小雪",
|
|
"401": "中雪",
|
|
"402": "大雪",
|
|
"403": "暴雪",
|
|
"404": "雨夹雪",
|
|
"405": "雨雪天气",
|
|
"406": "阵雨夹雪",
|
|
"407": "阵雪",
|
|
"408": "小到中雪",
|
|
"409": "中到大雪",
|
|
"410": "大到暴雪",
|
|
"456": "阵雨夹雪",
|
|
"457": "阵雪",
|
|
"499": "雪",
|
|
"500": "薄雾",
|
|
"501": "雾",
|
|
"502": "霾",
|
|
"503": "扬沙",
|
|
"504": "浮尘",
|
|
"507": "沙尘暴",
|
|
"508": "强沙尘暴",
|
|
"509": "浓雾",
|
|
"510": "强浓雾",
|
|
"511": "中度霾",
|
|
"512": "重度霾",
|
|
"513": "严重霾",
|
|
"514": "大雾",
|
|
"515": "特强浓雾",
|
|
"900": "热",
|
|
"901": "冷",
|
|
"999": "未知",
|
|
}
|
|
|
|
|
|
def format_observation_time(raw_time):
|
|
"""观测时间接口给的是带 T 的写法，报告里换成空格好看点"""
|
|
if not raw_time:
|
|
return None
|
|
return str(raw_time).replace("T", " ", 1)
|
|
|
|
|
|
def beijing_today(now=None):
|
|
"""
|
|
北京时间今天是哪天
|
|
|
|
<p>
|
|
服务器本地时区不能拿来当准：默认部署写的是 UTC，国内用户北京时间凌晨到
|
|
早上八点这段时间，服务器还在昨天，「今天」会说错一整天。
|
|
"""
|
|
if now is None:
|
|
now = datetime.datetime.now(datetime.timezone.utc)
|
|
if now.tzinfo is None:
|
|
now = now.replace(tzinfo=datetime.timezone.utc)
|
|
return now.astimezone(BEIJING_TZ).date()
|
|
|
|
|
|
def format_forecast_date(raw_date, today=None):
|
|
"""
|
|
预报日期统一成 今天/明天/后天 或者 星期加月日
|
|
|
|
<p>
|
|
接口给的是 2026-10-24 这种写法，网页上写的是 今天10月24日，
|
|
两条路径各写各的，用户听不出来是同一份数据。统一成一种写法。
|
|
"""
|
|
if not raw_date:
|
|
return raw_date
|
|
text = str(raw_date).strip()
|
|
today = today or beijing_today()
|
|
|
|
iso = re.fullmatch(r"(\d{4})-(\d{1,2})-(\d{1,2})", text)
|
|
if iso:
|
|
year = int(iso.group(1))
|
|
month = int(iso.group(2))
|
|
day = int(iso.group(3))
|
|
else:
|
|
found = re.search(r"(\d{1,2})月(\d{1,2})日", text)
|
|
if not found:
|
|
return text
|
|
year = today.year
|
|
month = int(found.group(1))
|
|
day = int(found.group(2))
|
|
|
|
try:
|
|
date_value = datetime.date(year, month, day)
|
|
except ValueError:
|
|
return text
|
|
|
|
delta = (date_value - today).days
|
|
if delta == 0:
|
|
prefix = "今天"
|
|
elif delta == 1:
|
|
prefix = "明天"
|
|
elif delta == 2:
|
|
prefix = "后天"
|
|
else:
|
|
prefix = WEEKDAY_NAMES[date_value.weekday()]
|
|
return f"{prefix}{month}月{day}日"
|
|
|
|
|
|
def normalize_unit(value):
|
|
"""把单位符号换成中文写法，换不动就原样留着"""
|
|
if value is None:
|
|
return None
|
|
text = str(value).strip()
|
|
if not text:
|
|
return text
|
|
matched = re.fullmatch(r"(.+?)([a-zA-Z°℃/]+)", text)
|
|
if not matched:
|
|
return text
|
|
unit = matched.group(2).lower()
|
|
if unit not in UNIT_NAMES:
|
|
return text
|
|
return f"{matched.group(1)}{UNIT_NAMES[unit]}"
|
|
|
|
|
|
def normalize_temperature(value):
|
|
"""温度只留数值，单位在报告里统一写一次"""
|
|
if value is None:
|
|
return None
|
|
matched = re.match(r"-?\d+(?:\.\d+)?", str(value).strip())
|
|
return matched.group(0) if matched else None
|
|
|
|
|
|
def normalize_wind_scale(value):
|
|
"""风力统一成 3级 这种带单位的写法"""
|
|
if value is None:
|
|
return None
|
|
text = str(value).strip()
|
|
if not text or text.endswith("级"):
|
|
return text
|
|
return f"{text}级"
|
|
|
|
|
|
async def fetch_city_info(location, api_key, api_host):
|
|
url = f"https://{api_host}/geo/v2/city/lookup"
|
|
async with httpx.AsyncClient(timeout=httpx.Timeout(5.0, connect=3.0)) as client:
|
|
response = await client.get(
|
|
url,
|
|
params={
|
|
"key": api_key,
|
|
"location": location,
|
|
"lang": "zh",
|
|
"number": 1,
|
|
},
|
|
headers=HEADERS,
|
|
)
|
|
try:
|
|
data = response.json()
|
|
except ValueError:
|
|
logger.bind(tag=TAG).error(
|
|
f"获取天气城市失败：HTTP {response.status_code} 返回了非 JSON 内容"
|
|
)
|
|
return None
|
|
if not isinstance(data, dict):
|
|
logger.bind(tag=TAG).error("获取天气城市失败：响应不是 JSON 对象")
|
|
return None
|
|
if response.status_code != 200 or data.get("code") != "200":
|
|
logger.bind(tag=TAG).error(
|
|
f"获取天气城市失败：HTTP {response.status_code}，代码 {data.get('code')}"
|
|
)
|
|
return None
|
|
return data.get("location", [])[0] if data.get("location") else None
|
|
|
|
|
|
async def fetch_weather_data(city_info, api_key, api_host):
|
|
location_id = city_info.get("id")
|
|
if not location_id:
|
|
return None
|
|
|
|
base_url = f"https://{api_host}/v7/weather"
|
|
params = {"key": api_key, "location": location_id, "lang": "zh"}
|
|
|
|
async def request(client, endpoint):
|
|
response = await client.get(f"{base_url}/{endpoint}", params=params, headers=HEADERS)
|
|
try:
|
|
data = response.json()
|
|
except ValueError:
|
|
return None
|
|
if response.status_code != 200 or data.get("code") != "200":
|
|
logger.bind(tag=TAG).error(
|
|
f"获取天气数据失败：接口={endpoint} HTTP {response.status_code}，代码 {data.get('code')}"
|
|
)
|
|
return None
|
|
return data
|
|
|
|
try:
|
|
async with httpx.AsyncClient(timeout=httpx.Timeout(8.0, connect=3.0)) as client:
|
|
now, daily = await asyncio.gather(
|
|
request(client, "now"), request(client, "7d")
|
|
)
|
|
except Exception as e:
|
|
logger.bind(tag=TAG).warning(f"获取天气 JSON 接口失败，将尝试网页接口: {e}")
|
|
return None
|
|
if now is None or daily is None:
|
|
return None
|
|
return now, daily
|
|
|
|
|
|
def format_weather_report(
|
|
city_name,
|
|
current_text,
|
|
current_basic,
|
|
temps_list,
|
|
observed_at=None,
|
|
data_updated_at=None,
|
|
):
|
|
"""两条路径的报告走同一个模板，日期和单位的写法才对得上"""
|
|
weather_report = f"您查询的位置是：{city_name}\n\n当前天气: {current_text}\n"
|
|
if observed_at:
|
|
weather_report += f"观测时间：{observed_at}\n"
|
|
elif data_updated_at:
|
|
# 顶层那个是接口的数据更新时间，不是观测时间，标签得说清楚
|
|
weather_report += f"数据更新时间：{data_updated_at}\n"
|
|
|
|
if current_basic:
|
|
weather_report += "详细参数：\n"
|
|
weather_report += "".join(
|
|
f" · {key}: {value}\n" for key, value in current_basic.items()
|
|
)
|
|
|
|
weather_report += "\n未来7天预报：\n"
|
|
weather_report += "".join(
|
|
f"{date}: {weather}，气温 {low}~{high}℃\n"
|
|
for date, weather, high, low in temps_list
|
|
)
|
|
weather_report += "\n（如需某一天的具体天气，请告诉我日期）"
|
|
return weather_report
|
|
|
|
|
|
def build_weather_report(city_info, now_data, daily_data):
|
|
city_name = city_info.get("name") or city_info.get("adm2") or "未知位置"
|
|
now = now_data.get("now", {})
|
|
|
|
current_basic = {
|
|
"温度": f"{now['temp']}℃" if now.get("temp") is not None else None,
|
|
"体感温度": f"{now['feelsLike']}℃" if now.get("feelsLike") is not None else None,
|
|
"湿度": f"{now['humidity']}%" if now.get("humidity") is not None else None,
|
|
"风向": now.get("windDir"),
|
|
"风力": normalize_wind_scale(now.get("windScale")),
|
|
"能见度": f"{now['vis']}公里" if now.get("vis") is not None else None,
|
|
"气压": f"{now['pressure']}百帕" if now.get("pressure") is not None else None,
|
|
"降水量": f"{now['precip']}毫米" if now.get("precip") is not None else None,
|
|
}
|
|
current_basic = {key: value for key, value in current_basic.items() if value}
|
|
|
|
temps_list = []
|
|
for item in daily_data.get("daily", [])[:7]:
|
|
date = format_forecast_date(item.get("fxDate") or "未知日期")
|
|
weather = item.get("textDay") or WEATHER_CODE_MAP.get(item.get("iconDay"), "未知")
|
|
high = normalize_temperature(item.get("tempMax")) or "未知"
|
|
low = normalize_temperature(item.get("tempMin")) or "未知"
|
|
temps_list.append((date, weather, high, low))
|
|
|
|
# 观测时间在 now.obsTime 里，没有的时候才用响应顶层那个数据更新时间，
|
|
# 两个不是一个意思，标签得分开写
|
|
observed_at = format_observation_time(now.get("obsTime"))
|
|
data_updated_at = format_observation_time(now_data.get("updateTime"))
|
|
return format_weather_report(
|
|
city_name,
|
|
now.get("text", "未知"),
|
|
current_basic,
|
|
temps_list,
|
|
observed_at,
|
|
data_updated_at,
|
|
)
|
|
|
|
|
|
async def fetch_weather_page(url):
|
|
# 网页只是兜底，等太久不值得，接口那边不通的话很快就该告诉用户
|
|
async with httpx.AsyncClient(
|
|
timeout=httpx.Timeout(5.0, connect=3.0), follow_redirects=True
|
|
) as client:
|
|
response = await client.get(url, headers=HEADERS)
|
|
return BeautifulSoup(response.text, "html.parser") if response.status_code == 200 else None
|
|
|
|
|
|
def parse_weather_info(soup):
|
|
"""
|
|
解析和风天气网页上的天气
|
|
|
|
<p>
|
|
页面的结构一变，取不到元素就返回 None。抛异常的话 Python 的原文
|
|
会被当成回复内容念给用户，听着像系统在报错。
|
|
"""
|
|
if soup is None:
|
|
return None
|
|
|
|
city_element = soup.select_one("h1.c-submenu__location")
|
|
if city_element is None:
|
|
logger.bind(tag=TAG).warning("网页天气解析失败：取不到城市名")
|
|
return None
|
|
city_name = city_element.get_text(strip=True)
|
|
|
|
current_abstract = soup.select_one(".c-city-weather-current .current-abstract")
|
|
current_abstract = (
|
|
current_abstract.get_text(strip=True) if current_abstract else "未知"
|
|
)
|
|
|
|
current_basic = {}
|
|
for item in soup.select(
|
|
".c-city-weather-current .current-basic .current-basic___item"
|
|
):
|
|
parts = item.get_text(strip=True, separator=" ").split(" ")
|
|
if len(parts) == 2:
|
|
key, value = parts[1], parts[0]
|
|
# 风向和风力挤在一项里，拆成两项跟接口那套对上
|
|
if key.endswith("风") and value.endswith("级"):
|
|
current_basic["风向"] = key
|
|
current_basic["风力"] = value
|
|
continue
|
|
key = ITEM_NAMES.get(key, key)
|
|
value = normalize_unit(value)
|
|
if not value or value == "0":
|
|
continue
|
|
current_basic[key] = value
|
|
|
|
# 网页上标的当前时间，取不到就当作没有。页面结构变了不该抛异常
|
|
observed_at = None
|
|
time_element = soup.select_one(".current-weather__bg .current-time")
|
|
if time_element is not None:
|
|
observed_at = time_element.get_text(strip=True) or None
|
|
|
|
temps_list = []
|
|
for row in soup.select(".city-forecast-tabs__row")[:7]: # 取前7天的数据
|
|
date_element = row.select_one(".date-bg .date")
|
|
icon_element = row.select_one(".date-bg .icon")
|
|
if date_element is None or icon_element is None:
|
|
logger.bind(tag=TAG).warning("网页天气解析失败：预报行缺日期或图标")
|
|
return None
|
|
icon_src = icon_element.get("src")
|
|
if not icon_src:
|
|
logger.bind(tag=TAG).warning("网页天气解析失败：预报行图标没有地址")
|
|
return None
|
|
weather_code = icon_src.split("/")[-1].split(".")[0]
|
|
weather = WEATHER_CODE_MAP.get(weather_code, "未知")
|
|
temps = [span.get_text(strip=True) for span in row.select(".tmp-cont .temp")]
|
|
high_temp, low_temp = (temps[0], temps[-1]) if len(temps) >= 2 else (None, None)
|
|
temps_list.append(
|
|
(
|
|
format_forecast_date(date_element.get_text(strip=True)),
|
|
weather,
|
|
normalize_temperature(high_temp) or "未知",
|
|
normalize_temperature(low_temp) or "未知",
|
|
)
|
|
)
|
|
|
|
return city_name, current_abstract, current_basic, temps_list, observed_at
|
|
|
|
|
|
@register_function("get_weather", GET_WEATHER_FUNCTION_DESC, ToolType.SYSTEM_CTL)
|
|
async def get_weather(conn: "ConnectionHandler", location: str = None, lang: str = "zh_CN"):
|
|
from core.utils.cache.manager import cache_manager, CacheType
|
|
|
|
weather_config = conn.config.get("plugins", {}).get("get_weather", {})
|
|
api_key = str(weather_config.get("api_key") or "").strip()
|
|
api_host = str(weather_config.get("api_host") or "").strip()
|
|
default_location = weather_config.get("default_location", "广州")
|
|
client_ip = conn.client_ip
|
|
|
|
# 凭据没配就没法取数据，直接告诉用户去申请，别拿空 key 去撞接口。
|
|
# 这句话要让开发者原样听到：走 LLM 转述的话，文档路径会被转述掉
|
|
if not api_key:
|
|
logger.bind(tag=TAG).warning(
|
|
"未配置和风天气 api_key，天气插件没法取数据。"
|
|
"申请和填写步骤见 docs/weather-integration.md 的「API Key 申请指南」"
|
|
)
|
|
return ActionResponse(Action.RESPONSE, None, WEATHER_CREDENTIAL_HINT)
|
|
if not api_host:
|
|
logger.bind(tag=TAG).warning(
|
|
"未配置和风天气 api_host，天气插件没法取数据。"
|
|
"申请和填写步骤见 docs/weather-integration.md 的「API Key 申请指南」"
|
|
)
|
|
return ActionResponse(Action.RESPONSE, None, WEATHER_HOST_HINT)
|
|
|
|
# 优先使用用户提供的location参数
|
|
if not location:
|
|
# 通过客户端IP解析城市
|
|
if client_ip:
|
|
# 先从缓存获取IP对应的城市信息
|
|
cached_ip_info = cache_manager.get(CacheType.IP_INFO, client_ip)
|
|
if cached_ip_info:
|
|
location = cached_ip_info.get("city")
|
|
else:
|
|
# 缓存未命中，调用API获取
|
|
ip_info = get_ip_info(client_ip, logger)
|
|
if ip_info:
|
|
cache_manager.set(CacheType.IP_INFO, client_ip, ip_info)
|
|
location = ip_info.get("city")
|
|
|
|
if not location:
|
|
location = default_location
|
|
else:
|
|
# 若无IP，使用默认位置
|
|
location = default_location
|
|
# 尝试从缓存获取完整天气报告
|
|
weather_cache_key = f"full_weather_{location}_{lang}"
|
|
failure_cache_key = f"failed_weather_{location}_{lang}"
|
|
cached_weather_report = cache_manager.get(CacheType.WEATHER, weather_cache_key)
|
|
if cached_weather_report:
|
|
return ActionResponse(Action.REQLLM, cached_weather_report, None)
|
|
# 刚失败过就不要再走一遍链路，用户等不起
|
|
cached_failure = cache_manager.get(CacheType.WEATHER, failure_cache_key)
|
|
if cached_failure:
|
|
return ActionResponse(Action.REQLLM, cached_failure, None)
|
|
|
|
# 缓存未命中，获取实时天气数据
|
|
city_info = await fetch_city_info(location, api_key, api_host)
|
|
if not city_info:
|
|
message = f"暂时查不到 {location} 的天气（城市未找到或天气服务不可用），请稍后再试"
|
|
cache_manager.set(
|
|
CacheType.WEATHER, failure_cache_key, message, ttl=WEATHER_FAILURE_TTL
|
|
)
|
|
return ActionResponse(Action.REQLLM, message)
|
|
weather_data = await fetch_weather_data(city_info, api_key, api_host)
|
|
if weather_data:
|
|
weather_report = build_weather_report(city_info, *weather_data)
|
|
else:
|
|
fx_link = city_info.get("fxLink")
|
|
soup = await fetch_weather_page(fx_link) if fx_link else None
|
|
parsed = parse_weather_info(soup)
|
|
if not parsed:
|
|
message = "暂时查不到天气数据，请稍后再试"
|
|
cache_manager.set(
|
|
CacheType.WEATHER, failure_cache_key, message, ttl=WEATHER_FAILURE_TTL
|
|
)
|
|
return ActionResponse(Action.REQLLM, message)
|
|
city_name, current_abstract, current_basic, temps_list, observed_at = parsed
|
|
weather_report = format_weather_report(
|
|
city_name, current_abstract, current_basic, temps_list, observed_at
|
|
)
|
|
|
|
# 缓存完整的天气报告
|
|
cache_manager.set(CacheType.WEATHER, weather_cache_key, weather_report)
|
|
|
|
return ActionResponse(Action.REQLLM, weather_report, None)
|