JobData/spiderJobs/platforms/job51/client.py

"""
前程无忧 (51Job) HTTP 客户端
在通用 HTTPClient 上叠加 51job 特有的 sign 签名和默认 headers

与 Boss/智联不同，51job 的 sign 依赖完整的 URL path + body，
因此需要在 post/get 方法中先构造签名再拼接最终 URL。
"""

from __future__ import annotations

import json
from typing import Any, Optional
from urllib.parse import quote

from crawler_core.http_client import HTTPClient
from crawler_core.qcwy.sign import Job51Sign

BASE_URL = "https://cupid.51job.com"

# 51job 小程序特有的默认请求头
JOB51_HEADERS = {
    "user-agent": (
        "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 "
        "(KHTML, like Gecko) Chrome/132.0.0.0 Safari/537.36 "
        "MicroMessenger/7.0.20.1781(0x6700143B) NetType/WIFI "
        "MiniProgramEnv/Mac MacWechat/WMPF MacWechat/3.8.7(0x13080712) "
        "UnifiedPCMacWechat(0xf2641702) XWEB/18788"
    ),
    "xweb_xhr": "1",
    "from-domain": "51job_weixin_wxapp",
    "sec-fetch-site": "cross-site",
    "sec-fetch-mode": "cors",
    "sec-fetch-dest": "empty",
    "referer": "https://servicewechat.com/wx1131e5c71e668b5d/426/page-frame.html",
    "accept-language": "zh-CN,zh;q=0.9",
    "priority": "u=1, i",
}


class Job51Client(HTTPClient):
    """
    前程无忧 HTTP 客户端

    继承通用 HTTPClient，每次请求自动计算 HMAC-SHA256 签名

    Args:
        signer:       Job51Sign 实例（可选）
        tunnel_proxy: 隧道代理地址（每次请求自动换 IP）
        proxy:        固定代理地址
        proxy_pool:   代理池列表
        timeout:      请求超时秒数
    """

    def __init__(
        self,
        signer: Optional[Job51Sign] = None,
        tunnel_proxy: Optional[str] = None,
        proxy: Optional[str] = None,
        proxy_pool: Optional[list[str]] = None,
        timeout: int = 10,
    ):
        super().__init__(
            base_url=BASE_URL,
            default_headers=JOB51_HEADERS,
            tunnel_proxy=tunnel_proxy,
            proxy=proxy,
            proxy_pool=proxy_pool,
            timeout=timeout,
        )
        self.signer = signer or Job51Sign()
        self._uuid = Job51Sign.generate_uuid()

    def _job51_headers(self, sign: str) -> dict:
        """构造每次请求的 51job 特有 headers"""
        property_obj = {
            "frompageUrl": "",
            "pageUrl": "pages/index/index",
            "isLogin": "否",
            "accountid": "",
            "resumeId": "",
            "firstFrompageUrl": "",
            "distinct_id": self._uuid,
        }
        return {
            "sign": sign,
            "partner": "",
            "property": quote(json.dumps(property_obj, ensure_ascii=False, separators=(",", ":")), safe=""),
            "uuid": self._uuid,
            "user-token": "",
            "account-id": "",
        }

    def post(self, path: str, body: dict, headers: Optional[dict] = None) -> tuple[int, Any]:
        """
        POST 请求，自动计算签名

        注意: path 参数为 endpoint（如 open/noauth/recommend/job-tab-dynamic-wx-mini）
        签名后会拼为 /endpoint?api_key=51job&timestamp=xxx

        关键: body 必须以 compact JSON 发送（无空格），与签名字符串完全一致
        不能使用 requests 的 json= 参数（会用默认带空格的序列化）
        """
        url_path, sign = self.signer.build_sign_path(path, "POST", body=body)

        job51_h = self._job51_headers(sign)
        job51_h["Content-Type"] = "application/json"
        if headers:
            job51_h.update(headers)

        # 必须用 compact JSON（与签名一致），通过 _post_raw 发送预序列化 body
        raw_body = json.dumps(body, ensure_ascii=False, separators=(",", ":"))
        return self._post_raw(url_path, raw_body, job51_h)

    def _post_raw(self, path: str, raw_body: str, headers: dict) -> tuple[int, Any]:
        """发送预序列化的 POST 请求（data= 而非 json=）"""
        merged_headers = self._merge_headers(headers)
        url = f"{self.base_url}{path}"

        if self._tunnel_proxy:
            import requests_go as requests
            s = self._new_session()
            try:
                resp = s.post(
                    url,
                    data=raw_body.encode("utf-8"),
                    headers=merged_headers,
                    proxies={"http": self._tunnel_proxy, "https": self._tunnel_proxy},
                    timeout=self.timeout,
                )
                return resp.status_code, resp.json()
            finally:
                s.close()

        proxies = self._get_proxies()
        kwargs: dict[str, Any] = {
            "data": raw_body.encode("utf-8"),
            "headers": merged_headers,
            "timeout": self.timeout,
        }
        if proxies:
            kwargs["proxies"] = proxies
        resp = self._session.post(url, **kwargs)
        return resp.status_code, resp.json()

    def get(self, path: str, params: Optional[dict] = None, headers: Optional[dict] = None) -> tuple[int, Any]:
        """
        GET 请求，自动计算签名

        注意: params 会被编入签名路径的 query string 中
        """
        url_path, sign = self.signer.build_sign_path(path, "GET", params=params)

        job51_h = self._job51_headers(sign)
        job51_h["content-type"] = "application/x-www-form-urlencoded"
        if headers:
            job51_h.update(headers)

        # GET 参数已经编入 url_path，不再传 params
        return super().get(url_path, params=None, headers=job51_h)


def create_client(
    signer: Optional[Job51Sign] = None,
    tunnel_proxy: Optional[str] = None,
    proxy: Optional[str] = None,
    proxy_pool: Optional[list[str]] = None,
) -> Job51Client:
    """创建 51job 客户端"""
    return Job51Client(signer=signer, tunnel_proxy=tunnel_proxy, proxy=proxy, proxy_pool=proxy_pool)