123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990 |
- # -*- coding: utf-8 -*-
- """
- Created on 2025-05-28
- ---------
- @summary: 沈阳联合产权交易所
- ---------
- @author: lzz
- """
- import feapder
- from items.spider_item import MgpListItem
- from collections import namedtuple
- class Sylhcqjys(feapder.BiddingListSpider):
- def start_callback(self):
- Menu = namedtuple('Menu', ['channel', 'code','typeone', 'typetwo','typethree', 'crawl_page'])
- self.site = "沈阳联合产权交易所"
- self.menus = [
- Menu('资产转让', 'a_sylhcqjys_zczr','wqzrxmGetList', 'zc', '1', 1),
- ]
- self.headers = {
- "Accept": "application/json, text/javascript, */*; q=0.01",
- "Accept-Language": "zh-CN,zh;q=0.9",
- "Cache-Control": "no-cache",
- "Connection": "keep-alive",
- "Content-Type": "application/x-www-form-urlencoded; charset=UTF-8",
- "Origin": "https://test.sprtc.com",
- "Pragma": "no-cache",
- "User-Agent": "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.127 Safari/537.36",
- "X-Requested-With": "XMLHttpRequest",
- }
- def start_requests(self):
- for menu in self.menus:
- start_url = 'https://test.sprtc.com/idtAppServiceV6/exclude/control/getGrid'
- yield feapder.Request(url=start_url,item=menu._asdict(),page=1, proxies=False)
- def download_midware(self, request):
- page = request.page
- menu = request.item
- data = {
- "script": f"{menu.get('typeone')}",
- "args": "[]",
- "keys": f"[\"{menu.get('typetwo')}\"]",
- "replaceMap": "{\"developer\":\"sjsxm\"}",
- "pageSize": "20",
- "nowPage": f"{page}"
- }
- request.data = data
- request.headers=self.headers
- def parse(self, request, response):
- menu = request.item
- info_list = response.json.get('data').get('DataList')
- for info in info_list:
- href_id = info.get('UUID').strip()
- href = f"https://www.sprtc.com/index/qrtwo.htm?id={href_id}"
- title = info.get('BDMC').strip()
- create_time = info.get('GPJSSJ').strip()
- area = "辽宁" # 省份
- city = "沈阳市" # 城市
- list_item = MgpListItem() # 存储数据的管道
- list_item.href = href # 标书链接
- list_item.channel = menu.get("channel") # 最上方定义的抓取栏目 (编辑器定的)
- list_item.spidercode = menu.get("code") # 最上方定义的爬虫code(编辑器定的)
- list_item.title = title # 标题
- list_item.publishtime = create_time # 标书发布时间
- list_item.site = self.site
- list_item.area = area # 城市默认:全国
- list_item.city = city # 城市 默认为空
- list_item.unique_key = ('href',)
- list_item.parse = "self.detail_get"
- list_item.deal_detail = ['//div[@id="formApp"]/div[@id="formScrollWrap"]']
- list_item.proxies = False
- list_item.parse_url = href
- list_item.render_time = 5
- yield list_item
- if __name__ == "__main__":
- Sylhcqjys(redis_key="lzz:Sylhcqjys").start()
|