hgg070/hgg070_spider/spiders/zuqiu.py

# -*- coding: utf-8 -*-
import scrapy


class ZuqiuSpider(scrapy.Spider):
    name = 'zuqiu'
    allowed_domains = ['m.hgg070.com']

    # 读取今日足球
    def start_requests(self):
        headers = {
            'Accept': '*/*',
            'Accept-Encoding': 'gzip, deflate',
            'Accept-Language': 'zh-CN,zh;q=0.9',
            'Content-Length': '130',
            'Content-type': 'application/x-www-form-urlencoded',
            'Cookie': '_ga=GA1.2.1009358217.1572056223; _gid=GA1.2.97506800.1572056223; _gat=1',
            'Host': 'm.hgg070.com',
            'Origin': 'http://m.hgg070.com',
            'Proxy-Connection': 'keep-alive',
            'Referer': 'http://m.hgg070.com/',
            'User-Agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 11_0 like Mac OS X) AppleWebKit/604.1.38 (KHTML, like Gecko) Version/11.0 Mobile/15A372 Safari/604.1'
        }
        url = "http://m.hgg070.com/app/member/get_league_list.php"
        h_types = [('FT', '', '130'), ('FU', 'P', '131'), ('FU', "", '130'), ('FU', '', '131')]
        for h_type in h_types:
            show_type, isp, length = h_type
            headers['Content-Length'] = length
            from_data = {
                'uid': '4f6573b99cc31f7be1579b063888a8a2bcda122ce94228510cda9e1bb32a477f',
                'langx': 'zh-cn',
                'ltype': '3',
                'gtype': 'FT',
                'showtype': show_type,
                'sorttype': '',
                'date': '',
                'isP': isp
            }
            yield scrapy.FormRequest(url=url, formdata=from_data, callback=self.parse, headers=headers,
                                     meta={'showtype': show_type, 'isp': isp}, dont_filter=True)

    def parse(self, response):
        leagues = response.xpath('//serverresponse/game/league')
        url = 'http://m.hgg070.com/app/member/get_game_list.php'
        headers = {
            'Accept': '*/*',
            'Accept-Encoding': 'gzip, deflate',
            'Accept-Language': 'zh-CN,zh;q=0.9',
            'Content-Length': '147',
            'Content-type': 'application/x-www-form-urlencoded',
            'Cookie': '_ga=GA1.2.1009358217.1572056223; _gid=GA1.2.97506800.1572056223; _gat=1',
            'Host': 'm.hgg070.com',
            'Origin': 'http://m.hgg070.com',
            'Proxy-Connection': 'keep-alive',
            'Referer': 'http://m.hgg070.com/',
            'User-Agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 11_0 like Mac OS X) AppleWebKit/604.1.38 (KHTML, like Gecko) Version/11.0 Mobile/15A372 Safari/604.1'
        }
        if leagues:
            showtype = response.meta['showtype']
            isp = response.meta['isp']
            if showtype == 'FT' and isp == '':
                date = ''
                headers['Content-Length'] = '147'
            elif showtype == 'FU' and isp == 'P':
                date = 'all'
                headers['Content-Length'] = '151'
            elif showtype == 'FU' and isp == '':
                date = 'all'
                headers['Content-Length'] = '150'
            else:
                date = 'all'
                headers['Content-Length'] = '151'
            for league in leagues:
                lid = league.xpath('.//league_id/text()').extract_first()
                from_data = {
                    'uid': '4f6573b99cc31f7be1579b063888a8a2bcda122ce94228510cda9e1bb32a477f',
                    'langx': 'zh-cn',
                    'ltype': '3',
                    'gtype': 'FT',
                    'showtype': showtype,
                    # 'showtype': "FT",
                    # 'lid': '103391',
                    'lid': lid,
                    'sorttype': 'league',
                    'date': date,
                    'isP': isp
                    # 'date': "",
                    # 'isP': ""
                }
                yield scrapy.FormRequest(url=url, formdata=from_data, callback=self.parse_match, headers=headers,
                                         meta={'showtype': showtype, 'isp': isp}, dont_filter=True)

    def parse_match(self, response):
        print(response.text)
        pass