黑光人才落户成都(黑光图库爬取)

黑光人才落户成都

黑光图库摄影

import requestsimport timeimport refrom urllib.parse import urlencodeimport asyncioasync def heiguang(): s = 1 shuru = input('请输入关键字:') for v in range(1, 1000): data = { "wd": shuru, "page": v, } headers = { "User-Agent": "Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/72.0.3610.2 Safari/537.36", "Referer": "https://tu.heiguang.com/search/all?wd=%E9%A3%8E%E6%99%AF", "origin": "https://tu.heiguang.com", "x-csrf-token": "83soU6DMxqVp3xBaHItySbHgE-G60v6OLxplSD1YtlyKCnwYja2S1SG3RRR0zD4A8KRfhs-ntLx7UgMDBDf8MA==", "x-requested-with": "XMLHttpRequest", "x-tingyun-id": "genFp9yIkGM;r=902391693", "Cookie": "__cfduid=dca794229b8f10a0e24721fbe1f5f5a721548685304; iswap=nwap; _csrf=74d5d67a8575d02f4673d0cea998798154517f427e0d37f8544855a7ddde9abba%3A2%3A%7Bi%3A0%3Bs%3A5%3A%22_csrf%22%3Bi%3A1%3Bs%3A32%3A%22yqTK-aTpHhUNhGLIADLguuJ2THfK9oJl%22%3B%7D; TY_SESSION_ID=88153885-f535-494e-b8b0-6a34d5ed2b10; yjs_id=64b0ebc3c8cc8de2a2105a6131cfc912; Hm_lvt_a0e2f5480dc36f754aa28437a28fc063=1548685251; Hm_lvt_b402e0a7169c75a55900fb37559d05b0=1548685252; UM_distinctid=16894d59955540-0f9608e9c0eac7-19187552-100200-16894d599563d5; Hm_lvt_3f8b19dbb40270224bdaa4f29c84357f=1548685253; Hm_lpvt_3f8b19dbb40270224bdaa4f29c84357f=1548685253; PHPSESSID=aj0liih3p3bq5umga19ccnupb2; ctrl_time=1; Hm_lpvt_a0e2f5480dc36f754aa28437a28fc063=1548858744; Hm_lpvt_b402e0a7169c75a55900fb37559d05b0=1548858744" } url = 'https://tu.heiguang.com/search/works?'+urlencode(data) html = requests.get(url, headers=headers, timeout=30) html_text = (html.text) html_list = re.findall('"(http.*?)"', html_text) html_list1 = re.findall('"(https.*?)"', html_text) # print(html_text) for html_list,html_list1 in zip(html_list, html_list1): a = (html_list)[:-1] aa = (html_list1)[:-1] b = requests.get(a) bb = requests.get(aa) c = b.content cc = bb.content op = open(r'e:/python/爬完数据存放地址/壁纸/成都%d.png'%s, 'wb') print('正在下载', a) print('正在下载', aa) op.write(c) op.write(cc) s+=1if __name__ == '__main__': start = time.time() loop = asyncio.get_event_loop() loop.run_until_complete(heiguang()) loop.close() print('[info]耗时:%s' % (time.time() - start))

黑光图库爬取,黑光图库摄影