| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176 |
- from datetime import datetime, timedelta
- import requests
- from requests.exceptions import Timeout
- import retrying
- import execjs
- from lxml import etree
- import json
- from loguru import logger
- import threading
- from queue import Queue
- #
- # 禁用SSL相关警告 (推荐)
- from requests.packages.urllib3.exceptions import InsecureRequestWarning
- import warnings
- requests.packages.urllib3.disable_warnings(InsecureRequestWarning)
- warnings.filterwarnings("ignore", category=DeprecationWarning) # 可选:过滤其他警告
- class GK:
- def __init__(self):
- self.search_flights_api = "https://booking.jetstar.com/hk/zh/booking/search-flights"
- self.headers = {
- 'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36',
- 'Accept-Encoding': 'gzip, deflate, br',
- 'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.7',
- 'Connection': 'keep-alive',
- 'Content-Type': 'application/x-www-form-urlencoded',
- 'accept-language': 'zh-CN,zh;q=0.9',
- 'cache-control': 'no-cache',
- 'pragma': 'no-cache',
- 'priority': 'u=0, i',
- 'referer': 'https://booking.jetstar.com/',
- 'sec-ch-ua': '"Google Chrome";v="135", "Not-A.Brand";v="8", "Chromium";v="135"',
- 'sec-ch-ua-mobile': '?0',
- 'sec-ch-ua-platform': '"Windows"',
- 'sec-fetch-dest': 'document',
- 'sec-fetch-mode': 'navigate',
- 'sec-fetch-site': 'same-origin',
- 'sec-fetch-user': '?1',
- 'upgrade-insecure-requests': '1'
- }
- with open('./akm逆向/逆向.js', encoding='utf-8') as f:
- js = f.read()
- self.ctx = execjs.compile(js)
- self.session = requests.Session()
- # self.ip = 100000000000
- # self.proxies = {
- # 'http': f'http://B_3351_HK___5_ss-{self.ip}:ev2pjj@proxy.renlaer.com:7778',
- # 'https': f'http://B_3351_HK___5_ss-{self.ip}:ev2pjj@proxy.renlaer.com:7778'
- # }
- # self.proxies = {
- # 'http': f'127.0.0.1:8888',
- # 'https': f'127.0.0.1:8888'
- # }
- self.cookies_queue = Queue()
- def get_cookie(self):
- logger.debug('正在获取 cookie bm-sz...')
- # akm js file url
- akm_url = "https://booking.jetstar.com/MkuYlo/pcp/LD0/PPluEQ/1ik7QcJffXbmL53i/QTcvXmg7/KS5kC3N/VRQcB"
- data = {
- 'sensor_data': self.ctx.call('encrypt1')
- }
- response = self.session.post(akm_url, headers=self.headers, verify=False, data=data,
- # proxies=self.proxies
- )
- # print(response.text)
- bmsz = response.cookies.get_dict()['bm_sz']
- data2 = {
- "sensor_data": self.ctx.call('encrypt2', bmsz)
- }
- data2 = json.dumps(data2)
- response2 = self.session.post(akm_url, headers=self.headers, data=data2, verify=False,
- # proxies=self.proxies
- )
- # print(response2.text)
- # print(response2.status_code)
- logger.debug(f'成功获取到 bm-sz :{bmsz}')
- @retrying.retry(stop_max_attempt_number=2)
- def send_get(self, url, params):
- response = self.session.get(
- url,
- headers=self.headers, params=params,
- timeout=15,
- verify=False,
- )
- response.raise_for_status()
- return response
- @retrying.retry(stop_max_attempt_number=3)
- def get_data(self, datetime_str):
- params = {
- "s": "true",
- "adults": "1", # 成年人
- "children": "0", # 儿童
- "infants": "0", # 婴儿
- "selectedclass1": "economy", # 选择类型:经济舱
- "currency": "CNY", # 货币
- "mon": "true",
- "channel": "DESKTOP",
- "origin1": "PVG", # 出发地
- "destination1": "NRT", # 目的地
- "departuredate1": datetime_str # 出发时间
- }
- logger.info(f'正在采集 {datetime_str} 航班数据...')
- try:
- response = self.send_get(self.search_flights_api, params)
- if not response:
- return
- return datetime_str, response
- # print(response.text)
- except Exception as e:
- logger.error(e)
- self.session = requests.Session()
- # self.ip += 1
- self.get_cookie()
- raise
- # return datetime_str, None
- def parse_data(self, datetime_str, response):
- if not response:
- return
- html = etree.HTML(response.text)
- data = html.xpath("//script[@id='bundle-data-v2']/text()")
- if data:
- json_data = json.loads(data[0])
- print(datetime_str, ' => ', json_data)
- else:
- logger.warning(f'{datetime_str} 当天暂无数据 / 触发验证码')
- print(response.text)
- @staticmethod
- def gen_datetime(start_date, end_date):
- current_date = datetime.strptime(start_date, '%Y-%m-%d')
- end_date = datetime.strptime(end_date, '%Y-%m-%d')
- date_list = []
- while current_date <= end_date:
- date_list.append(current_date.strftime('%Y-%m-%d')) # 转换为字符串格式存储
- current_date += timedelta(days=1)
- return date_list
- def run(self, start_date, end_date):
- self.get_cookie()
- # 获取采集时间
- for num, datetime_str in enumerate(self.gen_datetime(start_date, end_date)):
- # if num % 4 == 0:
- # self.session = requests.Session()
- # self.get_cookie()
- datetime_str, response = self.get_data(datetime_str)
- self.parse_data(datetime_str, response)
- # time.sleep(2)
- if __name__ == '__main__':
- gk = GK()
- gk.run(start_date='2025-05-15', end_date='2025-05-27')
|