Python爬蟲 - 抓取Tiktok小視頻

前言

好久沒更新了,更新一下
不廢話了

正文

確保電腦安裝了 charles / fiddler / wireshark 或其他能抓包的工具,然后在手機或者模擬器上配置好代理,記得安裝證書,接著開始抓包。

:最好在 Android7.0 以下配置,7.0及以上版本不能抓https了,有點麻煩,推薦使用模擬器

首先安裝好 tiktok ,打開后翻幾個視頻,會發(fā)現(xiàn)抓包工具里出現(xiàn)一堆請求
charles
我們需要找到它小視頻請求的鏈接,經(jīng)過我個把小時的分析,找到了這個
charles
接下來需要模擬請求,看看他的請求頭

accept-encoding = gzip
user-agent = okhttp/3.10.0.1
method = GET

接著我發(fā)現(xiàn)雖然是GET請求,但是url鏈接后面有一串賊雞兒長的參數(shù),wtm。。。
好吧,分析分析請求參數(shù)把,看著太費勁所以我寫了一個解析方法,解析成json格式,如下
 "type": "0",
 "max_cursor": "0",
 "min_cursor": "-1",
 "count": COUNT, # 加載個數(shù)
 "volume": "0.2",
 "pull_type": "2",
 "req_from": "",
 "gaid": "",
 "ad_user_agent": "Dalvik%2F2.1.0+%28Linux%3B+U%3B+Android+6.0.1%3B+MuMu+Build%2FV417IR%29", # 廣告ua
 "filter_warn": "0",
 "ts": TIMESTAMP[:-3], # 時間戳
 "js_sdk_version": "",
 "app_type": "normal",
 "os_api": "28",
 "device_type": "Xiaomi",
 "ssmix": "a",
 "manifest_version_code": "9999",
 "dpi": "240",
 "carrier_region": "JP", # 獲取的小視頻來源
 "region": "JP",
 "carrier_region_v2": "",
 "app_name": "trill",
 "version_name": "4.2.4",
 "ab_version": "4.2.4",
 "timezone_offset": "28800",
 "pass-route": "1",
 "pass-region": "1",
 "is_my_cn": "0",
 "fp":"",
 #"fp": "R2TrcrUtFMsrFYQZcrU1LSFeFWm1",
 "ac": "wifi",
 "update_version_code": "4240",
 "channel": "googleplay",
 "_rticket": TIMESTAMP, # 時間戳(保留毫秒)
 "device_platform": "android",
 "iid":"6678981313393919746", # 暫不知來源,必須有,不能更改
#     "iid": "6678981313393919746",
 "build_number": "9.5",
 "version_code": "424",
 "timezone_name": "Asia%2FShanghai",
 "openudid": "c55cd951e5464a56",
 "device_id": "6678979145927919106",
 "sys_region": "JP",
 "app_language": "zh-Hans",
 "resolution": RESOLUTION, # 畫質(zhì)
#     "resolution": "1280*720",
 "os_version": "9",
 "device_brand": "Android",
 "language": "jp",
 "aid": "1180", # 暫不知來源,必須有,不能更改
 "mcc_mnc": "",
 "as": "",
 "cp":"",
 "mas":""
 #"as": "a10548abb2087cfbe04299",
 #"cp": "8c8fc35c2606b9bae1Wc_g",
 #"mas": "019e07c3ae4fc3d6e172de9de34998f0ae9c9c4c2c0ca64666c6ec"
}
這里面有些能從字面意思看懂的就不多說了,主要看那后面幾個,fp、iid、aid、as、cp、mas 這里除了iid和aid這兩個參數(shù)不能更改以外,以他的參數(shù)隨便改,暫時也不知道 ? 和 ? 這兩個參數(shù)的來歷,保持不變即可
ok,模擬請求,代碼如下
import requests
?
headers = {
 "User-Agent": "okhttp/3.10.0.1",
 "Accpet-Encoding": "gzip"
}
?
url = "https://api-h2.tiktokv.com/aweme/v1/feed/"
?
def parse_url_params(param):
 """
 解析url參數(shù)為json
 """
 params = {}
 if '?' in param:
 p = param.split('?')[-1]
 ps = p.split('&')
 for i in ps:
 a = i.split('=')
 params.update({a[0]:'' if len(a) == 1 else a[1]})
 else:
 ps = param.split('&')
 for i in ps:
 a = i.split('=')
 params.update({a[0]:'' if len(a) == 1 else a[1]})
 return params
?
params = parse_url_params("type=0&max_cursor=0&min_cursor=-1&count=6&volume=0.2&pull_type=2&req_from&gaid&ad_user_agent=Dalvik%2F2.1.0+%28Linux%3B+U%3B+Android+6.0.1%3B+MuMu+Build%2FV417IR%29&filter_warn=0&ts=1555073922&js_sdk_version=&app_type=normal&os_api=23&device_type=MuMu&ssmix=a&manifest_version_code=9999&dpi=240&carrier_region=JP&region=CN&carrier_region_v2=&app_name=trill&version_name=4.2.4&ab_version=4.2.4&timezone_offset=28800&pass-route=1&pass-region=1&is_my_cn=0&fp=R2TrcrUtFMsrFYQZcrU1LSFeFWm1&ac=wifi&update_version_code=4240&channel=googleplay&_rticket=1555073923091&device_platform=android&iid=6678981313393919746&build_number=9.5&version_code=424&timezone_name=Asia%2FShanghai&openudid=c55cd951e5464a56&device_id=6678979145927919106&sys_region=CN&app_language=zh-Hans&resolution=1280*720&os_version=6.0.1&device_brand=Android&language=zh&aid=1180&mcc_mnc=&as=a10548abb2087cfbe04299&cp=8c8fc35c2606b9bae1Wc_g&mas=019e07c3ae4fc3d6e172de9de34998f0ae9c9c4c2c0ca64666c6ec")
?
resp = requests.get(url, headers=headers, params=params)
print(resp.json())
輸出一堆json內(nèi)容,由于內(nèi)容實在是太多,所以只看重點部分,下面是幾個
{'aweme_id': '6673431479185624322', 'desc': '', 'create_time': 1553847851,..., 'video': {'play_addr': {'uri': 'v07025860000bietc60src5j25b1epvg', 'url_list': [...], ... 'play_addr': {'uri': 'v07025860000bietc60src5j25b1epvg', 'url_list': [...], ... , 'url_list': [...], ... , 'url_list': [...] ... 'url_list': [...], 'width': 720, 'height': 720, 'url_key': 'v07025860000bietc60src5j25b1epvg_h264_360p_290447'}, 'is_h265': 0}], ...}
這是一個視頻的部分json內(nèi)容,我們需要的有 create_time、video - play_addr - uri,emmm好像就這些,還有作者的信息我沒有貼上來自行分析吧,接著上全部代碼
# Spider.py
import sys
import requests
?
from config import * # 這里我把請求參數(shù)寫在config里了,代碼在下面
?
?
class TiktokSpider:
?
 def __init__(self):
 self.url = 'https://api-h2.tiktokv.com/aweme/v1/feed'
 self.headers = {
 "User-Agent": "okhttp/3.10.0.1",
 "Accpet-Encoding": "gzip"
 }
?
 def get_aweme_list(self):
 """
 獲取視頻列表
 :return:
 """
 resp = requests.get(self.url, headers=self.headers, params=PARAMS)
 if resp.status_code != requests.codes.OK:
 print("獲取視頻列表失敗.")
 return None
 try:
 return resp.json()['aweme_list']
 except KeyError:
 print(resp.text)
?
 def parse_video(self):
 """
 解析json格式小視頻列表數(shù)據(jù)
 :return:
 """
 aweme_list = self.get_aweme_list()
 if aweme_list is None:
 return aweme_list
 video_result = []
 for aweme in aweme_list:
 play_addr = aweme['video']['play_addr']
 author = aweme['author']
 result = {
 'video_infos': {
 'url': play_addr['url_list'],
 'width': play_addr['width'],
 'height': play_addr['height'],
 'description': aweme['desc'],
 'create_time': str(time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(aweme['create_time']))),
 },
 'aweme_id': aweme['aweme_id'],
 'avatar_infos': {
 'url': author['avatar_thumb']['url_list'],
 'height': author['avatar_larger']['height'],
 'width': author['avatar_larger']['width']
 },
 'author_infos': {
 'nickname': author['nickname'],
 'uid': author['uid'],
 'short_id': author['short_id'],
 'signature': author['signature'],
 }
?
 }
 video_result.append(result)
 return video_result
?```
```python
# config.py
import time
import math
?
ORIGION='JP'
COUNT="10"
TIMESTAMP=str(math.floor(time.time()*1000))
RESOLUTION="1920*1080"
?
?
PARAMS = {
 "type": "0",
 "max_cursor": "0",
 "min_cursor": "-1",
 "count": COUNT, # 加載個數(shù)
 "volume": "0.2",
 "pull_type": "2",
 "req_from": "",
 "gaid": "",
 "ad_user_agent": "Dalvik%2F2.1.0+%28Linux%3B+U%3B+Android+6.0.1%3B+MuMu+Build%2FV417IR%29", # 廣告ua
 "filter_warn": "0",
 "ts": TIMESTAMP[:-3], # 時間戳
 "js_sdk_version": "",
 "app_type": "normal",
 "os_api": "28",
 "device_type": "Xiaomi",
 "ssmix": "a",
 "manifest_version_code": "9999",
 "dpi": "240",
 "carrier_region": "JP", # 獲取的小視頻來源
 "region": "JP",
 "carrier_region_v2": "",
 "app_name": "trill",
 "version_name": "4.2.4",
 "ab_version": "4.2.4",
 "timezone_offset": "28800",
 "pass-route": "1",
 "pass-region": "1",
 "is_my_cn": "0",
 "fp":"",
 #"fp": "R2TrcrUtFMsrFYQZcrU1LSFeFWm1",
 "ac": "wifi",
 "update_version_code": "4240",
 "channel": "googleplay",
 "_rticket": TIMESTAMP, # 時間戳(保留毫秒)
 "device_platform": "android",
 "iid":"6678981313393919746", # 暫不知來源,必須有,不能更改
#     "iid": "6678981313393919746",
 "build_number": "9.5",
 "version_code": "424",
 "timezone_name": "Asia%2FShanghai", # 時區(qū)
 "openudid": "c55cd951e5464a56",
 "device_id": "6678979145927919106",
 "sys_region": "JP", 
 "app_language": "zh-Hans",
 "resolution": RESOLUTION, # 畫質(zhì)
#     "resolution": "1280*720",
 "os_version": "9",
 "device_brand": "Android",
 "language": "jp",
 "aid": "1180", # 暫不知來源,必須有,不能更改
 "mcc_mnc": "",
 "as": "",
 "cp":"",
 "mas":""
 #"as": "a10548abb2087cfbe04299",
 #"cp": "8c8fc35c2606b9bae1Wc_g",
 #"mas": "019e07c3ae4fc3d6e172de9de34998f0ae9c9c4c2c0ca64666c6ec"
}

最后

emmm,其實我把抓到的數(shù)據(jù)用flask寫成了一個小網(wǎng)站,由于我css賊垃圾,所以做的不是很美觀,項目地址: Flask-Tiktok

原文:Python爬蟲 - 抓取國際版tiktok小視頻

最后編輯于
?著作權歸作者所有,轉載或內(nèi)容合作請聯(lián)系作者
【社區(qū)內(nèi)容提示】社區(qū)部分內(nèi)容疑似由AI輔助生成,瀏覽時請結合常識與多方信息審慎甄別。
平臺聲明:文章內(nèi)容(如有圖片或視頻亦包括在內(nèi))由作者上傳并發(fā)布,文章內(nèi)容僅代表作者本人觀點,簡書系信息發(fā)布平臺,僅提供信息存儲服務。

友情鏈接更多精彩內(nèi)容