|
|
@@ -8,6 +8,8 @@ import sys
|
|
|
import time
|
|
|
import requests
|
|
|
from hashlib import md5
|
|
|
+
|
|
|
+from common.mq import MQ
|
|
|
sys.path.append(os.getcwd())
|
|
|
from common.common import Common
|
|
|
from common.scheduling_db import MysqlHelper
|
|
|
@@ -54,6 +56,7 @@ class DouyinauthorScheduling:
|
|
|
|
|
|
@classmethod
|
|
|
def get_videoList(cls, log_type, crawler, user_dict, rule_dict, env):
|
|
|
+ mq = MQ(topic_name="topic_crawler_etl_" + env)
|
|
|
max_cursor = ""
|
|
|
# while True:
|
|
|
url = "https://www.douyin.com/aweme/v1/web/aweme/post/?device_platform=webapp&aid=6383&channel=channel_pc_web&sec_user_id={sec_user_id}&max_cursor={max_cursor}&show_live_replay_strategy=1&count=10&publish_video_strategy_type=2&pc_client_type=1&version_code=170400&version_name=17.4.0&cookie_enabled=true&screen_width=1440&screen_height=900&browser_language=zh-CN&browser_platform=MacIntel&browser_name=Chrome&browser_version=112.0.0.0&browser_online=true&engine_name=Blink&engine_version=112.0.0.0&os_name=Mac+OS&os_version=10.15.7&cpu_core_num=8&device_memory=8&platform=PC&downlink=10&effective_type=4g&round_trip_time=50".format(
|
|
|
@@ -146,12 +149,23 @@ class DouyinauthorScheduling:
|
|
|
Common.logger(log_type, crawler).info('视频已下载\n')
|
|
|
Common.logging(log_type, crawler, env, '视频已下载\n')
|
|
|
else:
|
|
|
- cls.download_publish(log_type=log_type,
|
|
|
- crawler=crawler,
|
|
|
- user_dict=user_dict,
|
|
|
- video_dict=video_dict,
|
|
|
- rule_dict=rule_dict,
|
|
|
- env=env)
|
|
|
+ # cls.download_publish(log_type=log_type,
|
|
|
+ # crawler=crawler,
|
|
|
+ # user_dict=user_dict,
|
|
|
+ # video_dict=video_dict,
|
|
|
+ # rule_dict=rule_dict,
|
|
|
+ # env=env)
|
|
|
+ video_dict["out_user_id"] = video_dict["user_id"]
|
|
|
+ video_dict["platform"] = crawler
|
|
|
+ video_dict["strategy"] = log_type
|
|
|
+ video_dict["out_video_id"] = video_dict["video_id"]
|
|
|
+ video_dict["width"] = video_dict["video_width"]
|
|
|
+ video_dict["height"] = video_dict["video_height"]
|
|
|
+ video_dict["crawler_rule"] = json.dumps(rule_dict)
|
|
|
+ video_dict["user_id"] = user_dict["uid"]
|
|
|
+ video_dict["publish_time"] = video_dict["publish_time_str"]
|
|
|
+
|
|
|
+ mq.send_msg(video_dict)
|
|
|
except Exception as e:
|
|
|
Common.logger(log_type, crawler).error(f"抓取单条视频异常:{e}\n")
|
|
|
Common.logging(log_type, crawler, env, f"抓取单条视频异常:{e}\n")
|
ehlxr