wangshaoqing/douyin-crawler-poc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: wangshaoqing:f60cf9c243fa0082cabcce88f3f2008e86242e93
Branches Tags
wangshaoqing:main
..
compare: wangshaoqing:cb7f2c89f78d802dcf4e044a7ab6aafb22f24385
Branches Tags
wangshaoqing:main

No commits in common. "f60cf9c243fa0082cabcce88f3f2008e86242e93" and "cb7f2c89f78d802dcf4e044a7ab6aafb22f24385" have entirely different histories.
f60cf9c243 ... cb7f2c89f7

2 changed files with 3 additions and 3 deletions
Show Stats Expand all files Collapse all files

2
Douyin.py
View File

@ -30,7 +30,7 @@ LISTEN_TARGET = "web/aweme/post/"
RECOMMENDATION_LISTEN_TARGET = "aweme/v2/web/module/feed/" RECOMMENDATION_LISTEN_TARGET = "aweme/v2/web/module/feed/"
SINGLE_VIDEO_LISTEN_TARGET = "web/aweme/detail/" SINGLE_VIDEO_LISTEN_TARGET = "web/aweme/detail/"
INVALID_FILENAME_CHARS = re.compile(r'[\\/:*?"<>|\r\n\t]') INVALID_FILENAME_CHARS = re.compile(r'[\\/:*?"<>|\r\n\t]')
RECOMMENDATION_URL_PATTERN = re.compile(r"^https?://www\.douyin\.com/?(?:jingxuan)?(?:\?.*)?$") RECOMMENDATION_URL_PATTERN = re.compile(r"^https?://www\.douyin\.com/?(?:\?.*)?$")
CREATOR_URL_PATTERN = re.compile(r"^https?://www\.douyin\.com/user/[^/?#]+(?:\?.*)?$") CREATOR_URL_PATTERN = re.compile(r"^https?://www\.douyin\.com/user/[^/?#]+(?:\?.*)?$")
VIDEO_URL_PATTERN = re.compile(r"^https?://www\.douyin\.com/video/(?P<aweme_id>\d+)(?:[/?#].*)?$") VIDEO_URL_PATTERN = re.compile(r"^https?://www\.douyin\.com/video/(?P<aweme_id>\d+)(?:[/?#].*)?$")
AWEME_ID_PATTERN = re.compile(r"^\d{5,}$") AWEME_ID_PATTERN = re.compile(r"^\d{5,}$")

4
login_douyin.py
View File

@ -7,7 +7,7 @@ import sys
import time import time
from pathlib import Path from pathlib import Path
DEFAULT_RECOMMENDATION_URL = "https://www.douyin.com/" from Douyin import DEFAULT_USER_URL
DEFAULT_CHROME_PATH = "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" DEFAULT_CHROME_PATH = "/Applications/Google Chrome.app/Contents/MacOS/Google Chrome"
DEFAULT_BROWSER_PORT = 9223 DEFAULT_BROWSER_PORT = 9223
@ -54,7 +54,7 @@ def build_parser() -> argparse.ArgumentParser:
default=DEFAULT_BROWSER_PORT, default=DEFAULT_BROWSER_PORT,
help="Chrome 调试端口，默认 9223", help="Chrome 调试端口，默认 9223",
) )
parser.add_argument("--user-url", default=DEFAULT_RECOMMENDATION_URL, help="启动后打开的抖音页面 URL，默认推荐流首页") parser.add_argument("--user-url", default=DEFAULT_USER_URL, help="启动后打开的抖音主页 URL")
return parser return parser