[Python] 纯文本查看 复制代码
import osimport time
import re
from lxml import etree
import requests
from selenium import webdriver
import numpy as np
# 定义函数get_video_ids(author_url),返回UP主全部短视频的ID的列表
# 参数author_url:抖音UP主的主页
# 例如，XXX的主页 https://www.douyin.com/user/xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
def get_video_ids(author_url):
ids = list()
count = 0
retry = 0
n = 0
flag = True
chrome_option = webdriver.ChromeOptions()
chrome_option.add_argument('headless') # 静默模式
driver = webdriver.Chrome(options=chrome_option)
driver.get(author_url)
while flag and retry <= 15:
driver.execute_script("window.scrollBy(0,2000)") # scrollBy(x, y)，JavaScript中操作内容滚动指定的像素数
n = n + 1
time.sleep(2)
html_source = driver.page_source
items = etree.HTML(html_source).xpath("//li[@class='ECMy_Zdt']")
count_items = len(items)
print("操作页面内容滚动{0:0>3}次后,获取视频ID{1:0>4}个。".format(n, count_items))
if count_items != count:
count = count_items
else:
if retry < 2:
retry = retry + 1
else:
flag = False
print("已经达到可获取视频ID的最大数量,开始逐个获取视频ID:\n")
for item in items:
video_id = item.xpath("a/@href")[0].split("/")[-1]
print("获取短视频ID:{}".format(video_id))
ids.append(video_id)
return ids
# 定义函数get_author_nickname，得到UP主的抖音名称
# 参考函数get_video_info写的
def get_author_nickname(video_id):
url_author = "https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids=" + video_id
url0_author = requests.get(url_author, headers={"user-agent": "Mozilla/5.0"})
# 获取json数据中的抖音名称
author_nickname = url0_author.json()["item_list"][0]['author']['nickname']
return author_nickname
# 定义函数get_video_info(video_id),返回元组(短视频下载地址,短视频标题)
# 参数video_id:抖音短视频唯一ID
def get_video_info(video_id):
# 通过url0获取json数据(Chrome浏览器，F12进入开发者模式，模拟手机端，可以看到url0)
url0 = "https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids=" + video_id
r_url0 = requests.get(url0, headers={"user-agent": "Mozilla/5.0"})
# 获取json数据中的视频地址及视频标题,replace是用于下载无水印视频
url1 = r_url0.json()["item_list"][0]['video']["play_addr"]["url_list"][0].replace('/playwm/', '/play/')
# 防止出现空标题，加上短视频ID
# 判断首字母是否为#，True保留#后面的内容，False丢弃#后面的内容
title_temp=r_url0.json()["item_list"][0]['share_info']["share_title"]
if title_temp.startswith('#'):
title = video_id + "-" + title_temp.split("#")[1].split("@")[0]
else:
title = video_id + "-" + title_temp.split("#")[0].split("@")[0]
# 获取url1重定向后的真实视频地址
r_url1 = requests.get(url1, headers={"user-agent": "Mozilla/5.0"}, allow_redirects=False)
url = r_url1.headers['Location']
return url, title
# 定义函数get_file_name(string),从字符串中提取合法文件名
def get_file_name(string):
#pattern = re.compile(r'[?*/\\|.:><]')
#txt = re.sub(pattern, '', string)
#有文章说，上面这块儿代码可以简化为下面这句，增加对空格的替换
txt = re.sub('[?*/\\|.:>< ]',"",string)
return txt
# 定义函数download_video(save_path, url, title),下载并以短视频标题作为文件名保存短视频到指定路径
def download_video(save_path, url, title):
if os.path.exists(save_path):
pass
else:
os.makedirs(save_path)
with requests.get(url, headers={"user-agent": "Mozilla/5.0"}, stream=True) as r:
total_size = int(int(r.headers["Content-Length"]) / 1024 + 0.5)
file_name = get_file_name(title)
full_path = save_path + get_file_name(title) + ".mp4"
with open(file=full_path, mode="wb") as f:
print('当前下载:【{}】,视频文件大小:【{}KB】'.format(file_name, total_size))
count = 0
scale = 50
start = time.perf_counter()
for chunk in r.iter_content(chunk_size=1024):
f.write(chunk)
count = count + 1
i = int(scale * (count / total_size))
a = "=" * i
b = "." * (scale - i)
c = (i / scale) * 100
dur = time.perf_counter() - start
speed = count / dur
print("\r下载进度:{0:^3.0f}%[{1:}>{2:}] 耗时:{3:.2f}s 平均下载速度:{4:.2f}KB/S。".format(c, a, b, dur, speed), end="")
print("\n视频文件下载完毕,存放于:【{0:}】。".format(full_path))
# 定义主程序
def main():
# 获取UP主全部短视频的ID
url = input("请输入抖音UP主的主页:")
print("\n获取UP主全部短视频的ID...")
ids = get_video_ids(url)
print("获取完毕!共获取短视频ID{}个!".format(len(ids)))
# 获取UP主nickname
author_nickname = get_author_nickname(ids[0])
# 拿ids对比0_ID.txt文件中的数据，输出新的ids_new
cwd = os.getcwd()
video_path = cwd + "/" + author_nickname + "/"
if os.path.exists(video_path):
pass
else:
os.makedirs(video_path)
author_ID_path = video_path +"0_ID.txt"
with open(author_ID_path, 'a'):
if os.path.exists(author_ID_path):
ids_old = np.loadtxt(author_ID_path, dtype=str)
ids_new = list(set(ids).difference(set(ids_old)))
else:
pass
# 根据短视频ID,批量获取下载地址、短视频标题
print("\n根据短视频的ID获取短视频的下载地址、标题信息...")
videos_info = list()
for video_id in ids_new: #把in ids 改为 in ids_new
video_info = get_video_info(video_id)
videos_info.append(video_info)
print("短视频标题:【{0:}】;下载地址:【{1:}】".format(video_info[1], video_info[0]))
# 批量下载短视频
print("\n开始批量下载短视频:")
#cwd = os.getcwd()
#video_path = cwd + "/" + author_nickname + "/"
total = len(videos_info)
for i in range(total):
print("\n将下载第【{0:0>4}/{1:0>4}】个短视频:".format(i + 1, total))
print("=" * 50)
download_video(video_path, videos_info[i][0], videos_info[i][1])
# 将下载完成的ID写入0_ID.txt
with open(author_ID_path, 'a') as g:
g.write(ids_new[i] + '\n')
if __name__ == "__main__":
main()
[复制链接]
发表于 2022-3-28 17:26
|
发表于 2022-3-28 18:43
