IEEE Xplore批量下載2018，selenium右鍵另存為

語言：python

需要安裝的包： selenium, pywin32, chrome-driver

需要安裝的軟件：Chrome

下載頻率請不要過於頻繁。

替換url可以下載當前鏈接頁面中所有的文章。

右鍵另存為使用的是win32api的方法。

'''
@author:Gawen
'''
import requests
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.common.action_chains import ActionChains
from selenium.webdriver.common.keys import Keys
import time
import win32api
import win32con

#替換url可以下載當前頁面所有的文章
url = 'https://ieeexplore.ieee.org/xpl/mostRecentIssue.jsp?punumber=8240062&punumber=8240062&filter=issueId%20EQ%20%228252286%22&rowsPerPage=75&pageNumber=2&rowsPerPage=75'
fore = 'https://ieeexplore.ieee.org'
s = 'Download or View the PDF:'
sites = []
errtitle = []
r = requests.get(url)
html = r.content.decode('utf-8')
soup = BeautifulSoup(html,'lxml')
h3 = soup.find('div', class_='cf jrnl-results-filter').find_all('h3')
h3text = []
for h in h3:
    h3text.append(h.text.strip())
print(h3text)
for i in range(len(h3text)):
    if((soup.find('a', attrs={'aria-label':s+'  '+h3text[i]}))==None):
        errtitle.append(h3text[i])
        continue
    pdf = soup.find('a', attrs={'aria-label':s+'  '+h3text[i]})['href']
    print(fore+pdf)
    sites.append(fore+pdf)
driver = webdriver.Chrome()
driver.maximize_window()
for site in sites:
    driver.get(site)
    element = driver.find_element_by_css_selector('body')
    driver.implicitly_wait(20)
    time.sleep(3)
    ActionChains(driver).context_click(element).perform()
    win32api.keybd_event(65,win32con.KEYEVENTF_KEYUP,0)
    time.sleep(1)
    win32api.keybd_event(18,0,0,0)#left alt
    win32api.keybd_event(83,0,0,0)
    win32api.keybd_event(83,0,win32con.KEYEVENTF_KEYUP,0)
    win32api.keybd_event(18,0,win32con.KEYEVENTF_KEYUP,0)#left alt up
    time.sleep(20)
driver.close()

print(str(len(errtitle))+'篇文章下載失敗,分別為:')
for title in errtitle:
    print(title+'\n')