gbanyan/Uptodatecrawler
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Found Uptodate API and apply with requests

Browse Source
This commit is contained in:
Gbanyan 2018-08-10 21:11:35 +08:00
parent edd4843f16
commit 2c327a733d
1 changed files with 13 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
Main.py
View File

@ -6,8 +6,10 @@ import html5lib
import re import re
import sys import sys
#Def the Uptodate URI for use
up_search_url = "https://www.uptodate.com/contents/search?search=" up_search_url = "https://www.uptodate.com/contents/search?search="
up_api_url = "https://www.uptodate.com/services/app/contents/search/2/json?&language=en&max=10&search="
up_prefix_url = "https://www.uptodate.com"
def do_uptodate_search_with_gecko(key_word): def do_uptodate_search_with_gecko(key_word):
print(up_search_url + key_word) print(up_search_url + key_word)
@ -19,18 +21,21 @@ def do_uptodate_search_with_gecko(key_word):
for links in articles_links: for links in articles_links:
print(links) print(links)
def do_uptodate_search(key_word): def do_uptodate_search_with_headless(key_word):
print(up_search_url + key_word) print(up_search_url + key_word)
driver = webdriver.Firefox() option = webdriver.ChromeOptions()
option.add_argument('headless')
driver = webdriver.Chrome(chrome_options=option)
driver.get(up_search_url + key_word) driver.get(up_search_url + key_word)
html = driver.page_source html = driver.page_source
soup = BeautifulSoup(html, 'html.parser') soup = BeautifulSoup(html, 'html.parser')
print(html) articles_links = soup.select("#search-results-container")
articles_links = soup.find_all("div.search-result") print(articles_links)
for link in articles_links: def do_uptodate_search_with_uptodate_api(key_word):
print(link) search_results = requests.get(up_api_url + key_word)
print(search_results.json())
if __name__ == '__main__': if __name__ == '__main__':
key_word = input("Please enter your keyword: ") key_word = input("Please enter your keyword: ")
do_uptodate_search_with_gecko(key_word) do_uptodate_search_with_uptodate_api(key_word)