diff --git "a/\347\254\254\344\272\214\346\234\237\350\256\255\347\273\203\350\220\245/2\347\217\255/2\347\217\255_Wong/jd_crawler_sel/jd_search.py" "b/\347\254\254\344\272\214\346\234\237\350\256\255\347\273\203\350\220\245/2\347\217\255/2\347\217\255_Wong/jd_crawler_sel/jd_search.py" new file mode 100644 index 0000000000000000000000000000000000000000..36219c960656901b33ac89983a17b6be331f036b --- /dev/null +++ "b/\347\254\254\344\272\214\346\234\237\350\256\255\347\273\203\350\220\245/2\347\217\255/2\347\217\255_Wong/jd_crawler_sel/jd_search.py" @@ -0,0 +1,30 @@ +from selenium import webdriver +from week10.jd_parser import search +import time + +driver_path = r"E:\Chromedriver\chromedriver.exe" + +chrome_options = webdriver.ChromeOptions() +chrome_options.add_argument("disable-blink-features=AutomationControlled") +browser = webdriver.Chrome(executable_path=driver_path, chrome_options=chrome_options) + +def sim_search(keyword, url): + browser.get(url) + search_input = browser.find_element_by_css_selector('input[aria-label="搜索"]') + search_input.send_keys(keyword) + search_button = browser.find_element_by_css_selector('button[aria-label="搜索"]') + search_button.click() + + +def main(keyword, url): + sim_search(keyword, url) + time.sleep(5) + item_array = search.parse_jd_item(browser.page_source) + print(item_array) + browser.close() + + +if __name__ =="__main__": + jd_url = r"https://www.jd.com/" + main("笔记本电脑", jd_url) +