使用Selenium模拟浏览器抓取淘宝商品美食信息

1.搜索关键词：利用Selenium驱动浏览器，得到商品列表。

2.分析页码并翻页：得到商品页码数，模拟翻页，得到后续页面的商品列表。

3.分析提取商品内容：利用PyQuery分析源码，解析得到商品列表。

4.存储到MongoDB：将商品列表信息存储到数据库MongoDB。

 1 from selenium import webdriver
 2 from selenium.common.exceptions import TimeoutException
 3 from selenium.webdriver.common.by import By
 4 from selenium.webdriver.support.ui import WebDriverWait
 5 from selenium.webdriver.support import expected_conditions as EC
 6 import re
 7 from pyquery import PyQuery as pq
 8 from urllib.parse import quote
 9 import pymongo
10 from config import *
11 
12 client = pymongo.MongoClient(MONGO_URL)
13 db = client[MONGO_DB]
14 browser = webdriver.Chrome()
15 wait = WebDriverWait(browser, 10)
16 
17 def search():
18     try:
19         browser.get('https://www.taobao.com')
20         input = wait.until(
21             EC.presence_of_element_located((By.CSS_SELECTOR, '#q'))
22         )
23         submit = wait.until(EC.element_to_be_clickable((By.CSS_SELECTOR, '#J_TSearchForm > div.search-button > button')))
24         input.send_keys('美食')
25         submit.click()
26         total = wait.until(EC.presence_of_element_located((By.CSS_SELECTOR, '#mainsrp-pager > div > div > div > div.total')))
27         get_products()
28         return total.text
29     except TimeoutEXception:
30         return search()
31 
32 def next_page(page_number):
33     try:
34         input = wait.until(
35             EC.presence_of_element_located((By.CSS_SELECTOR, '#mainsrp-pager > div > div > div > div.form > input'))
36         )
37         submit = wait.until(EC.element_to_be_clickable((By.CSS_SELECTOR, '#mainsrp-pager > div > div > div > div.form > span.btn.J_Submit')))
38         input.clear()
39         input.send_keys(page_number)
40         submit.click()
41         wait.until(EC.text_to_be_present_in_element((By.CSS_SELECTOR, '#mainsrp-pager > div > div > div > ul > li.item.active > span'), str(page_number)))
42         get_products()
43     except TimeoutException:
44         next_page(page_number)
45 
46 def get_products():
47     wait.until(EC.presence_of_element_located((By.CSS_SELECTOR, '#mainsrp-itemlist .items .item')))
48     html = browser.page_source
49     doc = pq(html)
50     items = doc('#mainsrp-itemlist .items .item').items()
51     for item in items:
52         product = {
53             'image': item.find('.pic .img').attr('src'),
54             'price': item.find('.price').text(),
55             'deal': item.find('.deal-cnt').text()[:-3],
56             'title': item.find('.title').text(),
57             'shop': item.find('.shop').text(),
58             'location': item.find('.location').text()
59         }
60         print(product)
61         save_to_mongo(product)
62 
63 def save_to_mongo(result):
64     try:
65         if db[MONGO_TABLE].insert(result):
66             print('保存MONGOBD成功',result)
67     except Exception:
68         print('存储到MONGODB失败',result)
69 
70 def main():
71     total = search()
72     total = int(re.compile('(\d+)').search(total).group(1))
73     for i in range(2,total+1):
74         next_page(i)
75     browser.close()
76 
77 if __name__ == '__main__':
78     main()

可以利用PhantomJS。

修改代码：

client = pymongo.MongoClient(MONGO_URL)
db = client[MONGO_DB]
#browser = webdriver.Chrome()
browser = webdriver.PhantomJS(service_args=SERVICE_ARGS)
wait = WebDriverWait(browser, 10)
browser.set_window_size(1400, 900)

MONGO_URL = 'localhost'
MONGO_DB = 'taobao'
MONGO_TABLE = 'products'

SERVICE_ARGS = ['--load-images=false', '--disk-cache=true']