1、简介
前面一篇告诉了大家如何爬取京东的评论,这次我来教大家如何大规模
并行的爬取自动获取商品id以及爬取评论
2、所需模块
除了上篇博客所需模块之外,这次需要加入selenium这个模块来爬取动态页面的数据
3、代码
代码下有注释,有疑问直接在下面评论
import requests
from bs4 import BeautifulSoup
import re
from selenium import webdriver
from collections import deque
import sys
import numpy as np
import jieba
import threading
from threading import current_thread,Lock
from time import ctime ,sleep
import pymysql
import json
import urllib
import math
import queue
class MyThread(threading.Thread):
def __init__(self, funcs, args, name=''):
threading.Thread.__init__(self)
self.funcs = funcs
self.name = name
self.args = args
def run(self):
self.funcs(*self.args)
def getContent(que):
while que:
try:
url = que.popleft()
print('正在爬的线程是'+current_thread().name+"爬的是"+url)
headers = {
'Accept': 'text / html, application / xhtml + xml, image / jxr, * / *',
'Accept - Encoding': 'gzip, deflate',
'Accept - Language': 'zh - Hans - CN, zh - Hans;q = 0.5',
'User - Agent': 'Mozilla / 5.0(Windows NT 10.0; Win64;x64) AppleWebKit / 537.36(KHTML, like Gecko) Chrome/51.0.2704.79 Safari/ 537.36Edge/14.14393',
'Connection': 'Keep - Alive'
}
req = requests.get(url, headers=headers)
req.encoding = 'gbk'
res = req.text
bs = BeautifulSoup(res)
bs = bs.find_all('div', class_='i-item')
total = []
for i in bs:
reg1 = i.find('dd')
reg2 = i.find('span', class_=re.compile(r'^sta'))
scores = reg2.get('class')
scores = ''.join(scores)
scores = str(scores)
if scores == 'starsa5' or scores == 'starsa4' or scores == 'starsa3':
j = 1
#
# j=1
#
if scores == 'starsa2' or scores == 'starsa1' or scores == 'starsa0':
j = 0
reg3 = re.compile("<[^>]*>")
content = reg3.sub('', reg1.prettify())
print(content)
total.append((content, scores, j))
db = pymysql.connect('localhost', 'root', 'xin123456789', 'test')
db.encoding = 'utf-8'
cursor = db.cursor()
cursor.execute('set names utf8')
sql = "INSERT INTO newjd (comment,scores ) VALUES ('%s','%s') "
sql2 = "INSERT INTO test1 (title) VALUES ('%s') "
cursor.execute(sql % (content, j))
db.commit()
cursor.close()
db.close()
sleep(3)
except Exception:
print('运行出错')
def main():
what = '手机'
what = urllib.parse.quote(what)
import queue
que = queue.Queue()
for i in range(10):
url = "http://search.jd.com/Search?keyword=%s&enc=utf-8&wq=%s&pvid=cf28fafa22df407d860947dac22f33fe&page=" % (what,what)
url=url + str(i)
que.put(url)
headers = {
'Accept': 'text / html, application / xhtml + xml, image / jxr, * / *',
'Accept - Encoding': 'gzip, deflate',
'Accept - Language': 'zh - Hans - CN, zh - Hans;q = 0.5',
'User - Agent': 'Mozilla / 5.0(Windows NT 10.0; Win64;x64) AppleWebKit / 537.36(KHTML, like Gecko) Chrome/51.0.2704.79 Safari/ 537.36Edge/14.14393',
'Connection': 'Keep - Alive'
}
driver = webdriver.PhantomJS()
id = []
while que:
url = que.get()
driver.get(url)
driver.implicitly_wait(10)
bs = BeautifulSoup(driver.page_source)
items = bs.find_all("li", class_="gl-item")
# print(items)
for i in items:
id.append(i.get('data-sku'))
print(i.get('data-sku'))
if que.empty():
break
que1 = deque()
visited = set()
commentpeypage = 30
for i in id:
itemsummaryurl = 'http://club.jd.com/ProductPageService.aspx?method=GetCommentSummaryBySkuId&referenceId=' + str(
i)
itemsummaryresponse = urllib.request.urlopen(itemsummaryurl)
itemsummaryjson_dict = json.loads(itemsummaryresponse.read().decode('utf-8'))
commentrange = int(math.ceil(itemsummaryjson_dict.get('CommentCount')) / commentpeypage)
for j in range(commentrange):
url = 'http://club.jd.com/review/' + str(i) + '-0-' + str(j) + '-0.html'
que1.append(url)
thread=[]
for i in range(4):
t = MyThread(getContent, (que1, ), name='thread' + str(i))
thread.append(t)
for i in range(4):
thread[i].start()
for i in range(4):
thread[i].join()
if __name__ =='__main__':
main()