码迷,mamicode.com
首页 > 其他好文 > 详细

微博抓取尝试

时间:2018-08-13 10:47:14      阅读:190      评论:0      收藏:0      [点我收藏+]

标签:tle   none   raw   add   extra   mpi   考点   dal   无限   

微信是比较封闭的,微博的好友信息比较开放,都可以抓到;

1)找人,通过关注列表;
2)提取出微博的数据,放到数据库;

微博昵称,头像;
关注,粉丝及微博数量;
根据一些基本的原则来决定是否将该用户的微博入待爬的队列;

指标:关注人数;
粉丝人数;但是有可能会很多人,而且有很多僵尸粉;(不好:第一,低效;
第二,平台也不会让你无限制的往下翻页,肯定会有限制)
微博数,粉丝数的数量是个重要的参考点;

怎么判断抓取的人不值得关注?可以先做一个定向的分析,分析你所抓的领域的人的微博大致情况;
    1)如果发布的微博数量特别少,可以认为是僵尸用户,不用爬;微博数小于某个下限;
    2)如果发布的微博数量特别多,比如每天发100多条,可能是小广告商或者机器人;
    3)对于转发的微博其实和僵尸的微博差不多,你可能爬了大量的微博发现都是重复的信息;

# 下面列出部分代码如下:

# -*- coding: utf-8 -*-
"""
Created on Sun Apr  1 10:18:42 2018

@author: Joe3223
"""
# -*- coding:utf-8 -*-
#!/usr/bin/env python3
import time
import os
import re
from bs4 import BeautifulSoup
from urllib.request import urlopen
from selenium import webdriver
from selenium.webdriver.common.desired_capabilities import DesiredCapabilities
#import pymongo
#from pymongo import MongoClient
import hashlib
from collections import deque
from lxml import etree
import threading

# 数据库的准备,这里用的是mongodb;
#client = MongoClient(‘localhost‘,27017)
#db = client.test
#followers = db.followers

# 注意:这里如果不设置user-agent,可能是无法跳转的
user_agent = (
    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_4) " +
    "AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.57 Safari/537.36"
)
##dcap = dict(DesiredCapabilities.PHANTOMJS)
##dcap["phantomjs.page.settings.userAgent"] = user_agent
dcap = dict(DesiredCapabilities.FIREFOX)
dcap["firefox.page.settings.userAgent"] = user_agent

#browserPath = ‘/opt/phantomjs-2.1.1-linux-x86_64/bin/phantomjs‘
#browserPath = ‘/usr/bin/phantomjs‘
# 基本参数的一些准备工作
parser = ‘html5lib‘
domain = "weibo.com"
url_home = "http://" + domain
download_bf = deque()                # 双向队列,用于保证多线程爬取是安全的
cur_queue = deque()
min_mblogs_allowed = 10              # 爬取的阈值设置
max_follow_fans_ratio_allowed = 3


# 这里有两个爬虫,一个爬取微博数据,一个爬取用户数据
weibo_driver = webdriver.Firefox()  # 微博爬虫
weibo_driver.set_window_size(1920, 1200)  # optional

# url入队列,当然,入队列前要先做查重    
def enqueueUrl(url):
    try:
        md5v = hashlib.md5(url).hexdigest()
        if md5v not in download_bf: # 去重
            print(url + ‘ is added to queue‘)
            cur_queue.append(url)
            download_bf.append(md5v)
        # else:
            # print ‘Skip %s‘ % (url)
    except ValueError:
        pass

# 队列左端弹出一个值
def dequeuUrl():
    return cur_queue.popleft()

# 到下一页取抓取        
def go_next_page(cur_driver):
    try:
        next_page = cur_driver.find_element_by_xpath(‘//a[contains(@class, "page next")]‘).get_attribute(‘href‘)
        print(‘next page is ‘ + next_page)
        cur_driver.get(next_page)
        time.sleep(3)
        return True
    except Exception:
        print(‘next page is not found‘)
        return False

# 通过xpath尝试获取元素,最多尝试6次    
def get_element_by_xpath(cur_driver, path):
    tried = 0
    while tried < 6:
        html = cur_driver.page_source
        tr = etree.HTML(html)
        elements = tr.xpath(path)
        if len(elements) == 0:
            time.sleep(1)
            continue
        return elements

# 滚屏,保证能抓到数据            
def scroll_to_bottom():
    # 最多尝试 50 次滚屏
    print(‘scroll down‘)
    for i in range(0,50):
        # print ‘scrolling for the %d time‘ % (i)
        weibo_driver.execute_script(‘window.scrollTo(0, document.body.scrollHeight)‘)
        html = weibo_driver.page_source
        tr = etree.HTML(html)
        next_page_url = tr.xpath(‘//a[contains(@class,"page next")]‘)
        if len(next_page_url) > 0:
            return next_page_url[0].get(‘href‘)
        if len(re.findall(‘点击重新载入‘, html)) > 0:
            print(‘scrolling failed, reload it‘)
            weibo_driver.find_element_by_link_text(‘点击重新载入‘).click()
        time.sleep(1)

# 提取微博数据
def extract_feed(feeds):
    for i in range(0,20):
    # 只有在抓取微博数据时需要滚屏
        scroll_to_bottom()
        for element in weibo_driver.find_elements_by_class_name(‘WB_detail‘):
            tried = 0
            while tried < 3:
                try:
                    feed = {}
                    feed[‘time‘] = element.find_element_by_xpath(‘.//div[@class="WB_from S_txt2"]‘).text
                    feed[‘content‘] = element.find_element_by_class_name(‘WB_text‘).text
                    feed[‘image_names‘] = []
                    for image in element.find_elements_by_xpath(‘.//li[contains(@class,"WB_pic")]/img‘):
                        feed[‘image_names‘].append(re.findall(‘/([^/]+)$‘, image.get_attribute(‘src‘)))
                    feeds.append(feed)
                    print(‘--------------------‘)
                    print(feed[‘time‘])
                    print(feed[‘content‘])
                    break
                except Exception:
                    tried += 1
                    time.sleep(1)
        # 微博信息的下一页
        if go_next_page(weibo_driver) is False:
            return feeds

def getFollows(pageInfo):
    pattern3 = re.compile(‘class="S_txt1" title="(.*?)".*?usercard‘)
    follows = re.findall(pattern3, pageInfo)
    print(follows)
    for i in follows:
        print(i)
        #follower = {"name":i,"type":"follower"}
        #rs = followers.insert_one(follower)
        #print(‘one insert:{0}‘.format(rs.inserted_id))
    
    ##urlsToScrawl = []
    ##urlsScrawled = []
    patterUrls = re.compile(‘<a bpfilter="page" class="page S_txt1"[\s\S]*?href="([\s\S]*?pids=Pl_Official_RelationMyfollow__92&amp;cfs=&amp;Pl_Official_RelationMyfollow__92_page=[\s\S]*?)"‘)
    follows = re.findall(patterUrls, pageInfo)
    for i in follows:
        print("http://weibo.com/"+i)
        ##if i not in urlsScrawled and i not in urlsToScrawl:
        ##urlsToScrapy.append("http://weibo.com/"+i)    


def login(current_driver,username, password):
    #driver = webdriver.PhantomJS(executable_path=browserPath)  #浏览器的地址
    #driver = webdriver.PhantomJS(desired_capabilities=dcap)
    #driver = webdriver.Firefox()
    #driver.set_window_size(1920, 1200)
    
    current_driver.get(url_home)  #访问目标网页地址
    #bsObj = BeautifulSoup(user_driver.page_source, parser)  #解析目标网页的 Html 源码
    time.sleep(10)
    #user_driver.save_screenshot("weiboLogin0.png")

    # 登录
    current_driver.find_element_by_id(‘loginname‘).send_keys(username)
    #user_driver.find_element_by_id(‘password‘).send_keys(password)
    #user_driver.find_element_by_xpath(‘//div[contains(@class,"input_wrap ")][0]/input‘).send_keys(password)
    current_driver.find_element_by_xpath(‘/html/body/div[1]/div[1]/div/div[2]/div[1]/div[2]/div/div[2]/div[1]/div[2]/div[1]/div/div/div/div[3]/div[2]/div/input‘).send_keys(password)
    # 执行 click()
    current_driver.find_element_by_xpath(‘//div[contains(@class,"login_btn")][1]/a‘).click()
    time.sleep(8)
    current_driver.save_screenshot("weiboLogin.png")

    ##verifyCode = input("Please input verify code:")            
    ##user_driver.find_element_by_xpath(‘/html/body/div[1]/div[1]/div/div[2]/div[1]/div[2]/div/div[2]/div[1]/div[2]/div[1]/div/div/div/div[3]/div[3]/div/input‘).send_keys(verifyCode)
    ##user_driver.find_element_by_xpath(‘//div[contains(@class,"login_btn")][1]/a‘).click()
    ##time.sleep(8)
    ##user_driver.save_screenshot("weiboLogin2.png")
    


def main(username, password):
    # 登录
    #login(user_driver,username, password)
    login(weibo_driver,username, password)
    
    # 等会操作
    time.sleep(30)
    #user_driver.save_screenshot("weibo.png")
          
    ## 从大V的入口进去爬取,真正的URL入口
    user_link = "https://weibo.com/u/3738542230?topnav=1&wvr=6&topsug=1&is_hot=1"
    print(‘downloading ‘ + user_link)
    weibo_driver.get(user_link)
    time.sleep(5)
    
    # 提取用户姓名
    account_name = get_element_by_xpath(weibo_driver, ‘//h1‘)[0].text
    photo = get_element_by_xpath(weibo_driver, ‘//p[@class="photo_wrap"]/img‘)[0].get(‘src‘)
    account_photo = re.findall(‘/([^/]+)$‘, photo)
    # 提取他的关注主页
    follows_link = get_element_by_xpath(weibo_driver, ‘//a[@class="t_link S_txt1"]‘)[0].get(‘href‘)
    print(‘account: ‘ + account_name)
    print(‘account_photo: ‘+account_photo[0])
    print(‘follows link is ‘ + follows_link)

    #user_driver.get("http"+follows_link)
    feeds = []
    #users = []
     # 起一个线程取获取微博数据
    t_feeds = threading.Thread(target=extract_feed, name=None, args=(feeds,))
    t_feeds.start()
    t_feeds.join()
    

if __name__ == ‘__main__‘:
    main("你的用户","你的密码")
    #login(user_driver,"570876459@qq.com", "xiaowuwu!!!")
    #login(weibo_driver,username, pass

微博抓取尝试

标签:tle   none   raw   add   extra   mpi   考点   dal   无限   

原文地址:https://www.cnblogs.com/daluozi/p/9466430.html

(0)
(0)
   
举报
评论 一句话评论(0
登录后才能评论!
© 2014 mamicode.com 版权所有  联系我们:gaon5@hotmail.com
迷上了代码!