爬虫那些事儿

随机更换user-agent

每次url请求更换一次user-agent

1	pip install fake-useragent

settings

1 2 3 4

DOWNLOADER_MIDDLEWARES = {    # \'ArticleSpider.middlewares.MyCustomDownloaderMiddleware\': 543,       \'ArticleSpider.middlewares.RandomUserAgentMiddleware\': 400, }

middlewares 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

from fake_useragent import UserAgent   class RandomUserAgentMiddleware(object):     def __init__(self, crawler):         super(RandomUserAgentMiddleware, self).__init__()           self.ua = UserAgent()         # 若settings中没有设置RANDOM_UA_TYPE的值默认值为random，         # 从settings中获取RANDOM_UA_TYPE变量，值可以是 random ie chrome firefox safari opera msie         self.ua_type = crawler.settings.get(\'RANDOM_UA_TYPE\', \'random\')        @classmethod     def from_crawler(cls, crawler):         return cls(crawler)       def process_request(self, request, spider):         def get_ua():             \'\'\'根据settings的RANDOM_UA_TYPE变量设置每次请求的User-Agent\'\'\'             return getattr(self.ua, self.ua_type)           ua = get_ua()         request.headers.setdefault(\'User-Agent\', get_ua())

 

回到目录

ip代理

方案一：免费版

自定义函数获取网上的一些免费代理ip

settings

1 2 3

DOWNLOADER_MIDDLEWARES = {       \'ArticleSpider.middlewares.RandomProxyMiddleware\': 400, }

middlewares 

1 2 3 4

class RandomProxyMiddleware(object):     #动态设置ip代理     def process_request(self, request, spider):         request.meta["proxy"] = get_random_ip() # 这个自定义函数返回一个随机代理ip：port

方案二：收费版

github上scrapy-proxies等等

回到目录

在线打码

 编码识别：由于验证码识别难度大，而且易更新，所以编码识别验证码(不推荐)

 在线打码：调用已经开发好的在线验证码识别软件接口识别验证码。识别率在90%以上，并且效率高(推荐)

 人工打码：识别率近100%，但是成本高（用于复杂的）

回到目录

cookie禁用

一些网站会跟踪cookie，如果不需要登陆的网站，可禁用cookie，降低被ban概率，scrapy默认开启cookie

1	COOKIES_ENABLED = False

回到目录

自动限速

调整某些参数，如

1 2	AUTOTHROTTLE_ENABLED = True DOWNLOAD_DELAY = 3

回到目录

selenium　

官方文档　http://selenium-python-docs-zh.readthedocs.io/zh_CN/latest/

作用：浏览器操控

安装selenium

1	pip install selenium

下载对应浏览器的驱动　　

http://selenium-python.readthedocs.io/installation.html

第三方(微博)登录知乎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

import time from selenium import webdriver from scrapy.selector import Selector   browser = webdriver.Chrome(executable_path="D:/Package/chromedriver.exe") time.sleep(2)  # 延时为了让页面加载完   browser.get("https://www.zhihu.com/#signin") browser.find_element_by_css_selector(".qrcode-signin-cut-button").click() browser.find_element_by_css_selector(".signup-social-buttons").click() browser.find_element_by_css_selector(".js-bindweibo").click() #browser.switch_to.window(browser.window_handles[-1]) browser.find_element_by_css_selector(".WB_iptxt").send_keys("xxx") browser.find_element_by_css_selector("input[node-type=\'passwd\']").send_keys("xxx") browser.find_element_by_css_selector("a[node-type=\'submit\']").click() time.sleep(2) # 延时为了让页面加载完 browser.find_element_by_css_selector("a[node-type=\'submit\']").click()

第三方(QQ)登录知乎

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

# -*- coding: utf-8 -*- __author__ = \'hy\' import time from selenium import webdriver from scrapy.selector import Selector    browser = webdriver.Firefox(executable_path="D:/Package/geckodriver.exe") # browser.get("https://www.zhihu.com/#signin") time.sleep(2)    # 点击QQ browser.find_element_by_css_selector(".qrcode-signin-cut-button").click() browser.find_element_by_css_selector(".signup-social-buttons").click() time.sleep(2) browser.find_element_by_css_selector(".js-bindqq").click() time.sleep(5)    browser.switch_to.window(browser.window_handles[-1]) browser.switch_to.frame("ptlogin_iframe")  # iframe必须逐级切入    # 用户名 密码   # 隐藏初始界面 browser.execute_script(\'document.getElementById("qlogin").style="display: none;"\') browser.execute_script(\'document.getElementsByClassName("authLogin").style="display: none;"\') # 显示用户、密码输入界面 browser.execute_script(\'document.getElementById("web_qr_login").style="display: block;"\') # browser.evaluate_script(\'document.getElementById("batch_quto").contentEditable = true\') time.sleep(5)    # 输入用户、密码 elem_user = browser.find_element_by_name("u").send_keys("xxx") elem_pwd = browser.find_element_by_name("p").send_keys("xxx") elem_but = browser.find_element_by_id("login_button").click() time.sleep(5)

回到目录

scrapy集成selenium　　

为什么集成selenium

selenium取代下载器，编码难度大的操作交给selenium

优点：反爬虫难度大

缺点：同步selenium效率低，需要结合Twisted成异步

middleware方式

方式一

settings

1 2 3

DOWNLOADER_MIDDLEWARES = {       \'ArticleSpider.middlewares.JSPageMiddleware\':1, }

middlewares 　　

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

from selenium import webdriver from scrapy.http import HtmlResponse import time     class JSPageMiddleware(object):     def __init__(self): # 使用同一个self，保证只打开一个浏览器，所有spider使用一个浏览器         self.browser = webdriver.Chrome(executable_path="D:/Package/chromedriver.exe")         super(JSPageMiddleware, self).__init__()       # 通过chrome请求动态网页     def process_request(self, request, spider):         if spider.name == "jobbole":             # self.browser = webdriver.Chrome(executable_path="D:/Package/chromedriver.exe")             self.browser.get(request.url)             time.sleep(1)             print("访问:{0}".format(request.url))             # browser.quit()             return HtmlResponse(url=self.browser.current_url, body=self.browser.page_source,                                 encoding="utf-8", request=request)

方式二

middlewares 

1 2 3 4 5 6 7 8 9 10 11 12 13 14

from scrapy.http import HtmlResponse import time   class JSPageMiddleware(object):     # 通过chrome请求动态网页     def process_request(self, request, spider):         if spider.name == "jobbole":             # self.browser = webdriver.Chrome(executable_path="D:/Package/chromedriver.exe")             spider.browser.get(request.url)             time.sleep(1)             print("访问:{0}".format(request.url))             # browser.quit()             return HtmlResponse(url=spider.browser.current_url, body=spider.browser.page_source,                                 encoding="utf-8", request=request)

spider

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

from selenium import webdriver from scrapy.xlib.pydispatch import dispatcher from scrapy import signals   class JobboleSpider(scrapy.Spider):     name = \'jobbole\'     allowed_domains = [\'blog.jobbole.com\']     start_urls = [\'http://blog.jobbole.com/all-posts/\']       def __init__(self): # 使用同一个self，每个spider使用一个浏览器         self.browser = webdriver.Chrome(executable_path="D:/Package/chromedriver.exe")         super(JobboleSpider, self).__init__()         dispatcher.connect(self.spider_closed, signals.spider_closed)  # 爬虫关闭后       def spider_closed(self, spider):         self.browser.quit()

回到目录

scrapy集成selenium模拟登录

为什么不直接用selenium替代原生下载器？

selenium是同步的方式，如果每个页面采用selenium则导致爬虫效率极低，目前并没有scrapy中的Twisted结合selenium的异步方案，因此selenium不推荐替代原生下载器

scrapy集成selenium能做什么？

 由于模拟登录是编码很难解决的问题 ，因此采用selenium解决；其它页面继续用原生下载器的异步下载方案

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

# -*- coding: utf-8 -*- import re import datetime    try:     import urlparse as parse except:     from urllib import parse    import scrapy from selenium import webdriver import time    class ZhihuSpider(scrapy.Spider):     name = "zhihu"     allowed_domains = ["www.zhihu.com"]     start_urls = [\'https://www.zhihu.com/\']     login_cookies = []       headers = {         "HOST": "www.zhihu.com",         "Referer": "https://www.zhizhu.com",         \'User-Agent\': "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:51.0) Gecko/20100101 Firefox/51.0"     }        # selenium登录保存cookies     def get_cookies(self):         browser = webdriver.Chrome(executable_path="D:/Package/chromedriver.exe")         time.sleep(2)  # 延时为了让页面加载完            browser.get("https://www.zhihu.com/#signin")         browser.find_element_by_css_selector(".qrcode-signin-cut-button").click()         browser.find_element_by_css_selector(".signup-social-buttons").click()         browser.find_element_by_css_selector(".js-bindweibo").click()         # browser.switch_to.window(browser.window_handles[-1])         browser.find_element_by_css_selector(".WB_iptxt").send_keys("xxx")         browser.find_element_by_css_selector("input[node-type=\'passwd\']").send_keys("xxx")         browser.find_element_by_css_selector("a[node-type=\'submit\']").click()         time.sleep(2)  # 延时为了让页面加载完         browser.find_element_by_css_selector("a[node-type=\'submit\']").click()         login_cookies = browser.get_cookies()         browser.close()        # 第一步:先于parse方法执行，处理登陆逻辑。可以猜测，start_requests携带的cookie会给后续所有的访问自动带上     def start_requests(self):         return [scrapy.Request(\'https://www.zhihu.com/#signin\', headers=self.headers, cookies=self.login_cookies,                                callback=self.parse)]        # 第二步:处理登陆后的逻辑     def parse(self, response):         my_url= \'https://www.zhihu.com/people/edit\'  # 该页面是个人中心页，只有登录后才能访问         yield scrapy.Request(my_url, headers=self.headers)

回到目录

爬取知乎文章和问答　　

scrapy shell调试　　

1 2	scrapy shell -s USER_AGENT="Mozilla/5.0 (Windows NT 6.1; WOW64; rv:51.0) Gecko/20100101 Firefox/51.0" https://www.zhihu.com/question/56320032

 页面分析　　

chrome安装jsonview插件

xhr页面查看json数据，这样获取数据更轻松

 

表设计

为了避免可能解析不到的字段或无法插入的情况，需要给字段设置默认值

 

 

 settings

 item

 pipeline

 spider

回到目录

scrapy-redis分布式爬虫

优点：利用多台机器的宽带加速爬取，利用多台机器的ip加速爬取（单台机器需要限速防止ip被ban）

缺点：编码难度大于单机爬虫

分布式需要解决的问题 

requests队列集中管理

去重集中管理　　

windows安装redis

1	https://github.com/MicrosoftArchive/redis/releases

创建项目　　

1	scrapy startproject ScrapyRedisTest

scrapy-redis:  https://github.com/rmax/scrapy-redis　　

scrapy-redis源码分析　　

import redis


# For standalone use.
DUPEFILTER_KEY = \'dupefilter:%(timestamp)s\'

PIPELINE_KEY = \'%(spider)s:items\'

REDIS_CLS = redis.StrictRedis
REDIS_ENCODING = \'utf-8\'
# Sane connection defaults.
REDIS_PARAMS = {
    \'socket_timeout\': 30,
    \'socket_connect_timeout\': 30,
    \'retry_on_timeout\': True,
    \'encoding\': REDIS_ENCODING,
}

SCHEDULER_QUEUE_KEY = \'%(spider)s:requests\'
SCHEDULER_QUEUE_CLASS = \'scrapy_redis.queue.PriorityQueue\'
SCHEDULER_DUPEFILTER_KEY = \'%(spider)s:dupefilter\'
SCHEDULER_DUPEFILTER_CLASS = \'scrapy_redis.dupefilter.RFPDupeFilter\'

START_URLS_KEY = \'%(name)s:start_urls\'
START_URLS_AS_SET = False

import six

from scrapy.utils.misc import load_object

from . import defaults


# Shortcut maps \'setting name\' -> \'parmater name\'.
SETTINGS_PARAMS_MAP = {
    \'REDIS_URL\': \'url\',
    \'REDIS_HOST\': \'host\',
    \'REDIS_PORT\': \'port\',
    \'REDIS_ENCODING\': \'encoding\',
}


def get_redis_from_settings(settings):
    """Returns a redis client instance from given Scrapy settings object.

    This function uses ``get_client`` to instantiate the client and uses
    ``defaults.REDIS_PARAMS`` global as defaults values for the parameters. You
    can override them using the ``REDIS_PARAMS`` setting.

    Parameters
    ----------
    settings : Settings
        A scrapy settings object. See the supported settings below.

    Returns
    -------
    server
        Redis client instance.

    Other Parameters
    ----------------
    REDIS_URL : str, optional
        Server connection URL.
    REDIS_HOST : str, optional
        Server host.
    REDIS_PORT : str, optional
        Server port.
    REDIS_ENCODING : str, optional
        Data encoding.
    REDIS_PARAMS : dict, optional
        Additional client parameters.

    """
    # 把settings文件的配置和defaults配置更新到params
    params = defaults.REDIS_PARAMS.copy()
    params.update(settings.getdict(\'REDIS_PARAMS\'))
    # XXX: Deprecate REDIS_* settings.
    for source, dest in SETTINGS_PARAMS_MAP.items():
        val = settings.get(source)
        if val:
            params[dest] = val

    # Allow ``redis_cls`` to be a path to a class.
    if isinstance(params.get(\'redis_cls\'), six.string_types):
        params[\'redis_cls\'] = load_object(params[\'redis_cls\'])

    return get_redis(**params)  # 调用get_redis


# get_redis_from_settings函数的别名：from_settings，从这里可以知道这个文件是准备给其它文件调用的(这里没用。。)
# Backwards compatible alias.
from_settings = get_redis_from_settings


# 连接redis
def get_redis(**kwargs):
    """Returns a redis client instance.

    Parameters
    ----------
    redis_cls : class, optional
        Defaults to ``redis.StrictRedis``.
    url : str, optional
        If given, ``redis_cls.from_url`` is used to instantiate the class.
    **kwargs
        Extra parameters to be passed to the ``redis_cls`` class.

    Returns
    -------
    server
        Redis client instance.

    """
    redis_cls = kwargs.pop(\'redis_cls\', defaults.REDIS_CLS)
    url = kwargs.pop(\'url\', None)
    if url:
        return redis_cls.from_url(url, **kwargs)
    else:
        return redis_cls(**kwargs)