温馨提示×

温馨提示×

您好,登录后才能下订单哦!

密码登录×
登录注册×
其他方式登录
点击 登录注册 即表示同意《亿速云用户服务条款》

Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词

发布时间:2022-03-04 09:39:33 来源:亿速云 阅读:133 作者:小新 栏目:开发技术

小编给大家分享一下Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词,希望大家阅读完这篇文章之后都有所收获,下面让我们一起去探讨吧!

环境部署

scrapy安装

pip install scrapy -i https://pypi.douban.com/simple

selenium安装

pip install selenium -i https://pypi.douban.com/simple

jieba安装

pip install jieba -i https://pypi.douban.com/simple

IDE:PyCharm

google chrome driver下载对应版本:google chrome driver下载地址

检查浏览器版本,下载对应版本。

Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词

实现过程

下面开始搞起。

创建项目

使用scrapy命令创建我们的项目。

scrapy startproject csdn_hot_words

项目结构,如同官方给出的结构。

Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词

定义Item实体

按照之前的逻辑,主要属性为标题关键词对应出现次数的字典。代码如下:

# Define here the models for your scraped items
#
# See documentation in:
# https://docs.scrapy.org/en/latest/topics/items.html
 
import scrapy
 
 
class CsdnHotWordsItem(scrapy.Item):
    # define the fields for your item here like:
    # name = scrapy.Field()
    words = scrapy.Field()

关键词提取工具

使用jieba分词获取工具。

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time    : 2021/11/5 23:47
# @Author  : 至尊宝
# @Site    : 
# @File    : analyse_sentence.py
 
import jieba.analyse
 
 
def get_key_word(sentence):
    result_dic = {}
    words_lis = jieba.analyse.extract_tags(
        sentence, topK=3, withWeight=True, allowPOS=())
    for word, flag in words_lis:
        if word in result_dic:
            result_dic[word] += 1
        else:
            result_dic[word] = 1
    return result_dic

爬虫构造

这里需要给爬虫初始化一个浏览器参数,用来实现页面的动态加载。

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time    : 2021/11/5 23:47
# @Author  : 至尊宝
# @Site    : 
# @File    : csdn.py
 
import scrapy
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
 
from csdn_hot_words.items import CsdnHotWordsItem
from csdn_hot_words.tools.analyse_sentence import get_key_word
 
 
class CsdnSpider(scrapy.Spider):
    name = 'csdn'
    # allowed_domains = ['blog.csdn.net']
    start_urls = ['https://blog.csdn.net/rank/list']
 
    def __init__(self):
        chrome_options = Options()
        chrome_options.add_argument('--headless')  # 使用无头谷歌浏览器模式
        chrome_options.add_argument('--disable-gpu')
        chrome_options.add_argument('--no-sandbox')
        self.browser = webdriver.Chrome(chrome_options=chrome_options,
                                        executable_path="E:\\chromedriver_win32\\chromedriver.exe")
        self.browser.set_page_load_timeout(30)
 
    def parse(self, response, **kwargs):
        titles = response.xpath("//div[@class='hosetitem-title']/a/text()")
        for x in titles:
            item = CsdnHotWordsItem()
            item['words'] = get_key_word(x.get())
            yield item

代码说明

1、这里使用的是chrome的无头模式,就不需要有个浏览器打开再访问,都是后台执行的。

2、需要添加chromedriver的执行文件地址。

3、在parse的部分,可以参考之前我文章的xpath,获取到标题并且调用关键词提取,构造item对象。

中间件代码构造

添加js代码执行内容。中间件完整代码:

# Define here the models for your spider middleware
#
# See documentation in:
# https://docs.scrapy.org/en/latest/topics/spider-middleware.html
 
from scrapy import signals
from scrapy.http import HtmlResponse
from selenium.common.exceptions import TimeoutException
import time
 
from selenium.webdriver.chrome.options import Options
 
# useful for handling different item types with a single interface
from itemadapter import is_item, ItemAdapter
 
 
class CsdnHotWordsSpiderMiddleware:
    # Not all methods need to be defined. If a method is not defined,
    # scrapy acts as if the spider middleware does not modify the
    # passed objects.
 
    @classmethod
    def from_crawler(cls, crawler):
        # This method is used by Scrapy to create your spiders.
        s = cls()
        crawler.signals.connect(s.spider_opened, signal=signals.spider_opened)
        return s
 
    def process_spider_input(self, response, spider):
        # Called for each response that goes through the spider
        # middleware and into the spider.
 
        # Should return None or raise an exception.
        return None
 
    def process_spider_output(self, response, result, spider):
        # Called with the results returned from the Spider, after
        # it has processed the response.
 
        # Must return an iterable of Request, or item objects.
        for i in result:
            yield i
 
    def process_spider_exception(self, response, exception, spider):
        # Called when a spider or process_spider_input() method
        # (from other spider middleware) raises an exception.
 
        # Should return either None or an iterable of Request or item objects.
        pass
 
    def process_start_requests(self, start_requests, spider):
        # Called with the start requests of the spider, and works
        # similarly to the process_spider_output() method, except
        # that it doesn't have a response associated.
 
        # Must return only requests (not items).
        for r in start_requests:
            yield r
 
    def spider_opened(self, spider):
        spider.logger.info('Spider opened: %s' % spider.name)
 
 
class CsdnHotWordsDownloaderMiddleware:
    # Not all methods need to be defined. If a method is not defined,
    # scrapy acts as if the downloader middleware does not modify the
    # passed objects.
 
    @classmethod
    def from_crawler(cls, crawler):
        # This method is used by Scrapy to create your spiders.
        s = cls()
        crawler.signals.connect(s.spider_opened, signal=signals.spider_opened)
        return s
 
    def process_request(self, request, spider):
        js = '''
                        let height = 0
                let interval = setInterval(() => {
                    window.scrollTo({
                        top: height,
                        behavior: "smooth"
                    });
                    height += 500
                }, 500);
                setTimeout(() => {
                    clearInterval(interval)
                }, 20000);
            '''
        try:
            spider.browser.get(request.url)
            spider.browser.execute_script(js)
            time.sleep(20)
            return HtmlResponse(url=spider.browser.current_url, body=spider.browser.page_source,
                                encoding="utf-8", request=request)
        except TimeoutException as e:
            print('超时异常:{}'.format(e))
            spider.browser.execute_script('window.stop()')
        finally:
            spider.browser.close()
 
    def process_response(self, request, response, spider):
        # Called with the response returned from the downloader.
 
        # Must either;
        # - return a Response object
        # - return a Request object
        # - or raise IgnoreRequest
        return response
 
    def process_exception(self, request, exception, spider):
        # Called when a download handler or a process_request()
        # (from other downloader middleware) raises an exception.
 
        # Must either:
        # - return None: continue processing this exception
        # - return a Response object: stops process_exception() chain
        # - return a Request object: stops process_exception() chain
        pass
 
    def spider_opened(self, spider):
        spider.logger.info('Spider opened: %s' % spider.name)

制作自定义pipeline

定义按照词频统计最终结果输出到文件。代码如下:

# Define your item pipelines here
#
# Don't forget to add your pipeline to the ITEM_PIPELINES setting
# See: https://docs.scrapy.org/en/latest/topics/item-pipeline.html
 
 
# useful for handling different item types with a single interface
from itemadapter import ItemAdapter
 
 
class CsdnHotWordsPipeline:
 
    def __init__(self):
        self.file = open('result.txt', 'w', encoding='utf-8')
        self.all_words = []
 
    def process_item(self, item, spider):
        self.all_words.append(item)
        return item
 
    def close_spider(self, spider):
        key_word_dic = {}
        for y in self.all_words:
            print(y)
            for k, v in y['words'].items():
                if k.lower() in key_word_dic:
                    key_word_dic[k.lower()] += v
                else:
                    key_word_dic[k.lower()] = v
        word_count_sort = sorted(key_word_dic.items(),
                                 key=lambda x: x[1], reverse=True)
        for word in word_count_sort:
            self.file.write('{},{}\n'.format(word[0], word[1]))
        self.file.close()

settings配置

配置上要做一些调整。如下调整:

# Scrapy settings for csdn_hot_words project
#
# For simplicity, this file contains only settings considered important or
# commonly used. You can find more settings consulting the documentation:
#
#     https://docs.scrapy.org/en/latest/topics/settings.html
#     https://docs.scrapy.org/en/latest/topics/downloader-middleware.html
#     https://docs.scrapy.org/en/latest/topics/spider-middleware.html
 
BOT_NAME = 'csdn_hot_words'
 
SPIDER_MODULES = ['csdn_hot_words.spiders']
NEWSPIDER_MODULE = 'csdn_hot_words.spiders'
 
# Crawl responsibly by identifying yourself (and your website) on the user-agent
# USER_AGENT = 'csdn_hot_words (+http://www.yourdomain.com)'
USER_AGENT = 'Mozilla/5.0'
 
# Obey robots.txt rules
ROBOTSTXT_OBEY = False
 
# Configure maximum concurrent requests performed by Scrapy (default: 16)
# CONCURRENT_REQUESTS = 32
 
# Configure a delay for requests for the same website (default: 0)
# See https://docs.scrapy.org/en/latest/topics/settings.html#download-delay
# See also autothrottle settings and docs
DOWNLOAD_DELAY = 30
# The download delay setting will honor only one of:
# CONCURRENT_REQUESTS_PER_DOMAIN = 16
# CONCURRENT_REQUESTS_PER_IP = 16
 
# Disable cookies (enabled by default)
COOKIES_ENABLED = False
 
# Disable Telnet Console (enabled by default)
# TELNETCONSOLE_ENABLED = False
 
# Override the default request headers:
DEFAULT_REQUEST_HEADERS = {
    'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
    'Accept-Language': 'en',
    'User-Agent': 'Mozilla/5.0 (Windows NT 6.2; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/27.0.1453.94 Safari/537.36'
}
 
# Enable or disable spider middlewares
# See https://docs.scrapy.org/en/latest/topics/spider-middleware.html
SPIDER_MIDDLEWARES = {
   'csdn_hot_words.middlewares.CsdnHotWordsSpiderMiddleware': 543,
}
 
# Enable or disable downloader middlewares
# See https://docs.scrapy.org/en/latest/topics/downloader-middleware.html
DOWNLOADER_MIDDLEWARES = {
   'csdn_hot_words.middlewares.CsdnHotWordsDownloaderMiddleware': 543,
}
 
# Enable or disable extensions
# See https://docs.scrapy.org/en/latest/topics/extensions.html
# EXTENSIONS = {
#    'scrapy.extensions.telnet.TelnetConsole': None,
# }
 
# Configure item pipelines
# See https://docs.scrapy.org/en/latest/topics/item-pipeline.html
ITEM_PIPELINES = {
    'csdn_hot_words.pipelines.CsdnHotWordsPipeline': 300,
}
 
# Enable and configure the AutoThrottle extension (disabled by default)
# See https://docs.scrapy.org/en/latest/topics/autothrottle.html
# AUTOTHROTTLE_ENABLED = True
# The initial download delay
# AUTOTHROTTLE_START_DELAY = 5
# The maximum download delay to be set in case of high latencies
# AUTOTHROTTLE_MAX_DELAY = 60
# The average number of requests Scrapy should be sending in parallel to
# each remote server
# AUTOTHROTTLE_TARGET_CONCURRENCY = 1.0
# Enable showing throttling stats for every response received:
# AUTOTHROTTLE_DEBUG = False
 
# Enable and configure HTTP caching (disabled by default)
# See https://docs.scrapy.org/en/latest/topics/downloader-middleware.html#httpcache-middleware-settings
# HTTPCACHE_ENABLED = True
# HTTPCACHE_EXPIRATION_SECS = 0
# HTTPCACHE_DIR = 'httpcache'
# HTTPCACHE_IGNORE_HTTP_CODES = []
# HTTPCACHE_STORAGE = 'scrapy.extensions.httpcache.FilesystemCacheStorage'

执行主程序

可以通过scrapy的命令执行,但是为了看日志方便,加了一个主程序代码。

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time    : 2021/11/5 22:41
# @Author  : 至尊宝
# @Site    : 
# @File    : main.py
from scrapy import cmdline
 
cmdline.execute('scrapy crawl csdn'.split())

执行结果

执行部分日志

Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词

得到result.txt结果。

Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词

看完了这篇文章,相信你对“Python如何通过Scrapy框架实现爬取CSDN全站热榜标题热词”有了一定的了解,如果想了解更多相关知识,欢迎关注亿速云行业资讯频道,感谢各位的阅读!

向AI问一下细节

免责声明:本站发布的内容(图片、视频和文字)以原创、转载和分享为主,文章观点不代表本网站立场,如果涉及侵权请联系站长邮箱:is@yisu.com进行举报,并提供相关证据,一经查实,将立刻删除涉嫌侵权内容。

AI