Python爬虫：scrapy利用splash爬取动态网页

依赖库：

pip install scrapy-splash

配置settings.py

# splash服务器地址
SPLASH_URL = 'http://localhost:8050'

# 支持cache_args(可选)
SPIDER_MIDDLEWARES = {
    'scrapy_splash.SplashDeduplicateArgsMiddleware': 100,
}

# 下载中间件设置
DOWNLOADER_MIDDLEWARES = {
    'scrapy_splash.SplashCookiesMiddleware': 723,
    'scrapy_splash.SplashMiddleware': 725,
    'scrapy.downloadermiddlewares.httpcompression.HttpCompressionMiddleware': 810
}

# 设置去重过滤器
DUPEFILTER_CLASS = 'scrapy_splash.SplashAwareDupeFilter'

# 启用缓存系统
HTTPCACHE_STORAGE = 'scrapy_splash.SplashAwareFSCacheStorage'

代码示例

将原有​​Request​​​ 替换成 ​​SplashRequest​​即可

# -*- coding: utf-8 -*-
import scrapy
from scrapy import cmdline
from scrapy_splash import SplashRequest


class ToscrapeJsSpider(scrapy.Spider):
    name = "toscrape_js"
    allowed_domains = ["toscrape.com"]
    start_urls = (
        'http://quotes.toscrape.com/js/',
    )

    def start_requests(self):
        for url in self.start_urls:
            yield SplashRequest(url, args={"timeout": 5, "image": 0, 'wait': '5'})

    def parse(self, response):
        quotes = response.css(".quote .text::text").extract()
        for quote in quotes:
            print(quote)


if __name__ == '__main__':
    cmdline.execute("scrapy crawl toscrape_js".split())

​​SplashRequest​​参数说明

url: 待爬页面

headers: 与​​Request​​相同

cookies：与​​Request​​相同

args: {dict}传递给​​Splash​​的参数, wait很重要，需要给js足够的执行时间

cache_args: {list}让​​Splash​​缓存的参数

endpoint: ​​Splash​​端点服务，默认​​render.html​​

splash_url: ​​Splash​​服务器地址，默认为配置文件中的​​SPLASH_URL​​

设置代理：

args={
    'proxy': 'http://proxy_ip:proxy_port'
}

参考：
​​using proxy with scrapy-splash​​