scrapy只支持静态网页的抓取，通过scrapy-splash，能实现对JavaScript的解析。

一、搭建Docker服务器

Docker的相关知识参考本站的Docker相关文章。

Scrapy-Splash采用Splash HTTP API，需要搭建一个Splash实例，用Docker搭建最方便：

1
2
3

$ docker run -d -p 8050:8050 --restart=always --name=splash -m 200m scrapinghub/splash
 
# -m 200m 内存限制使用200兆

在服务器运行好Docker后，就可以通过IP+端口(例如：http://123.206.211.100:8050 )访问了。

二、Scrapy项目

1，安装scrapy-splash

1	$ pip install scrapy-splash

2，配置(setting.py)

增加Splash服务器地址

1	SPLASH_URL = 'http://123.206.211.100:8050'

开启Splash中间件

DOWNLOADER_MIDDLEWARES = {
    'scrapy_splash.SplashCookiesMiddleware': 723,
    'scrapy_splash.SplashMiddleware': 725,
    'scrapy.downloadermiddlewares.httpcompression.HttpCompressionMiddleware': 810,
    ...
}

其它设置

1
2
3

DUPEFILTER_CLASS = 'scrapy_splash.SplashAwareDupeFilter'
HTTPCACHE_STORAGE = 'scrapy_splash.SplashAwareFSCacheStorage'
SPLASH_COOKIES_DEBUG = True

3，spider.py使用SplashRequest

import scrapy
from scrapy_splash import SplashRequest

class SpiderS1(scrapy.Spider):

    name = "s1_spider"

    def start_requests(self):
    	
    	urls = ['http://sports.sina.com.cn/g/seriea/2017-05-23/doc-ifyfkqiv6736172.shtml',
                'http://sports.sina.com.cn/basketball/nba/2017-05-23/doc-ifyfkqiv6683532.shtml']
    	
        requests = []
        for url in urls:
       	    url = url.strip()
       	    request = SplashRequest(url, callback=self.parse, args={'wait':3})
        	requests.append(request)
        return requests
    
    def parse(self, response):
        self.log(response.url)
        ...

使用非常简单，具体请求参数参考 Scrapy&JavaScript integration through Splash。