amberwest/Jd

Python

Jd

使用scrapy_redis爬取京东评论

spider.py

一般的spider继承RedisSpider，全站爬取就继承RedisCrawlSpider
设置redis_key
不用写start_urls

settings.py

添加参数并开启redispipeline

# 确保所有的爬虫通过redis去重
DUPEFILTER_CLASS = "scrapy_redis.dupefilter.RFPDupeFilter"
# 启用Redis调度存储请求队列
SCHEDULER = "scrapy_redis.scheduler.Scheduler"
SCHEDULER_ORDER = 'BFO'
# 不清除redis队列，这样可以暂停/恢复 爬取
SCHEDULER_PERSIST = True
SCHEDULER_QUEUE_CLASS = 'scrapy_redis.queue.SpiderPriorityQueue'
ITEM_PIPELINES = {
    'scrapy_redis.pipelines.RedisPipeline': 300,
    }

设置数据库redis字段

REDIS_URL = None
REDIS_HOST = '127.0.0.1'
REDIS_PORT = 6379
REDIS_PARAMS = {
    'password': '',
    'db': 1
}

启动redis

登录redis

redis-server redis.conf
redis-cli
127.0.0.1:6379> auth password

选择对应数据库并输入start_url

127.0.0.1:6379> select 1
OK
127.0.0.1:6379[1]> lpush jd_comment:start_urls http://jd.com
(integer) 1