本文实例讲述了python Scrapy框架第一个入门程序。分享给大家供大家参考,具体如下:
首先创建项目:
scrappy start project maitian
第二步: 明确要抓取的字段items.py
import scrapy
class MaitianItem(scrapy.Item):
# define the fields for your item here like:
# name = scrapy.Field()
title = scrapy.Field()
price = scrapy.Field()
area = scrapy.Field()
district = scrapy.Field()
第三步: 在spider目录下创建爬虫文件: zufang_spider.py
2.1 创建一个类,并继承scrapy的一个子类: scrapy.Spider
2.2 自定义爬取名, name="" 后面运行框架需要用到;
2.3 定义爬取目标网址
2.4 定义scrapy的方法
下面是简单项目:
import scrapy
from maitian.items import MaitianItem
class MaitianSpider(scrapy.Spider):
name = "zufang"
start_urls = ['http://bj.maitian.cn/zfall/PG1']
def parse(self, response):
for zufang_itme in response.xpath('//div[@class="list_title"]'):
yield {
'title': zufang_itme.xpath('./h1/a/text()').extract_first().strip(),
'price': zufang_itme.xpath('./div[@class="the_price"]/ol/strong/span/text()').extract_first().strip(),
'area': zufang_itme.xpath('./p/span/text()').extract_first().replace('㎡', '').strip(),
'district': zufang_itme.xpath('./p//text()').re(r'昌平|朝阳|东城|大兴|丰台|海淀|石景山|顺义|通州|西城')[0],
}
next_page_url = response.xpath(
'//div[@id="paging"]/a[@class="down_page"]/@href').extract_first()
if next_page_url is not None:
yield scrapy.Request(response.urljoin(next_page_url))
第四步: 在settings.py文件中设置数据保存到数据库
.
.
.
ITEM_PIPELINES = {'maitian.pipelines.MaitianPipeline': 300,}
MONGODB_HOST = '127.0.0.1'
MONGODB_PORT = 27017
MONGODB_DBNAME = 'maitian'
MONGODB_DOCNAME = 'zufang'
第五步: 通过管道pipelines.py 连接上面的操作
import pymongo
from scrapy.conf import settings
class MaitianPipeline(object):
def __init__(self):
host = settings['MONGODB_HOST']
port = settings['MONGODB_PORT']
db_name = settings['MONGODB_DBNAME']
client = pymongo.MongoClient(host=host, port=port)
db = client[db_name]
self.post = db[settings['MONGODB_DOCNAME']]
def process_item(self, item, spider):
zufang = dict(item)
self.post.insert(zufang)
return item
其中 middlewares.py 文件暂不处理
以上简单的scrapy爬虫项目搭建完毕了
希望本文所述对大家基于Scrapy框架的Python程序设计有所帮助。
标签:scrapy
Powered By python教程网 鲁ICP备18013710号
python博客 - 小白学python最友好的网站!