Scrapy爬取天天美剧封面照及剧集下载地址
其实我只是想试试爬取图片而已,先看看网页，需要爬的地方有两个，一是封面图，二是下载地址，挺简单的Item定义：import scrapy
class TiantianmeijuItem(scrapy.Item):
name = scrapy.Field()
p_w_picpath_urls = scrapy.Field()
p_w_picpaths = scrapy.Field()
p_w_picpath_paths = scrapy.Field()
episode = scrapy.Field()
episode_url = scrapy.Field()name是保存名字
p_w_picpath_urls和p_w_picpaths 是爬取图片的pipeline用的，一个是保存图片URL，一个是保存图片存放信息p_w_picpath_paths其实没什么实际作用，只是记录下载成功的图片地址epiosde和episode_url是保存集数和对应下载地址
Spider：import scrapy
from tiantianmeiju.items import TiantianmeijuItem
import sys
reload(sys) # Python2.5 初始化后会删除 sys.setdefaultencoding 这个方法，我们需要重新载入
sys.setdefaultencoding('utf-8')
class CacthUrlSpider(scrapy.Spider):
name = 'meiju'
allowed_domains = ['cn163.net']
start_urls = ["http://cn163.net/archives/{id}/".format(id=id) for id in ['16355', '13470', '18766', '18805']]
def parse(self, response):
item = TiantianmeijuItem()
item['name'] = response.xpath('//*[@id="content"]/div/div/h2/text()').extract()
item['p_w_picpath_urls'] = response.xpath('//*[@id="entry"]/div/img/@src').extract()
item['episode'] = response.xpath('//*[@id="entry"]/p/a/text()').extract()
item['episode_url'] = response.xpath('//*[@id="entry"]/p/a/@href').extract()
yield item</pre>页面比较简单<br />Pipelines：这里写了两个管道，一个是把下载链接保存到文件，一个是下载图片<pre class="brush:python;toolbar:false">import json import os
from scrapy.pipelines.p_w_picpaths import ImagesPipeline
from scrapy.exceptions import DropItem
from scrapy.http import Request
from settings import IMAGES_STORE
class TiantianmeijuPipeline(object):
def process_item(self, item, spider):
return item
class WriteToFilePipeline(object):
def process_item(self, item, spider):
item = dict(item)
FolderName = item['name'].replace('/', '')
downloadFile = 'download_urls.txt'
with open(os.path.join(IMAGES_STORE, FolderName, downloadFile), 'w') as file:
for name,url in zip(item['episode'], item['episode_url']):
file.write('{name}: {url}\n'.format(name=name, url=url))
return item
class MyImagesPipeline(ImagesPipeline):
def get_media_requests(self, item, info):
for p_w_picpath_url in item['p_w_picpath_urls']:
yield Request(p_w_picpath_url, meta={'item': item})
def item_completed(self, results, item, info):
p_w_picpath_paths = for ok,x in results if ok]
if not p_w_picpath_paths:
raise DropItem("Item contains no p_w_picpaths")
item['p_w_picpath_paths'] = p_w_picpath_paths
return item
def file_path(self, request, response=None, info=None): item = request.meta['item']
FolderName = item['name'].replace('/', '')
p_w_picpath_guid = request.url.split('/')[-1]
filename = u'{}/{}'.format(FolderName, p_w_picpath_guid)
return filename
get_media_requests和item_completed，因为默认的图片储存路径是<IMAGES_STORE>/full/3afec3b4765f8f0a07b78f98c07b83f013567a0a.jpg，我需要把full改成以美剧名字目录来保存，所以重写了file_path
settings打开pipelines相关配置：ITEM_PIPELINES = {
'tiantianmeiju.pipelines.WriteToFilePipeline': 2,
'tiantianmeiju.pipelines.MyImagesPipeline': 1,
}
IMAGES_STORE = os.path.join(os.getcwd(), 'p_w_picpath') # 图片存储路径
IMAGES_EXPIRES = 90
IMAGES_MIN_HEIGHT = 110
IMAGES_MIN_WIDTH = 110爬下来之后就是这个效果了：
页:
[1]