aykhans/PostScrape
1
0
Fork 0
mirror of https://github.com/aykhans/PostScrape.git synced 2025-04-21 00:07:16 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

images added

Browse Source
This commit is contained in:
ayxan 2022-09-06 04:25:11 +04:00
parent 67c45e3c70
commit bc3caf3b51
1 changed files with 3 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/post_scrape/spiders/car_spider.py
View File

@ -27,6 +27,8 @@ class ToScrapeCSSSpider(scrapy.Spider):
yield response.follow(f"https://turbo.az{next_page}", callback=self.parse, headers=self.headers) yield response.follow(f"https://turbo.az{next_page}", callback=self.parse, headers=self.headers)
def parse_detail_url(self, r): def parse_detail_url(self, r):
images = [r.xpath('//a[@class="product-photos-large"]/@href').get()]
images += r.xpath('//div[@class="product-photos"]/div/a/@href').getall()
if r.xpath('//div[@class="shop-container"]'): if r.xpath('//div[@class="shop-container"]'):
avto_salon = True avto_salon = True
phone = r.xpath('//div[@class="shop-contact--phones-list"]//a[@class="shop-contact--phones-number"]/text()').getall() phone = r.xpath('//div[@class="shop-contact--phones-list"]//a[@class="shop-contact--phones-number"]/text()').getall()
@ -90,6 +92,7 @@ class ToScrapeCSSSpider(scrapy.Spider):
yield { yield {
'url': r.url, 'url': r.url,
'avto_salon': avto_salon, 'avto_salon': avto_salon,
'images': images,
'phone': phone, 'phone': phone,
'extra_fields': extra_fields, 'extra_fields': extra_fields,
'description': description, 'description': description,