python - 下载的图像是空白图像，而不是实际图像

Question

出于学习目的，我正在尝试下载 Buzzfeed 文章的所有帖子图像。

这是我的代码:

import lxml.html
import string
import random
import requests

url ='http://www.buzzfeed.com/mjs538/messages-from-creationists-to-people-who-believe-in-evolutio?bftw'
headers = headers = {
        'User-Agent': 'Mozilla/5.0',
        'From': 'admin@jhvisser.com'
}

page= requests.get(url)

tree = lxml.html.fromstring(page.content)

#print(soup.prettify()).encode('ascii', 'ignore')

images = tree.cssselect("div.sub_buzz_content img")

def id_generator(size=6, chars=string.ascii_uppercase + string.digits):
    return ''.join(random.choice(chars) for x in range(size))

for image in images:
    with open(id_generator() + '.jpg', 'wb') as handle:
        request = requests.get(image.attrib['src'], headers=headers, stream=True)

        for block in request.iter_content(1024):
            if not block:
                break
            handle.write(block)

检索到的都是110字节大小的图像，查看它们只是一张空图像。我在这里的代码中做错了什么导致了这个问题吗？如果有更简单的方法，我就不必使用请求。

Answer 1

如果您仔细查看要抓取的网页的源代码，您会发现您想要的图像网址并未在 img 的 src 属性中指定 标记，但位于 rel:bf_image_src 属性中。

将 image.attrib['src'] 更改为 image.attrib['rel:bf_image_src'] 应该可以解决您的问题。

我无法复制您的代码(它声称未安装 cssselect)，但此代码带有 BeautifulSoup和 urllib2在我的电脑上运行顺利，并下载了全部22张图片。

from itertools import count
from bs4 import BeautifulSoup
import urllib2
from time import sleep


url ='http://www.buzzfeed.com/mjs538/messages-from-creationists-to-people-who-believe-in-evolutio?bftw'
headers = {
    'User-Agent': 'Non-commercical crawler, Steinar Lima. Contact: https://stackoverflow.com/questions/21616904/images-downloaded-are-blank-images-instead-of-actual-images'
}

r = urllib2.Request(url, headers=headers)
soup = BeautifulSoup(urllib2.urlopen(r))
c = count()

for div in soup.find_all('div', id='buzz_sub_buzz'):
    for img in div.find_all('img'):
        print img['rel:bf_image_src']
        with open('images/{}.jpg'.format(next(c)), 'wb') as img_out:
            req = urllib2.Request(img['rel:bf_image_src'], headers=headers)
            img_out.write(urllib2.urlopen(req).read())
            sleep(5)