how to fetch images in soup

[unkn0wn]

This is the html code in page.

/html/body/section/section/section/article/a/figure/picture 

<picture>
<source media="__" srcset="Actual link for image">
<img style="__" data-src="Image link"
</picture>

I am using below to fetch images.

for img in soup.findAll('img', attrs={'data-src'}):
            img['src'] = img['data-src']
        return soup

Above doesn't work, but html has the link to the image in .png extension.

How do i make it work

[kovidgoyal]

That will work fine provided you do it in preprocess_html() and data-src has the correct url.

[unkn0wn]

<div class="itgimage">
 <img class="lazyload" 
data-src="https://akm-img-a-in.tosshub.com/indiatoday/images/bodyeditor/202112/4-PK-b-x721.jpg?X3hMZ0Z5Dz9GQjbHLA1rUxxmdMmXNu8v" 
src="https://akm-img-a-in.tosshub.com/indiatoday/../sites/all/themes/itg/images/itg_image370x208.jpg" alt=""></div>

How do i get images from here? the data-src is the actual image link while the img attribute is a dummy image..

data-src has .jpg extension with ?______ .. if calibre can open this link it can save the image while fetching.. I think it fails to try opening this link as its not seen as image link..
EDIT
Okay looks like data-src isn't fetching

[kovidgoyal]

img['src'] = img['data-src'].split('?')[0]

[unkn0wn]

def preprocess_html(self, soup):
for img in soup.findAll


Does above not work when 'auto_cleanup = True'?

I think auto_cleanup ignores all images even after preprocess for finding images is used!

[kovidgoyal]

Yes, auto_cleanup does its own thing, you cant use any of the regular facilities with auto_cleanup.