Missing Covers in AutoTag/Identify

[pod1908]

Each time I try and tag a book there are no cover images found. I have both Attempt to download cover boxes checked in the GCD metadata settings.

[mizaki]

I believe GCD have turned CloudFlare on more aggressively. In the logs you should see a message about Cloud Flare when covers fail to download. If that'll change with the API being worked on, I don't know.

[killo3967]

Could you use nodriver library for scraping instead of beautifulsoup?

https://github.com/ultrafunkamsterdam/nodriver

I had good results with cloudflare blocking, which is our case.

Here you have an example code for this

`''' get image with nodriver'''

import nodriver as uc
import time

async def main():

browser = await uc.start(
    headless=False,
    user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
    cookies={}
)
page = await browser.get("https://www.comics.org/issue/4303/cover/4/")
    
time.sleep(2)
    
elems = await page.select_all('*[src]')
for elem in elems:
    image = (elem.attributes[1])
    if "covers_by_id" in image:
        # Get the image URL
        image = image.split('?')[0].replace('//','/')
        print (f"Cover URL: {image}")
    
        # Save the image to a file
        await elem.save_screenshot('screenshot.jpg')
        break

# Close page
await page.close()

# Force browser to close
uc.cdp.browser.close

if name == 'main':
uc.loop().run_until_complete(main())
`

[mizaki]

Not seeing the covers in the GUI isn't the end of the world but for matching it's more of a problem. I have asked again for the inclusion of images hashes, so feel free to add support for the request: GrandComicsDatabase/gcd-django#615

For the official version, I don't see anything like that getting baked in. Possibly adding the ability to use a proxy FlareSolverr etc. is a maybe but I'd rather have it "fixed" on GCD side.

[killo3967]

I'm sorry, but I disagree. I think viewing the homepage is essential for identification when using "Search Online", and of course for "Auto-tag", unless a hash is used, but this creates another problem. What type of hash should be used? Will everyone use the same one?

What problem do you see with using nodriver? Are there any limitations or compatibility issues?

Flaresolver seems fine, but you're already dependent on third-party packages and where it's being implemented. In your plugin or in Comictagger?