API Retrieve comic cover

[strauzen]

Hi,

It would be quite useful if the comic cover could be retrieved via API calls (as well as alternative covers) as for automatic comicbook metadata retrieval one of the checks that is usually done is validating of the covers match with a certain level of confidence.

[bpepple]

Oof, that's a really inefficient way to match comics. A much better solution is to provide a cover hash like was suggested in #615.

[strauzen]

Oof, that's a really inefficient way to match comics. A much better solution is to provide a cover hash like was suggested in #615.

That does sound like a nicer approach depending on the type of hash used and how it accounts for cover variations.

Sadly that doesn't work for other sources I use for metadata of European comics so the image comparison would still be required (and it's already implemented) for my use cases.

But if we have a nice hashed based solution for the covers that can account for layout changes then it could also do the job and my request would be supperceeded 🙂

An example of layout changes a hash can't account for but image comparison can are the covers of Asterix trench editions over the years. Correct me if I'm wrong, I'd love to improve my solution!

[bpepple]

An example of layout changes a hash can't account for but image comparison can are the covers of Asterix trench editions over the years. Correct me if I'm wrong, I'd love to improve my solution!

Are those the Mad Cave / Papercutz editions? And are indexed at GCD?

[strauzen]

No, I was referring to other usecases not related to GCD.

For Asterix I collect the comics and metadata in French. But I can imagine encountering similar situations with variations innGCD. I haven't dug enough into English comics yet to have a concrete example atm.

Example:

[jochengcd]

The API path to a cover will be via the issue. We won't provide direct access to covers in bulk.

[bpepple]

I don't think that he was asking for direct access to the covers in bulk thru an endpoint, or at least I hope he wasn't.

Most tagger software does a search by name, and then from that list they download each cover for the issues and do a comparison of cover hashes hamming distances to determine which one is correct. That's why we provide the cover hash in our issue list response, so they aren't downloading the issue covers needlessly.

[strauzen]

No indeed, not in bulk, you go issue by issue and retrieve the cover.

[DamianNiolet]

Hello, all. First off, apologies if this is in the wrong place; didn't see a better option and need urgent assistance. Second, greatly appreciate all of your generous efforts to create and improve GCD. I am using it as the basis for an app, which I'm using Claude Code to help me develop. I wasn't paying close enough attention, and Claude bombarded files1.comics.org with 16K cover requests at once, affectively block my CF worker. See below. How can I resolve this and get back on track? The original rate was within your guidelines and I will ensure Claud keeps it that way going forward. Just a little hiccup, but with huge impact on my app's usefulness. I appreciate any advice you can provide.

curl -I https://files1.comics.org/img/gcd/covers_by_id/0/w400/525.jpg
→ HTTP/2 403 · content-type: text/html · server: cloudflare (Cloudflare challenge)

curl https://www.comics.org/api/issue/1/?format=json
→ HTTP/2 200 · content-type: application/json (works, same server IP)

[gcd-github]

cloudflare blocks due to high load are temporarily some blocks for bots need to be always on Am 03.10.26 um 21:00 schrieb 'Damian Niolet' via gcd-tech:

…

Hello, all. First off, greatly appreciate all of your generous effort to create and improve GCD. I am using it as the basis for an app, which I'm using Claude to help me develop. I wasn't paying lose enough attention, and Claude bombarded files1 with 16K cover requests at once, affectively block my CF worker. See below. How can I resolve this and get back on track? The original rate was within your guidelines and I will ensure Claud keeps it that way going forward. Just a little hiccup, but with huge impact on my app's usefulness. I appreciate any advice you can provide. curl -I https://files1.comics.org/img/gcd/covers_by_id/0/w400/525.jpg <https://files1.comics.org/img/gcd/covers_by_id/0/w400/525.jpg> → HTTP/2 403 · content-type: text/html · server: cloudflare (Cloudflare challenge) curl https://www.comics.org/api/issue/1/?format=json <https:// www.comics.org/api/issue/1/?format=json> → HTTP/2 200 · content-type: application/json (works, same server IP) — Reply to this email directly, view it on GitHub <https://github.com/ GrandComicsDatabase/gcd-django#739? email_source=notifications&email_token=ADREI2E3QBYOFILFPKFDSID5SFEFZA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOBXGM2DONRSUZZGKYLTN5XKU43VMJZWG4TJMJSWJJLFOZSW45FMMZXW65DFOJPWG3DJMNVQ#discussioncomment-18734762>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADREI2ETF3P2LJDIH3KUY7L5SFEFZAVCNFSNUABGKJSXA33TNF2G64TZHM3DANRRGM2DSO2ENFZWG5LTONUW63R3GEYDKOBUHE2DPILWAI>. Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/ notifications/mobile/ios/ ADREI2D37XDFO62ZTTO3OH35SFEFZA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOBXGM2DONRSUZZGKYLTN5XKU43VMJZWG4TJMJSWJJLFOZSW45FKMZXW65DFOJPWS33T> and Android <https://github.com/notifications/mobile/android/ADREI2FBDHPX6ZQQCXIR34L5SFEFZA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOBXGM2DONRSUZZGKYLTN5XKU43VMJZWG4TJMJSWJJLFOZSW45FOMZXW65DFOJPWC3TEOJXWSZA>. Download it today! You are receiving this because you are subscribed to this thread.Message ID: <GrandComicsDatabase/gcd-django/repo-discussions/739/ ***@***.***> -- You received this message because you are subscribed to the Google Groups "gcd-tech" group. To unsubscribe from this group and stop receiving emails from it, send an email to ***@***.*** <mailto:gcd- ***@***.***>. To view this discussion visit https://groups.google.com/d/msgid/gcd- tech/GrandComicsDatabase/gcd-django/repo-discussions/739/ comments/18734762%40github.com <https://groups.google.com/d/msgid/gcd- tech/GrandComicsDatabase/gcd-django/repo-discussions/739/ comments/18734762%40github.com?utm_medium=email&utm_source=footer>.

[DamianNiolet]

Thank you for the reply. Do we know how temporary? I turned off my app completely for 24 hours to make sure it wasn't hitting your servers and it's still blocked. And if your servers think the app is a bot, how do we change that?

[gcd-github]

Cloudflare checks for bots, not we ourselves. For all access there is a cloudflare challenge at least once per day. This won't work with curl, but browsers. Only for API access we turned the challenge off. See also the covers/Cloudflare thread on gcd-tech mailinglist Am 04.10.26 um 09:42 schrieb 'Damian Niolet' via gcd-tech:

…

Thank you for the reply. Do we know how temporary? I turned off my app completely for 24 hours to make sure it wasn't hitting your servers and it's still blocked. And if your servers think the app is a bot, how do we change that? — Reply to this email directly, view it on GitHub <https://github.com/ GrandComicsDatabase/gcd-django#739? email_source=notifications&email_token=ADREI2D5P7NMKKTNMY2Z53T5SH5MVA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOBXGQYDOMZRUZZGKYLTN5XKOY3PNVWWK3TUUVSXMZLOOSWGM33PORSXEX3DNRUWG2Y#discussioncomment-18740731>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADREI2FMRSRUMHMWIN6MCGT5SH5MVAVCNFSNUABGKJSXA33TNF2G64TZHM3DANRRGM2DSO2ENFZWG5LTONUW63R3GEYDKOBUHE2DPILWAI>. Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/ notifications/mobile/ios/ ADREI2A3Q7AFOVYTQS57UCD5SH5MVA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOBXGQYDOMZRUZZGKYLTN5XKOY3PNVWWK3TUUVSXMZLOOSVGM33PORSXEX3JN5ZQ> and Android <https://github.com/notifications/mobile/android/ADREI2EDZQIM6E6JX5EX7LD5SH5MVA5CNFSNUABIM5UWIORPF5TWS5BNNB2WEL2ENFZWG5LTONUW63SDN5WW2ZLOOQXTCOBXGQYDOMZRUZZGKYLTN5XKOY3PNVWWK3TUUVSXMZLOOSXGM33PORSXEX3BNZSHE33JMQ>. Download it today! You are receiving this because you commented.Message ID: <GrandComicsDatabase/gcd-django/repo-discussions/739/ ***@***.***> -- You received this message because you are subscribed to the Google Groups "gcd-tech" group. To unsubscribe from this group and stop receiving emails from it, send an email to ***@***.*** <mailto:gcd- ***@***.***>. To view this discussion visit https://groups.google.com/d/msgid/gcd- tech/GrandComicsDatabase/gcd-django/repo-discussions/739/ comments/18740731%40github.com <https://groups.google.com/d/msgid/gcd- tech/GrandComicsDatabase/gcd-django/repo-discussions/739/ comments/18740731%40github.com?utm_medium=email&utm_source=footer>.

[DamianNiolet]

Thanks, that's clarifying — and I've been following the Covers/Cloudflare thread. Since my app fetches covers server-side (a script, not a browser), the signed-URL approach in #778 looks like exactly what I need: the issue API is already reachable for me (returns 200), so I can get the signed cover_url — the one open question is whether a script can then fetch that signed files1 URL and have the Cloudflare token rule accept it.

That's the piece still untested in the Covers/Cloudflare thread (beta's been tried by hand in a browser so far). I'd be glad to be the script-side test consumer once the token rule is live — a slow, small-batch, rate-limited run reporting exactly what a non-browser client sees. (My end is already throttled hard, so I'm not adding load in the meantime.)