Hi! I would like to try to work on this task, could I be assigned to this?

@jjessieyang I assigned you to work on it. Good luck!

I'm having some trouble setting things up locally. I think I've followed all the guides for setting up docker, production book data, etc. as best I can, but I still can't see any books other than what's provided in the pre-loaded dataset or links to view by Subjects. I've read through issue #1864 but I still can't get it to work locally. How should I go about working on this task? This is my first time contributing to an open source project so any advice is greatly appreciated. Thank you!

@jjessieyang
In this case, you don't need those particular books. However, if you want to load those you can run a command like this to load those particular books.
docker compose exec -e PYTHONPATH=. web bash -c "./scripts/copydocs.py --search 'author_key:OL12180252A' --search-limit 100"

Let me know if you have further questions.

Okay thank you! Do you have any suggestions on where I should look to fix this bug? I think I'm also just a bit confused why this book is the only one that has this error.

@jjessieyang, one approach can be to try to figure out what's different between the working case and the not working case. Here, it looks as if this cover is coming from Internet Archive's cover service rather than from Open Library's. Were I to hazard a guess as to what's going on, I would say that the places that the cover doesn't show up aren't using Internet Archive's cover service as a fall back.

If you visit https://openlibrary.org/works/OL33647732W.json and https://openlibrary.org/books/OL45606286M.json you'll see that this book doesn't have a cover, yet curiously one shows up. C.f. https://openlibrary.org/works/OL66554W.json, which does have a cover.

If you mouse over the cover image where it does display for the book in this issue, you can see that unlike many covers on Open Library, the link isn't to covers.openlibrary.org/something, but to https://archive.org/download/heraldosdelpacha0000chur/page/cover_w180_h360.jpg.

If you do git grep "archive.org/download" there are some promising targets, including return f"https://archive.org/download/{itemid}/page/cover_w{w}_h{h}.jpg" from openlibrary/plugins/upstream/models.py, which looks pretty spot on. That will lead to get_ia_cover() and get_cover_url(), which will probably begin to unravel the mystery:

Note: a better git grep might be git grep -E "download/.+/page/cover_.+"

From here you'll want to look into how the 'broken' images are being loaded. As a quick sketch, you might want to right click -> 'inspect' the elements that are broken. For example, if you do that on image in the lower left of the screen where it says "1 edition selected" (i.e. the place with the red arrow pointing to it in the first image in this issue summary), you'll see:

<img title="OL45606286M" src="https://covers.openlibrary.org/b/olid/OL45606286M-M.jpg?default=https://openlibrary.org/images/icons/avatar_book-lg.png">

With the help of git grep, you should find your way to utils/SelectionManager/SelectionManager.js. You'll see there that the image resolution strategy is different.

Hopefully that gets you started.

If you get stuck, please ask questions.