[BUG] Error queueing book: 'NoneType' oject is not callable

[Ikepuska]

Bug Description

When attempting to download a book that does not exist on welib, an error is thrown and the download does not occur. It also does not move on to attempting AA.
I have tried this with both the built-in and external cloudflare bypasser, regardless the error is the same.
The current version also doesn't seem to use the ALLOW_USE_WELIB variable at all. The default of true doesn't appear in the logs, and setting the variable doesn't show up in the logs either.

Steps To Reproduce

1. Find a book that isn't on welib but is with AA. (In my case MD5 is dea3460fc2a6b33133fcca1bdc63b499
2. Select download, or info
3. Errors occur in the logs and nothing happens with downloading.

Expected Behavior

The download script determines that there is no available download from welib and moves on to AA and downloads successfully.

Debug Information

Important: Please enable debug mode before submitting a bug report by setting the environment variable:

DEBUG=true

_calibre-web-automated-calibre-web-automated-downloader-1_logs.txt
When running in debug mode, a DEBUG button will appear in the interface. Please click this button to generate a debug zip file and attach it to this issue.

Environment Information

• BASE OS: [e.g. Windows 10, Ubuntu 22.04]
• Proxmox VE running on an AMD platform

docker-compose

flaresolverr:
depends_on:
- calibre-web-automated
image: ghcr.io/flaresolverr/flaresolverr:latest
calibre-web-automated-downloader:
depends_on:
- calibre-web-automated
- flaresolverr
image: ghcr.io/calibrain/calibre-web-automated-book-downloader-extbp:latest
environment:
FLASK_PORT: 8084
LOG_LEVEL: debug
BOOK_LANGUAGE: en
ALLOW_USE_WELIB: false
TZ: America/New_York
APP_ENV: prod
UID: 1000
GID: 100
CWA_DB_PATH: /auth/app.db
EXT_BYPASSER_URL: http://flaresolverr:8191
ports:
- 8084:8084
restart: unless-stopped
volumes:
# This is where the books will be downloaded to, usually it would be
# the same as whatever you gave in "calibre-web-automated"
- /calibreweb/cwa-book-ingest:/cwa-book-ingest
# This is the location of CWA's app.db, which contains authentication
# details
- /calibreweb/config/app.db:/auth/app.db:ro

Additional Context

[julie-sullivan]

Did you try this fix listed in #286

[Ikepuska]

I did see that issue, I thought the PR was already included in the release I am using, but apparently not. The fix did work to allow the books to download, although I would say that the fact that it was ignoring the setting about allowing welib is also an issue.
However the filename after implementing the fix is stripping the extension and leaving the bare md5 with a . at the end and no extension. E.g. 85d2a1857fef21780f726379fd099135. instead of 85d2a1857fef21780f726379fd099135.epub This results in CWA not processing the file at all even when the download is successful. When downloading the books through the browser direct from the calibre-downloader they are all author, author(1), author(2), etc when downloading multiple books from the same author, which makes it hard to figure out which is which without opening them each one by one before renaming them to include the title.

[Caffa]

I tried the fix listed in #286 and originally I had this issue

calibre-web-automated-book-downloader-1  | 2025-10-15 01:18:15,258 - backend - ERROR - logger.py:16 - Error queueing book: 'NoneType' object is not callable
calibre-web-automated-book-downloader-1  | Traceback (most recent call last):
calibre-web-automated-book-downloader-1  |   File "/app/backend.py", line 69, in queue_book
calibre-web-automated-book-downloader-1  |     book_info = book_manager.get_book_info(book_id)
calibre-web-automated-book-downloader-1  |   File "/app/book_manager.py", line 150, in get_book_info
calibre-web-automated-book-downloader-1  |     return _parse_book_info_page(soup, book_id)
calibre-web-automated-book-downloader-1  |   File "/app/book_manager.py", line 244, in _parse_book_info_page
calibre-web-automated-book-downloader-1  |     title=divs[7].text.strip(),
calibre-web-automated-book-downloader-1  | TypeError: 'NoneType' object is not callable

But restarting the container helped. Steps I took:

1. Turn off container in orbstack aka docker-compose down
2. do docker-compose pull
3. docker compose up -d
4. docker exec -it HASH bash where the HASH is what I got from docker ps
5. edit that line
6. restart container

[DigitGram]

However the filename after implementing the fix is stripping the extension and leaving the bare md5 with a . at the end and no extension. E.g. 85d2a1857fef21780f726379fd099135. instead of 85d2a1857fef21780f726379fd099135.epub

+1
Can confirm.

I get "author." with no file extension instead.

[cavernerg]

However the filename after implementing the fix is stripping the extension and leaving the bare md5 with a . at the end and no extension. E.g. 85d2a1857fef21780f726379fd099135. instead of 85d2a1857fef21780f726379fd099135.epub

+1
Can confirm.

I get "author." with no file extension instead.

+1

and to make it worse: when downloading multiple books from the same author, just one file remains

[Ikepuska]

+1 Can confirm.

I get "author." with no file extension instead.

To be clear, if you access the container shell and look at the file that is actually in the relevant folder you see the md5. If you download the file via the webui, you get author

[calibrain]

Could you share the md5 of the book causing these issues ?

[Ikepuska]

It's not just one book, and the MD5 in the earlier post is an example.
85d2a1857fef21780f726379fd099135

[calibrain]

Oh, I am sorry, I did not release the fix, (it was available in main but not on latest )
Sorry about it ! It is now released under v0.2.4 and should be available in ~1h after deployment