[FEATURE] Add ebook-only copy export to a CWA/watch folder

[SturmB]

Is your feature request related to a problem? Please describe.

I'm evaluating Chaptarr for a dual-format setup: ebooks go to Calibre-Web Automated (CWA), while audiobooks go to Audiobookshelf. The missing piece for my setup is a supported ebook-only copy handoff to CWA's ingest folder, without running an additional Calibre Content Server or maintaining a custom bridge script.

CWA consumes/removes files from its ingest folder and can rewrite ebooks during processing. That makes two seemingly simple configurations unsafe or unsuitable:

• Making the ingest folder Chaptarr's managed ebook root would leave Chaptarr tracking files that CWA removes.
• Hardlinking an ebook into ingest could expose the original torrent payload to CWA's in-place modifications and break seeding integrity.

This is a feature request based on evaluating v0.9.958 and the current integration proposals, not a claim that I've reproduced a Chaptarr failure in production.

Describe the solution you'd like

An optional connection/export destination that copies successfully imported ebooks to a configurable CWA/watch folder while retaining Chaptarr's normal managed library file.

The important behavior would be:

1. Ebooks only. Audiobooks continue to their separate managed root/Audiobookshelf route, with no global import-mode switching.
2. An independent export copy. Never move the managed file or hardlink the export to it/the torrent payload. CWA must be able to rewrite or delete the exported file without changing either source.
3. Safe publication. The consumer should see only a complete file, not a partially written copy. Filename collisions should not silently overwrite a different book.
4. Repeat-safe delivery. Retry/replay of the same import should not create uncontrolled duplicate exports, including after CWA consumes the previous copy. An intentional upgrade or explicit re-export should still be possible.
5. Visible failures and retry. Distinguish a successful library import from a failed export, expose the failure, and allow retry without re-downloading the book.
6. One-way handoff. CWA consuming a copy should not mark Chaptarr's managed file missing. This need does not require synchronizing deletions or writing Calibre's database directly.

A generic ebook copy-to-folder connection would meet the need; it needn't be tightly coupled to CWA. I'm open to a different implementation that provides those guarantees.

Describe alternatives you've considered

• Calibre Content Server connector (feature: calibre content server connector #123; earlier Calibre content server #89): related, but a different transport. It targets Calibre Content Server write endpoints; CWA's existing web endpoint is not that service. A copy-to-ingest option would work with CWA's existing ingestion workflow without another server.
• Custom import script: could filter ebooks and stage copies, but leaves retry, deduplication, error visibility, and ongoing compatibility in a separately maintained component.
• Use ingest as the managed root / hardlink into ingest: unsuitable for the lifecycle and mutation reasons above.
• Current setup: Bindery retains a managed ebook file and independently copies it to CWA ingest, while audiobooks take a separate route. That is the behavior I'm trying to preserve when evaluating Chaptarr, rather than requesting a database migration tool.

Additional context

Related proposal: #123

CWA documents its ingest-folder workflow here: https://github.com/crocodilestick/Calibre-Web-Automated#readme

I searched existing public issues, PRs, and discussions and found the Content Server proposals, but not a dedicated copy-to-ingest request. If this is already planned or covered elsewhere, please point me to the appropriate thread.

Media Type

Does this feature apply to:

• Audiobooks
• eBooks
• Both

Use Case

Docker-based book automation with qBittorrent seeding, CWA managing its Calibre library locally on a NAS, and Audiobookshelf serving a separate audiobook library. Chaptarr would retain its managed roots, deliver disposable ebook copies to the shared ingest folder, and leave CWA as the owner of Calibre database writes. No direct database writes over the network share, no changes to seeding files, and no manual switching between ebook and audiobook import modes.

[benjitobz]

Not exactly what you're asking for but there is also this in addition to the content server connector:
#132

Your number 6 is the exact reason I went the calibre content server route the ingest folder isn't exactly friendly for applications that track files. You could use a 'custom script' connector to achieve some of this.

Notes:

• This PR is massive and I foresee it needing a lot of testing and potentially reworking
• This PR also requires a working calibre content server available in CWA. I'll be creating a PR shortly.

[SturmB]

Thanks—after looking more closely, a Calibre-managed root may actually fit what I’m trying to achieve. My request was framed around preserving our existing CWA ingest workflow, but ingest itself isn’t a hard requirement. I’m open to Chaptarr directing imports and conversion through Calibre, with CWA continuing to provide web/Kobo access. I’ll follow #132 and the companion CWA work; the main things I’d want to validate are seeding safety, metadata/embedded-file handling, and preserving existing CWA/Kobo library state.