[Script request]: bookshelf (revival of Readarr)

[Shebuka]

Application Name

bookshelf

Website

https://github.com/pennydreadful/bookshelf

Description

bookshelf is a revival of Readarr. The images published are configured to use working Goodreads or Hardcover metadata out of the box.

Bookshelf is an ebook and audiobook collection manager for Usenet and BitTorrent users. It can monitor multiple RSS feeds for new books from your favorite authors and will grab, sort, and rename them. Note that only one type of a given book is supported. If you want both an audiobook and ebook of a given book you will need multiple instances.

Due Diligence

• I have searched existing scripts and found no duplicates.
• I have searched existing discussions and found no duplicate requests.

[Shebuka]

• The Hardcover metadata source should be the default, although it requires registration it is in all ways better than Goodreads.
• The Goodreads metadata source should be only used if you want a drop-in replacement for Readarr.

[tremor021]

It has no release on github

[Shebuka]

I'm new to PVE 😅, can you clarify:

• We can't use the container they provide?
• Should I ask the maintainer to create a release?

[tremor021]

Our scripts don't use docker containers, unless really necessary, and such apps are rare. We do bare-metal installs of github releases OR official packages.

So, in order for us to ship these apps, we need to have something to latch on to, either github release or a official install script or official packages. W/e this dev can provide will suffice, with some caveats:

1. If they provide official install script, all support is then up to them. We will not deal with it unless its a clear problem with some backend interaction that breaks the script.
2. If its a official .deb package or binary, its also up to the devs to deal with the bugs because in that case we are installing their own official packages or binaries (following his documentation)
3. If the app is built from source, we need clear documentation on how to install, update and maintain. Many of the apps don't have this sorted out, so we have to resort into our own reverse engineering of the github repo/docs/dockerfiles and that takes time. Any help from the devs that can make a bare-metal installation easy on the user is highly appreciated.

[Shebuka]

Given that:

• bookshelf is a fork of Readarr;
• Readarr has the same build structure as Radarr;

can we assume/test that the build from source will work in the same way?

So the maintainer of bookshelf needs to create a release, and we can copy the build and install scripts from Radarr, and everything should 'in theory' just work?

[tremor021]

Correct, we need a release tarball. Also, Readarr is not built from source, it has their own official binaries which we fetch like this:

curl -fsSL 'https://readarr.servarr.com/v1/update/develop/updatefile?os=linux&runtime=netcore&arch=x64' -o readarr.tar.gz

What i meant by "built from source", it means WE do all the commands to build it via our install script. If they have official prebuilt binaries then we don't have to do that. We will just use those

[Shebuka]

All clear now 👍