Plugin ideas... add yours here

[nqk]

Originally Posted by BetterRed

If you mean these The calibre:// URL scheme — calibre 7.23.0 documentation then you can put them Comments or a another long text column using Insert link tool on the Rich Text editor. I have added a Links custom column to my libraries where I put links to 'related' books etc.



BR

I mean regular html url that can be inserted in the comments field in book metadata. For instance the full url to a book is: <a href="http://192.168.2.30:8080/#book_id=12817&library_id=Calibre&panel=book_detai ls">Book name</a>

which can be shortened to <a href="#book_id=12817&panel=book_details">Book name</a>

This works fine in the Content Server, but does nothing, as expected in the Main Calibre, and I hope it would actually jump to the target book when clicking.



Sent from my Pixel 7 Pro using Tapatalk

[nqk]

Originally Posted by BetterRed

If you mean these The calibre:// URL scheme — calibre 7.23.0 documentation then you can put them Comments or a another long text column using Insert link tool on the Rich Text editor. I have added a Links custom column to my libraries where I put links to 'related' books etc.



BR

I would try to avoid creating more columns. You know in many cases the text in the Comments may say "this book is similar to that book", and I would just enter the url to that specific book in rich text format.

Or the url can be used to do searches in the database.

I have been using it this way in the Content Server for long, and I would hope I can do it in the Main Calibre. In my view, it is simpler and more convenient than creating a bunch of columns.

[BetterRed]

Originally Posted by nqk

I mean regular html url that can be inserted in the comments field in book metadata. For instance the full url to a book is: <a href="http://192.168.2.30:8080/#book_id=12817&library_id=Calibre&panel=book_detai ls">Book name</a>

which can be shortened to <a href="#book_id=12817&panel=book_details">Book name</a>

This works fine in the Content Server, but does nothing, as expected in the Main Calibre, and I hope it would actually jump to the target book when clicking.



Sent from my Pixel 7 Pro using Tapatalk

Originally Posted by nqk

I would try to avoid creating more columns. You know in many cases the text in the Comments may say "this book is similar to that book", and I would just enter the url to that specific book in rich text format.

Or the url can be used to do searches in the database.

I have been using it this way in the Content Server for long, and I would hope I can do it in the Main Calibre. In my view, it is simpler and more convenient than creating a bunch of columns.

I used to do something similar, but the day Kovid released the calibre URL scheme I decided to switch to using it - I dislike interactive web browser interfaces.

Added: for me the calibre URL scheme is the #1 calibre enhancement.

BR

[nqk]

calibre:// would be fine for me if it worked in the Content Server, which I use regularly to read books and leave comments... You know, I read on multiple mobile devices (phone, eink reader, tablet, PC) so I need the Content Server to synchronize my progress and annotations.

Sent from my Pixel 7 Pro using Tapatalk

[kovidgoyal]

Originally Posted by nqk

calibre:// would be fine for me if it worked in the Content Server, which I use regularly to read books and leave comments... You know, I read on multiple mobile devices (phone, eink reader, tablet, PC) so I need the Content Server to synchronize my progress and annotations.

Sent from my Pixel 7 Pro using Tapatalk

https://github.com/kovidgoyal/calibr...deb4031d8fa61d

[nqk]

Originally Posted by kovidgoyal

https://github.com/kovidgoyal/calibr...deb4031d8fa61d

Oh, that was quick.

Thank you, Kovid. I would update as soon as the new version is released.

Sent from my Pixel 7 Pro using Tapatalk

[StickyC]

Right now, when you grab metadata for a book, it'll search the various sources given for alternative covers. The results may or may not be a good match and may or may not be larger/better than the existing cover. The covers are sorted by size, so oftentimes a higher quality/size replacement is not the first on the list. This makes looking for larger covers require a lot of cognitive overhead.

While that part of Calibre that normally handles downloading/choosing covers is probably not really 'touchable' code (is it?). It shouldn't be too difficult to write a plugin that:

- Downloads alternative book covers for a book (probably need to throttle the requests to avoid getting blocked)
- Rejects any candidates smaller than the existing cover
- Generates a histogram of the existing cover image and remaining candidates
- Automatically replaces the existing cover with the downloaded match that is larger/better, or at least provides a list sorted by histogram similarity so that a likely better cover is shown first.

I have not actually tried doing cover comparisons this way, so I have no idea how effective it would actually be. I do know that a lot of "Duplicate image finder" apps use histograms to compare identical pics of different sizes.

Ideally, it'd be something you could run against a library or individual book with options for minimum size to seek replacements and/or minimum size increase and, of course, the ability to accept/decline. Possibly even a 'minimum match' for comparing histograms.

I've not looked into what it takes to interface with Calibre, but the histogram comparison element seems like it'd be fairly easy to implement.

[theducks]

@stickyC
Beware the covers with whitespace borders. (eg 600x600). I see these frequently from a Google search. The trimmed cover was nnnx230

IMHO use the Quality check Plugin to identify covers that are less than your desired minimum and then MANUALLY use the Get Covers part of Get Metadata to cherry pick (or just keep what you have)

[JSWolf]

I quite like the Apple Books covers plugin. I have it set to download the cover from the US & the UK.

[StickyC]

Originally Posted by theducks

@stickyC
Beware the covers with whitespace borders. (eg 600x600). I see these frequently from a Google search. The trimmed cover was nnnx230

Ahh, good point. Add that to the features - Trim whitespace around the perimeter

Originally Posted by theducks

IMHO use the Quality check Plugin to identify covers that are less than your desired minimum and then MANUALLY use the Get Covers part of Get Metadata to cherry pick (or just keep what you have)

I hadn't thought of that - it doesn't really fit in my workflow, but maybe it's time to change my workflow

[theducks]

I got tired and frustrated trying to get various templates to work upon adding books and discovered that once I learned REGEX (I already knew basic HTML from the 3.x days), That I could cast an eye on the code and fix many (perceived ) problems in less than a minute (I have made a collection of saved searches and snippets).
Since I now only do 1 or 2 books at after a shopping time, I find the GUI based workflow satisfactory.
As Always YMMV

[DNSB]

I've pretty much given up on trying to bulk download covers or other metadata. I don't do that many books at a time and it's faster to make sure I have the correct cover than to bulk download and go back to correct mistakes. Not to mention that bulk download is getting blocked more and after fewer requests often making it less useful.

[calmingcolor]

Kokoro is an open-weight TTS model (82M parameters). Its lightweight and delivers comparable quality to larger models. With Apache-licensed weights, it can be deployed anywhere! I think it would be great if someone could make a plug in... It is beyond me - but would be great to have something other than the Piper read aloud.

Is this a good idea? Possible?

https://github.com/hexgrad/kokoro?tab=readme-ov-file
https://github.com/Lyrcaxis/KokoroSharp

[kovidgoyal]

text to speech engines aren't implemented as plugins, so you would need to modify calibre code to add a new backend.

[calmingcolor]

Originally Posted by kovidgoyal

text to speech engines aren't implemented as plugins, so you would need to modify calibre code to add a new backend.

Ah good to know! Thanks for the heads up before I started working on it. I could try and start pulling stuff together for a backend. Are you, or do you know anyone else I could contact who, is interested in extending TTS functionality and could help?
\The model is benchmarking close to eleven labs, but so much more lightweight!