3 wk. ago • 97%

Searchable Ebook Library

Am using Calibre and audiobookshelf. I'd love a solution where I can search the actual contents of the books. Like being able to search for topics inside all of my books.

Would be a cool AI feature - similar to how immich works.

Does anyone have a solution for that?

14 comments

Calibre can create a full text index to search through everything (well, for files that actually contain the text, and it needs a lot of space).

14
I had the same idea a while back and was wondering why no one has implemented something like this yet. This seems like an actual useful application for LLMs.

I am using Zotero (Citation Management Software) to collect scientific Articles I have read. Sometimes I forget in which Article I read about something specific. A search, where you could describe what you are looking for in a sentence, which then returns the Article with the relevant part, would be a gamechanger.

8
- What you are looking for is a RAG and is one of the few legitimately useful implementations of LLMs outside the wall of hype.
  
  5
  
  Thanks for the link! Learned something new today.
  
  3
  
  Thanks for the link! Learned something new today.
  
  1
  
  nice, that is exactly what I'm looking for - thanks : )
  
  1
- I've had the idea for a while to use an LLM to gather metadata about books for me as well as generate tag lists for themes, plot, writing style, etc for everything in my ebook library. You could also generate non spoiler plot summaries and produce recommendations for similar books.
  
  1
- A search, where you could describe what you are looking for in a sentence, which then returns the Article with the relevant part, would be a gamechanger.
  
  Yeah, exactly that
  
  1
Paperlessngx will store pdfs and index their contents for searching. It's not necessarily meant for books but I think it would work.

5
Keep the actual epubs and search in those? (You basically want a transcript of a read out book...which is the book itself.)

3
- He wants all of his books in one index.
  
  5
  
  And I think they want a solution that'll index audio books, too. An LLM that'll listen to, transcribed, and index audio books.
  
  2
- yeah I'm only talking about ebooks - I just mentioned audiobookshelf because it can also do ebooks and I've read here that people use it as a ebook management thing
  
  1
I did this a few years ago with a stack of pi 4s connected to a four port PoE switch. One was an openWRT router, one was a plex server connected to some spinning discs via usb, and I had another you could plug an hdmi cable into and use to view the media. I eventually found out I could host the whole thing on a single pi, but it was still a fun project. Could probably do it all on a pi 5 with an nvme hat no problem. Might look into that when I get the spare tinkering money.

1

You've viewed 14 comments.