Thread: How to add ISBN to an EPUB file, to have it visible on my Elipsa?
View Single Post
Old 10-08-2021, 12:07 AM   #5
davidfor
Grand Sorcerer
davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.davidfor ought to be getting tired of karma fortunes by now.
 
Posts: 24,905
Karma: 47303824
Join Date: Jul 2011
Location: Sydney, Australia
Device: Kobo:Touch,Glo, AuraH2O, GloHD,AuraONE, ClaraHD, Libra H2O; tolinoepos
Quote:
Originally Posted by Grimaud View Post


I read them in examples from EPUB 3.2 specifications (like this one) and creating a fake book with Sigil also suggested the same way of definining it. I don’t think any of them would explain an invalid way of defining an ISBN inside an opf file.
Nowhere has Kobo said they support the EPUB 3.2 specs. And, I wouldn't be surprised if no ereader or ereading app did. There isn't that much point as yet as it will be rare in actual books.
Quote:
After a small modification (the unique-identifier tag is mandatory according to epubcheck), it did work with a side loaded epub, but didn’t with the same epub ‘kepubified’, even after asking for a full scan with NickelMenu

Code:
<package version="3.0" id="package" unique-identifier="isbn" xml:lang="fr" xmlns="http://www.idpf.org/2007/opf">
  <metadata xmlns:dc="http://purl.org/dc/elements/1.1/">
		<dc:identifier id="isbn">isbn:9791039102759</dc:identifier>
The unique-identifier tag is mandatory for one dc:identifier in the OPF. It doesn't have to be the ISBN. And there is at least one line in there that suggests it shouldn't be. In any case, I was copying lines from books I have that work for a Kobo device to read the ISBN when importing the books.
Quote:
I didn’t know about that page. I’ll try to ask them, maybe they will have a bit of time to give me an answer
That page isn't really useful for this. It is about how to prepare books to be loaded to Kobo's CMS and sold through the store. I do not believe there is anything in it about how metadata is read from sideloaded books.
Quote:
Edit : I tested my first way of doing it one more time. With a side loaded .epub (something I didn't tried before), the urn:isbn:[isbn] is shown. With a kepubifyed epub, nope, nothing, what's an ISBN, etc

I tried with the exact same epub, only changing the extension from epub to kepub.epub: my Elipsa is able to read the ISBN from the .epub file, but unable from the .kepub.epub (which unlike previous tests where I used Kepubify, here, my .kepub.epub is litteraly the exact same file, except for the extension)
It would have been a good idea to say what format you were sending to the device. Your first post makes no mention of kepub. I think anyone who read it would assume you were sideloading epubs. I definitely did. And the only reason I tested kepubs was because I couldn't remember what happened. And the test I did was the simple rename as well.

But, there is no real reason to expect the same results for both epubs and kepubs. They are different formats, follow different specs and are handled by completely different code. Epubs will be handled by the Adobe RMSDK, so you get whatever that supports. Kepubs are handled by something else, and as sideloaded kepubs are not officially supported, there is no way to actually know what should be done with them.
davidfor is offline   Reply With Quote