Cover page getting duplicated in spine - user error or bug?

[ownedbycats]

I am converting some AZW3 books (mostly Amazon freebies acquired last year) to ePub.

Occasionally, the automatic split of chapter breaks will split apart the cover page. Then I get this:



This file is the cover page for the book. This file occurs more than once in the spine.

I'm usually able to fix this by either re-running the conversion without split or editing the content.opf.

Is this user error, or a bug?

[kovidgoyal]

https://www.mobileread.com/forums/sh...d.php?t=186697

[ownedbycats]

Something odd here: The conversion was done with Action Chains. Since that doesn't have a log, I reproduced all the settings I could and it didn't occur.

I'll ask in the plugin thread about it.

[DNSB]

I've seen that a couple of time when the cover image was tagged with a hx tag and a caption also had a hx tag. More commonly in the title page where the book title gets a h1 tag and the series/subtitle and author get a h2 tag. One reason why I use KindleUnpack for azw3 files that I obtain since it does not split files in the background and does spit out something much closer to the original ePub input file.

[ownedbycats]

@kovidgoyal: I just uploaded a ticket with the debug log and copyrighted book files. It's flagged security because I wasn't sure how else to make it private.

@DNSB - The H1/H2 does match up to what I saw too. I think opening the ePub directly in ebook-edit and saving as ePub will also get something close.

This is part of my 'send to device + add to reading list' Action Chain so I'm thinking I might be better off disabling the split entirely - especially since most books split as intended instead of having a single 500-page HTML file that freezes my Kobo.

[DNSB]

Originally Posted by ownedbycats

@kovidgoyal: I just uploaded a ticket with the debug log and copyrighted book files. It's flagged security because I wasn't sure how else to make it private.

@DNSB - The H1/H2 does match up to what I saw too. I think opening the ePub directly in ebook-edit and saving as ePub will also get something close.

This is part of my 'send to device + add to reading list' Action Chain so I'm thinking I might be better off disabling the split entirely - especially since most books split as intended instead of having a single 500-page HTML file that freezes my Kobo.

Or as suggested using KindleUnpack to create the epub from the azw3/KF8 file and avoiding many of the issues that I have with a calibre conversion.

[ownedbycats]

This was fixed for 6.27:

https://github.com/kovidgoyal/calibr...bfdbe00a98d5d0

[JSWolf]

Originally Posted by DNSB

I've seen that a couple of time when the cover image was tagged with a hx tag and a caption also had a hx tag. More commonly in the title page where the book title gets a h1 tag and the series/subtitle and author get a h2 tag. One reason why I use KindleUnpack for azw3 files that I obtain since it does not split files in the background and does spit out something much closer to the original ePub input file.

I find a calibre conversion better then KindleUnpack when the eBook it one big html file.

[ownedbycats]

Originally Posted by JSWolf

I find a calibre conversion better then KindleUnpack when the eBook it one big html file.

I find the 'big HTML file' a bit less common than properly split books, but I guess this depends on what kind of books you read.