Losing details in content.opf when parsing for conversion

[JBReadin]

Hello!

First post here so I hope I describe this issue well for everyone's clarity

I am having an issue when converting a manga from .KPF to .KFX, though I have since noticed this issue when converting from and to any combination of types. The files convert without throwing any errors, but I lose a lot of the HTML tags and attributes in the parsing of the .OPF that are necessary to make the book function like a manga, for example:

• <meta name="fixed-layout" content="true"/>
• <meta name="original-resolution" content="..."/>
• <itemref ... properties="page-spread-left">
• etc.

all of which are defined as necessary by KDP here: Creating Fixed-Layout Books with Image Pop-Ups or Virtual Panels

I have attatched the debug folder from when I ran the converstion so that you can see the content.opf file losing these tags throughout the conversion

What I noticed is rather strange, that the initial content.opf file in the "input" folder has the correct tags and attributes, but after parsing, the content.opf file in the "parsed" folder doesn't have the correct tags or attributes anymore. For this reason I believe that the extension that reads .KPF is functioning correctly, and that it is some other component (please excuse that I am not familiar with the internal process of Calibre) which is responsible for parsing the input which is not catching and continuing the necessary tags and attributes

I really hope that this had made sense! And if there are any further questions I am more than happy to elaborate!

[jhowell]

The calibre conversion feature is designed for reflowable books. It does not support fixed layout formats such as Kindle comics.

My KFX Input plugin has a command line interface (CLI) that can be used to perform conversion of KPF to EPUB bypassing calibre's conversion pipeline. That will retain the fixed layout formatting of a comic. And the KFX Output plugin command line interface can convert KPF comics to KFX.

If you are instead starting from a KF8 (azw3) file you can use the KindleUnpack plugin to extract a fixed layout EPUB without conversion.

[JBReadin]

Originally Posted by jhowell

the KFX Output plugin command line interface can convert KPF comics to KFX

That's exactly what I needed! Thank you so much!

[Quoth]

Originally Posted by jhowell

And the KFX Output plugin command line interface can convert KPF comics to KFX.

Doesn't that need also some Amazon program installed? Conversion to KFX doesn't work on Linux, at least not for me. The only way I get KFX (for testing) is to upload epub to Amazon KDP, publish, and buy a copy using Whispernet delivery direct to a Kindle that supports KFX (like my PW3). Even Send-To-Kindle doesn't do the exact same epub to azw3 as KDP publishing or Calibre conversion (which is near identical to published azw3).

Amazon seem determined to make life difficult.

[jhowell]

Originally Posted by Quoth

Doesn't that need also some Amazon program installed?

Conversion from KPF to KFX does not require any extra software.

Conversion to KFX from other formats, such as EPUB, does require the Kindle Previewer to be installed.

[Quoth]

Originally Posted by jhowell

Conversion from KPF to KFX does not require any extra software.

Conversion to KFX from other formats, such as EPUB, does require the Kindle Previewer to be installed.

That's good to know!
Thanks.