[Conversion Plugin] KFX Input

[Martinoptic]

Calibre may not be able to convert it but the From KFX plugin should be able to - the message is quite specific: For best results use either the From KFX user interface plugin or the KFX Input Plugin CLI for conversion. See the KFX Input Plugin documentation for more information.
It is likely because calibre doesn’t support fixed layout conversion. The From KFX plugin bypasses the calibre conversion stage.

[jhowell]

Originally Posted by Anonymous Guy

Greetings, I'd like to report a bug.

It would seem that when converting KFX comics/manga, an error occurs.
Exception('This book has a layout that is incompatible with calibre conversion. For best results use either the From KFX user interface plugin or the KFX Input Plugin CLI for conversion. See the KFX Input Plugin documentation for more information.')"

It likely has something to do with right-to-left image layouts, but not sure.

The plugin is working as intended.

The issue has to do with books that use a fixed-layout, such as comics, manga, and print replica. Calibre's conversion system is designed for books that use reflowable layout. Using it on a fixed-layout book can mess with the book's formatting. For those books it is better to use either "From KFX" or the KFX Input CLI for conversion since those avoid the calibre conversion pipeline. If you decide to ignore this advice you can select the KFX Input conversion option to "Allow conversion to complete even if errors are detected".

For image based books such as comics/manga conversion to CBZ or PDF is generally superior to EPUB as an output format since those are well supported by many reading apps. An EPUB produced from a Kindle comic/manga will contain a lot of proprietary Amazon-specific formatting internally and may not render properly in a non-Amazon reading app.

[Anonymous Guy]

Originally Posted by jhowell

The plugin is working as intended.

The issue has to do with books that use a fixed-layout, such as comics, manga, and print replica. Calibre's conversion system is designed for books that use reflowable layout. Using it on a fixed-layout book can mess with the book's formatting. For those books it is better to use either "From KFX" or the KFX Input CLI for conversion since those avoid the calibre conversion pipeline. If you decide to ignore this advice you can select the KFX Input conversion option to "Allow conversion to complete even if errors are detected".

For image based books such as comics/manga conversion to CBZ or PDF is generally superior to EPUB as an output format since those are well supported by many reading apps. An EPUB produced from a Kindle comic/manga will contain a lot of proprietary Amazon-specific formatting internally and may not render properly in a non-Amazon reading app.

Thank you for the swift reply. I do not quite understand "From KFX" or the CLI. I have never heard of them. Can you go into more detail about how to use those? And yes, I have been using that "fix."

Understood about CBZ. Though my main concern is simply being able to extact the images themselves in the order they appear.

[Anonymous Guy]

Oof. I found From KFX. You have to add it to the toolbar manually. Now it works perfectly. Thanks!

[Anonymous Guy]

Though I have another issue. I have an uncompressed KFX file. Everything looks great, except the cover. For some reason the cover is a compressed jpg instead of an uncompressed png like everything else. I'm not sure if it's due to Calibre, the plugin, or what.

[jhowell]

Originally Posted by Anonymous Guy

I do not quite understand "From KFX" or the CLI. I have never heard of them. Can you go into more detail about how to use those?

The documentation for those features is in the first post of this thread.

Originally Posted by Anonymous Guy

Though I have another issue. I have an uncompressed KFX file. Everything looks great, except the cover. For some reason the cover is a compressed jpg instead of an uncompressed png like everything else. I'm not sure if it's due to Calibre, the plugin, or what.

That is normal for KFX comics. The cover is a single JPEG file and so that is retained unchanged. In most cases the following pages are each composed of a grid of multiple JPEG images. The plugin combines them into a single image per page and stores it in PNG format to avoid introducing any new compression artifacts.

[Anonymous Guy]

Originally Posted by jhowell

The documentation for those features is in the first post of this thread.



That is normal for KFX comics. The cover is a single JPEG file and so that is retained unchanged. In most cases the following pages are each composed of a grid of multiple JPEG images. The plugin combines them into a single image per page and stores it in PNG format to avoid introducing any new compression artifacts.

Thanks. Guess I'll have to look into other methods of getting hi-res comics without KFX. Amazon doesn't seem to deliver them to older versions of the Android app now. The method seems to be a closely guarded secret.

[AlfaeBooksManagerFan]

Is it possible to use this plugin standalone in Python, without Calibre?

python kfx_input.py
Traceback (most recent call last):
File "C:\Users\codrut\KFX_Input\kfx_input.py", line 6, in <module>
from calibre.gui2.convert import Widget
ModuleNotFoundError: No module named 'calibre'

that's because of:

from PyQt5 import (QtCore, QtWidgets)
from calibre.gui2.convert import Widget

[jhowell]

Originally Posted by codrut_popescu

Is it possible to use this plugin standalone in Python, without Calibre?

Not currently. The plugin does have a command line interface that bypasses the calibre GUI, but that still requires calibre to be installed. See the Command Line Interface section in the first post of this thread for more information.

There are a number of dependencies that are fulfilled by using calibre as the platform for running the plugin. It would be possible to change the plugin so that it could be run without calibre but doing that would increase the effort needed to support it so I have chosen to rely on calibre being present.

[fireattack]

I understand this project does not come with kfx_symbol_catalog.ion, but I wonder has anyone did so and shared such data?

Thanks in advance.

[jhowell]

Originally Posted by fireattack

I understand this project does not come with kfx_symbol_catalog.ion, but I wonder has anyone did so and shared such data?

You are the first person interested enough to ask about it here. I intentionally do not include the symbol catalog in the plugin to avoid copying anything significant directly from Amazon.

The format of the kfx_symbol_catalog.ion file should be:

$ion_1_0
$ion_shared_symbol_table::
{
  name: "YJ_symbols",
  version: 10,
  symbols: [
    "language",
    "font_family",
    "font_style",
    ...
    "standalone_entities",
    "document_regions",
    "yj.user_margin_bounds"
  ]
}

Where ... is replaced by around 800 additional symbols.

The full list of symbols can be obtained by decompiling the Java file EpubToKFXConverter-4.0.jar, which is part of Amazon's Kindle Previewer software, using a tool such as CFR.

In version 3.86 of the Previewer the enumeration corresponding to the symbol table (b_0) can be found in the class com.amazon.kaf.c.b. Those names will be different in other versions of the Previewer due to Java obfuscation.

[silverhikari]

want to know if am assuming right: that when using kfx input and adding a manga to calibre that it will grab the associated .azw.res files in the folder to get the high-res images and package it into .kfx file(if de-drm is working) and then using the "from kfx" option will convert the kfx file into a .cbz or pdf that will contain the high-res images?

P.S: does converting it to a pdf mess with the quality and should use cbz for the original quality or is it a 1:1 conversion?

[jhowell]

Originally Posted by silverhikari

want to know if am assuming right: that when using kfx input and adding a manga to calibre that it will grab the associated .azw.res files in the folder to get the high-res images and package it into .kfx file(if de-drm is working) and then using the "from kfx" option will convert the kfx file into a .cbz or pdf that will contain the high-res images?

Yes, the highest resolution that is in the set of files that make up the book will be used. However that resolution will vary greatly depending on which device type or app was used to download the KFX file. KFX sent to a Kindle is the worst. KFX sent to a PC is better. The best image quality comes from the Kindle apps for iOS or Android but DRM removal is not usually possible in that case.

Originally Posted by silverhikari

P.S: does converting it to a pdf mess with the quality and should use cbz for the original quality or is it a 1:1 conversion?

Stick with CBZ if the the original image quality matters to you.

[boksb]

from kindle for pc 2.3.1(70682) downloaded my purchased ebook and it shows as folder in kindle content with multiple files one named as BZ----.azw and others are named as CR!a-----.azw.res, CR!b----.azw.res and so on 12 files. once i import the azw file to calibre, it shows as KFX-ZIP and when i click Form KFX option from toolbar, it displays, "Found KFX-ZIP format" and "DRM Protected". i have all the latest dedrm tools and calibre also still not able to convert or view this files. help please

[jhowell]

Originally Posted by boksb

from kindle for pc 2.3.1(70682) downloaded my purchased ebook and it shows as folder in kindle content with multiple files one named as BZ----.azw and others are named as CR!a-----.azw.res, CR!b----.azw.res and so on 12 files.

That is normal. Books in KFX format are usually split up into multiple files. The AZW file is the main one. Importing that will cause this plugin to gather up the the files that make up the book into a KFX-ZIP file.

Originally Posted by boksb

i click Form KFX option from toolbar, it displays, "Found KFX-ZIP format" and "DRM Protected". i have all the latest dedrm tools and calibre also still not able to convert or view this files. help please

There are several versions of the DeDRM plugin that could be considered “the latest”. You need to use the auto-release version linked to in the first post in this thread under DRM Concerns.

If that does not solve the problem then ask for assistance in the Issues section at the site hosting the DRM removal plugin. Detailed discussion of DRM removal is not allowed here on MobileRead.