[Conversion Output] KFX Output

[jhowell]

Originally Posted by CyberPaul

ERRORE: KFX conversion failed: <p>TITLE<br><br><b>Kindle Previewer error:</b> Error(YJConvValidator):E120014: First Line Style applied to non content container. ContainerType=BOX, ContainerSubType=CONTENT_LIST</p>

Any error message that has "Kindle Previewer error" indicates that the Amazon Kindle Previewer failed to convert the book. The Previewer will only convert books that meet the strict (and undocumented) requirements needed to support Enhanced TYpesetting.

The rest of the error message comes from the Previewer and can provide some guidance as to what went wrong. In this particular case there is apparently a ::first-line selector in the book that the Previewer doesn't like.

You can probably fix this if you are willing and able to manually edit the book's contents. To do this requires some expertise in EPUB internals and use of an EPUB editor, such as the one included in calibre. I'm afraid that I can't help with that.

[CyberPaul]

You provided me a very helpful hint! I solved removing the first-line selector.
Thank you so much!

[auqui]

Hi!
Please can you help me with this error?

ERROR: KFX conversion failed: <p>TITLE<br><br><b>Kindle Previewer error:</b> Error(YJEpubAdapter):E00403: "Unsupported Transformer for tag : TagName : span Display : inline-block";</p>

Thanks.

[jhowell]

Originally Posted by auqui

Please can you help me with this error?

ERROR: KFX conversion failed: <p>TITLE<br><br><b>Kindle Previewer error:</b> Error(YJEpubAdapter):E00403: "Unsupported Transformer for tag : TagName : span Display : inline-block";</p>

This is another example of an error message produced by the Amazon Previewer, rather than the plugin itself. It appears to be reporting that there is a <span> element with style "display: inline-block" in the book and it doesn't allow that.

If you want this book to convert you will have to edit it manually. You are on your own for that.

[dumberock]

Originally Posted by jhowell

Version 1.2.0 - 15 Aug 2016

Add support for Kindle Previewer 3.4 beta.

thanks a lot, now I can join the party

tried some ebooks, about 2/3 working, quite expected.

hyphens work nicely in my native french, the conversion quality is so-so though, it's quite the "beta" indeed, doesn't even support <br/> (for instance long chapter titles or whatever list using it will still be split but with class margin between) or margins in percentage properly... ofc I should adress my complaints to amazon directly, and I will

[jhowell]

Originally Posted by dumberock

...the conversion quality is so-so though, it's quite the "beta" indeed...

Beside the conversion problems that appear to be caused by bugs there are also intentional transformations that occur in the conversion to KFX. These can be seen in the Previewer and also in books sold by Amazon. For example, I frequently see spacing adjusted to reduce white space on pages.

[halabund]

The plugin works well on the books I tried.

Is it also possible to use Calibre itself to send the book to the device in KFX format? I prefer to use Calibre to manage the books on the device.

If I right click a book and choose Send to Device -> Send Specific Format To -> Main Memory, then I get a list of formats to choose from. KFX is not in the list. I did verify that the KFX version exists on the disk (as converted with this plugin) and that Calibre knows about it (it does show up in View Specific Format).

Currently I just copy KFX books manually instead of using Calibre.

[theducks]

Originally Posted by halabund

The plugin works well on the books I tried.

Is it also possible to use Calibre itself to send the book to the device in KFX format? I prefer to use Calibre to manage the books on the device.

If I right click a book and choose Send to Device -> Send Specific Format To -> Main Memory, then I get a list of formats to choose from. KFX is not in the list. I did verify that the KFX version exists on the disk (as converted with this plugin) and that Calibre knows about it (it does show up in View Specific Format).

Currently I just copy KFX books manually instead of using Calibre.

I did not see KFX in the list of supported formats for Drivers Fire or Voyager.
Kinda hard to send without the ability to Tick (enable)

[jhowell]

Originally Posted by halabund

Is it also possible to use Calibre itself to send the book to the device in KFX format?

Calibre's current support for KFX is very limited. The kindle device driver is able to identify KFX books over USB as they appear when sent by Amazon. These books are mostly encrypted, but some metadata is readable in an unencrypted auxiliary KFX file.

The KFX Output conversion plugin produces books in a single unencrypted monolithic KFX file. There is currently no support in calibre for this KFX variant.

Besides sending monolithic KFX files to a Kindle, it would nice for calibre to have support for importing them and extracting their metadata. I will look into it. I am unfamiliar with the parts of the calibre source code that deal with interacting with devices and my available time is limited, so it may be a while before I can come up with a patch for calibre.

ETA: I have decided that adding better support for KFX in calibre would be more work than I am willing to put in on something that I do not need myself. It will have to wait for someone else with the proper skills and an interest in this to step forward.

[HowlerX]

Hi! Thanks for this wonderful plugin. Just wanted to inform everyone of a few things I've encountered when converting epubs to KFX format.

The plugin was recently updated to replace &nbsp; or just regular space within table data cells with a some text. I've found that you can replace those spaces with a zero-width non-joiner breaking space "&zwnj;" without altering the look of a simple html table.

Also bolded or italicized text of any kind within table data cells will break enhanced typesetting. To create pseudo-bolded text you can create a css style and use the "-webkit-text-stroke: 0.8px;" to make text look bolded. It's a workaroud and seems to work for Kindle PW2 and Voyage. That will allow "enhanced typesetting" to work. Still haven't figured out a workaround for italicized text, sorry.

I was wondering if there's a way we can manually upload a cover for books not available on Amazon. Perhaps in the *.sdr folder somewhere?

That's all for now. Thanks for releasing this great plugin!

[jhowell]

Originally Posted by HowlerX

The plugin was recently updated to replace &nbsp; or just regular space within table data cells with a some text. I've found that you can replace those spaces with a zero-width non-joiner breaking space "&zwnj;" without altering the look of a simple html table.

Thanks for the tip! Without some content in each table cell the Previewer fails to convert books. I tried a number of invisible characters and none of them worked so I had settled on using a small visible character. This will be changed to ZWNJ in the next plugin release.

Originally Posted by HowlerX

Also bolded or italicized text of any kind within table data cells will break enhanced typesetting. To create pseudo-bolded text you can create a css style and use the "-webkit-text-stroke: 0.8px;" to make text look bolded. It's a workaroud and seems to work for Kindle PW2 and Voyage. That will allow "enhanced typesetting" to work. Still haven't figured out a workaround for italicized text, sorry.

And another good tip. You have obviously spent some time on this.

Originally Posted by HowlerX

I was wondering if there's a way we can manually upload a cover for books not available on Amazon. Perhaps in the *.sdr folder somewhere?

The ExtractCoverThumbs program can examine the books on your kindle and generate any missing cover thumbnails, but it doesn't support KFX. Perhaps the developer would be interested in adding this.

[quiris]

Originally Posted by jhowell

The ExtractCoverThumbs program can examine the books on your kindle and generate any missing cover thumbnails, but it doesn't support KFX. Perhaps the developer would be interested in adding this.

The ExtractCoverThumbs depends on KindleUnpack work, https://github.com/quiris11/ExtractC...ndle_unpack.py so unfortunately I have to wait for KindleUnpack KFX support

[jhowell]

Version 1.3.0 - 20 Sep 2016

An invisible character will be added to empty table cells to work around a Previewer bug. (A visible character was used previously.)

If no ASIN is specified the converted book will be marked as a "personal document" and a unique ASIN will be generated. (An ASIN is required for the support of cover thumbnails on e-ink Kindles.)

Add support for Kindle Previewer 3.5 beta. (KFX generation from Previewer version 3.5 is unchanged from version 3.4.)

[jhowell]

Originally Posted by quiris

The ExtractCoverThumbs depends on KindleUnpack work, https://github.com/quiris11/ExtractC...ndle_unpack.py so unfortunately I have to wait for KindleUnpack KFX support

-- Test software removed since this has new been added to ExtractCoverThumbs. --

[maximalmobi7]

So I was playing around with kfx format plugin and found It seems to work best if I convert an epub format into the rb format, then selecting the rb format to kfx format, that seems to of gotten me past the converting bugs that happen when you try epub to kfx format, it keep crashing with the Conversion Errors, but when it took the rb format and converted it to epub in the process of outputting it to kfx format it seem to of gotten no conversion errors, at least that worked for one ebook so far, I hope that helps if some people keep getting Conversion Errors like me, I am just playing with the format of kfx format since its so new, it seem like an fun thing to try for ebooks on my kindle paperwhite 2 device, to see if they really look better than normal azw3 formats or not. I hope this information helps , it seems once you convert it to the rb format it removes any errors you may have with an epub to kfx format, again it seems to make the epub more compatible when first converting it to rb format from epub format, then selecting in calibre the rb format as the input then the kfx format as the output format under calibre again I hope this helps users who are having the problems with Conversion Errors from kfx format. I got some errors that the main page doesn't even show with epub format to kfx format, that's why I was trying to find an easy way to make most of my ebooks into kfx format it check it out in any case.

the only thing I need to work on is getting the thumbnail image showing up, and the font size correctly set in the kfx format since I need to set the font right under rb format , also I had to manually copy the kfx format to my kindle paperwrite 2 device, since the calibre cannot transfer both thumbnail image or kfx format to the device yet, I guess that because the format is too new, and they have not made an transfer mode yet for the ebooks, otherwise the ebook file seems to load fine under the device, I think the an extracting zip plugin for the thumbnail option I have yet to try the plugin yet, otherwise my method of converting the files seems to work.

I may of misunderstand the last post I thought it was an plugin for calibre but I think I was wrong, in any case, I have to try the python command manually or find another way to place the thumbnail under the device hard drive. like see if the thumbnails for mobi and azw3 work for the kfx format or not. I have to play around if it, in any case I hope calibre is updated at some point to support thumbnail and kfx format tranfering in the future versions of it.

so the extract_cover_thumbs modified for kfxmeta , I was not able to import the zip file I got an error when I tried I have to find the website again that had it and read more to see exactly what I did wrong, if anything, the ebook is easy to put on the device, I just have to figure out how to place the thumbnail image their for kfx format, otherwise it seems to work without an thumbnail, but I would prefer to be able to load one if I can figure it out, any help about the thumbnail, or when calibre be updated to transfer the kfx format and thumbnail for it too, also calibre has an plugin for x-ray files, but I don't think the new kfx format supports that feature as far as I can tell.
one last thing if I understand it right the thumbnail has something to do with its support of ASIN of the KFX file created when converted, I just have to make sure it matches the information under calibre, and see if it generates an thumbnail then or not, again I am new to the kfx format plugin under calibre. so I have to experiment and see if I can get the thumbnail to be included in the file or not.

to not get the error below when converting from epub format to RB format into kfx format, make sure your source from epub format to rb format, has under structure detection the box label insert metadata as page of start of book that feature needs to be uncheck at all times when going from epub format to rb format, also needs to be unchecked when converting from rb format to kfx format, otherwise you will get the error below, I was able to convert my ebook once I made sure that feature was not selected in calibre.

Error(YJConvValidator):E120013: Data table found in the book