[GUI Plugin] KindleUnpack - The Plugin

[PoP]

Just for information (at the occasion of playing with locking fonts of a book in this thread).

I noticed in the KindleUnpack of the sample AZW3, in the OPF <!-- BEGIN INFORMATION ONLY section, the entry <meta name="Unknown_(403)_(hex)" content="01" />

After removing this header, I found that the book opens with the ereader fonts instead of publisher fonts. I don't know which <meta name=?> triggers Kindlegen to generate the 403 header, but I thought KindleUnpack could possibly improve its decode from "Unknown_" to something like "Open With Publisher Fonts".

To illustrate:

Spoiler:

Before:
After:

[jhowell]

Originally Posted by PoP

Just for information (at the occasion of playing with locking fonts of a book in this thread).

I noticed in the KindleUnpack of the sample AZW3, in the OPF <!-- BEGIN INFORMATION ONLY section, the entry <meta name="Unknown_(403)_(hex)" content="01" />

After removing this header, I found that the book opens with the ereader fonts instead of publisher fonts. I don't know which <meta name=?> triggers Kindlegen to generate the 403 header, but I thought KindleUnpack could possibly improve its decode from "Unknown_" to something like "Open With Publisher Fonts".

I did some testing of the free sample of that book patched as you describe on my Kindle Oasis and have not been able to duplicate the results you describe. I am however seeing something very odd that works the same in both the patched and unpatched book.

I sideload the book and then immediately opened it and as you indicated the publisher fonts were not displayed. I then selected another font in the Aa menu and switched back to publisher font. After that the publisher font was displayed correctly.

In another test I sideload the book but then opened another book, closed it, and opened the sideloaded book. In that case the publisher fonts were displayed correctly from the start.

Finally I opened the sideloaded book and didn't see publisher fonts. Then closed it, opened another book, and went back to the sideloaded book. The publisher fonts were then shown.

So as far as I can tell there is a bug, but it does not seem to depend on the EXTH 403 field.

[PoP]

OK PoP just made another bad guess Perhaps <meta name="Mysterious_(403)_(hex)" content="01" /> is more appropriate then.

[jhowell]

Originally Posted by PoP

OK PoP just made another bad guess Perhaps <meta name="Mysterious_(403)_(hex)" content="01" /> is more appropriate then.

You made a reasonable guess.

I have be wondering about the purpose of EXTH 403 for a while. As far as I can tell it cannot be produced by kindlegen.

The other EXTH types numerically close to it control restrictions related to DRM, so it may be related to that. I surveyed the books I have purchased from Amazon a while ago and about 10% have it set to one, more have it set to zero. I haven't found any consistent difference in capabilities associated with it.

[mikapanja]

I used the plug-in to make the necessary edits to the MOBI file and now would like to repack it. What would be the easiest way to do so? Is there a KindleRepack plug-in or something like that? If not, would it be possible to make one?

[jhowell]

Originally Posted by mikapanja

I used the plug-in to make the necessary edits to the MOBI file and now would like to repack it. What would be the easiest way to do so? Is there a KindleRepack plug-in or something like that? If not, would it be possible to make one?

Kindleunpack does not just unpack the MOBI. It attempts to (as much as possible) undo the transformation performed by kindlegen when creating MOBI and KF8 format.

Repacking would mean passing the result back into kindlegen. (Kindlegen is now part of Amazon's Kindle Previewer.)

[democrite]

Hello Diap,

I am not sure if such has been mentioned before.

I think some of us purchase Kindle yet don’t have a Kindle device because sometimes Amazon is the only source for an ebook. As such, I think some of us always unpack azw4 and azw3 to EPUB. I am not familiar with the plug-in API yet it seems possible. Might you or someone someday be able to add an option to auto unpack such and add as a format? That would be so great and save effort for something that some of us always do.

[DNSB]

Originally Posted by democrite

Hello Diap,

I am not sure if such has been mentioned before.

I think some of us purchase Kindle yet don’t have a Kindle device because sometimes Amazon is the only source for an ebook. As such, I think some of us always unpack azw4 and azw3 to EPUB. I am not familiar with the plug-in API yet it seems possible. Might you or someone someday be able to add an option to auto unpack such and add as a format? That would be so great and save effort for something that some of us always do.

I don't feel that having the plugin insert itself into the book import chain is a worthwhile use of the programmer's time. You could use the ActionChains plugin if you wish to automate the process. You can see some possibilities in the Action Chains Scripts forum.

[DiapDealer]

I've got to work on making sure the plugin will work with the upcoming change to PyQt6 before considering anything new.

[DiapDealer]

Updated to v0.83.1 to support the upcoming Calibre 6 and (Py)Qt6. Tested to work with the latest Calibre 6 betas and back as far as Calibre 0.8.60.

No new features or bugfixes--just a boring maintenance release to support the upcoming Calibre6.

[stefano.sb]

Originally Posted by DiapDealer

.....

There is no "easy and quick" way to determine a file's KF8-ness. Your way works as well as another. There are some scripts you could run to dump the mobi headers... you could also open the ebook in the Kindle App (PC/Mac). For the time being, KF8 books can only be read in single column view with the apps; so multi-column mode just won't change anything—telling you it's a KF8. But I don't know how long that might be relied on.

Hi DiapDealer and first of all thanks for your plug-in :-)

I have quoted this old post because I'd like to determinate if a MOBI file is a mobi7, mobi8 or it's a mobi7+mobi8 version without unpacking it. If I understand well when I convert an ebook using Calibre I can use MOBI Output option with OLD=Mobi7, NEW=mobi8 and BOTH=mobi7+mobi8

It would be nice to have an option in KindleUnpack plug-in that for every MOBI file selected sets a Calibre custom column with 0->MOBI type doesen't recognized, 1->MOBI7, 2->MOBI8, 3->BOTH

Is there in yuor source code a specific function that I can use to try to understand how to do that?

Thanks

Stefano

[DiapDealer]

I've no clue, unfortunately, where the best place to set a calibre custom column in this plugin would be. This plugin's codebase is something that I have to re-familiarize myself with whenever a Calibre update makes it necessary to revisit it. Plus... other than inserting an extracted epub/pdf back into the calibre library, I don't have a lot of interest in adding much in the way of calibre-specific features to this plugin. I'm not all that familiar with calibre's codebase--and utterly clueless when it comes to setting custom columns programmatically. Happy to consider incorporating a patch or pull-request if you (or someone else) get something non-invasive working, though.

As for what type of mobi it is before unpacking: the options available in the KindleUnpack menu should tell you that (with individual books). If there is no ability to split, or extract an Epub (or pdf), it's a plain-jane Mobi-7. If the option to extract an epub is present, it's a KF8. If the option to split is present, it's a combination Mobi7/KF8. If the option to extract a pdf is present, it's an azw4.

But I'm afraid modifying this plugin to merely mark books for later unpacking isn't very likely. Perhaps a new "What's My Mobi" type of plugin would be more suitable for that. This one is very much geared toward extracting or unpacking from the get-go, rather than just flagging

[stefano.sb]

Originally Posted by DiapDealer

.... Perhaps a new "What's My Mobi" type of plugin would be more suitable for that. This one is very much geared toward extracting or unpacking from the get-go, rather than just flagging

Thanks for the clear answer. I'm actually considering the idea of ​​creating a small ad-hoc plug-in that we can easily call "What's My Mobi" :-)

Looking in the code I found the "build_single_book_menus (self, book_id)" function in the action.py file which, according to the MOBI type of the selected book, it changes the menu on the fly. Starting from here I think I can realize the function I need ...

Stefano

[DiapDealer]

I think I actually have a "What's my Mobi?" type python script somewhere that generates a report/log about kindlebook files in directories. You're welcome to it if I can find it (and you can use it).

[stefano.sb]

Originally Posted by DiapDealer

I think I actually have a "What's my Mobi?" type python script somewhere that generates a report/log about kindlebook files in directories. You're welcome to it if I can find it (and you can use it).