[GUI Plugin] KindleUnpack - The Plugin

[theducks]

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).

That sounds like something that could/should be added to the Quality check PI.
MobiType: <type>

[stefano.sb]

Originally Posted by theducks

That sounds like something that could/should be added to the Quality check PI.
MobiType: <type>

Very good idea... In this plug-in it would be also very useful an option to store the output of a massive check (i.e. "missing EBOK cdetype") into two custom columns

• the check result
• last check running date

Stefano

[theducks]

Originally Posted by stefano.sb

Very good idea... In this plug-in it would be also very useful an option to store the output of a massive check (i.e. "missing EBOK cdetype") into two custom columns

• the check result
• last check running date

Stefano

I just add a Tag (I have a custom column #usertag i use for notes, which device has the book...) Multi-select (ctrl-a) the results, then use the Metadata editor (bulk mode)

QCPI already has some Mobi checks

[stefano.sb]

Originally Posted by theducks

I just add a Tag (I have a custom column #usertag i use for notes, which device has the book...) Multi-select (ctrl-a) the results, then use the Metadata editor (bulk mode)

QCPI already has some Mobi checks

Good suggestion! I will try to add the three new MOBI checks that I need to the QCPI (MOBI7, MOBI8, BOTH) using the same approach of the others check currently present (i.e. Chek Mobi -> check missing ASIN identifier)

After that in Calibre I will be able to use the Edit metadata in bulk function to set values in my custum columns

P.S. Next posts in the [GUI Plugin] Quality Check thread....

Thanks

Stefano

[DiapDealer]

Originally Posted by stefano.sb

P.S. Next posts in the [GUI Plugin] Quality Check thread....

I don't mind a short derailment, but this is probably a good idea.

Still looking for my script. I've found a few variations that use Python2/Qt4 (or Tkinter), but there should be something newer than that.

[stefano.sb]

Originally Posted by stefano.sb

Good suggestion! I will try to add the three new MOBI checks that I need to the QCPI (MOBI7, MOBI8, BOTH) using the same approach of the others check currently present (i.e. Chek Mobi -> check missing ASIN identifier)

After that in Calibre I will be able to use the Edit metadata in bulk function to set values in my custum columns

P.S. Next posts in the [GUI Plugin] Quality Check thread....

Thanks

Stefano

Here we are :-) I released this new function inside Quality Check Plug-in. See here

It’s a beta version… I am not a good Python programmer so if you want to test it or give me suggestions….

Thanks

Stefano

[motherat]

This is excellent, thanks so much for your work!

Question:

KindleUnpack can convert azw3 files to epubs directly via the KF8 to epub... menu item. For mobi files, nothing like this seems to be available -- the plugin can unpack them, but there's no option to turn them into epubs.

I assume that's working as intended? Or am I overlooking something? Is the only way to turn mobis into epubs Calibre's built-in conversion function?

[jhowell]

Originally Posted by motherat

KindleUnpack can convert azw3 files to epubs directly via the KF8 to epub... menu item. For mobi files, nothing like this seems to be available -- the plugin can unpack them, but there's no option to turn them into epubs.

The internal content of a KF8/azw3 file is very similar to EPUB so it is a fairly straightforward task to create an EPUB from it.

The content inside of the older MOBI format is based on a subset of HTML3. This needs extensive changes to turn it into a proper EPUB, which is something that would out of place in this plugin. It is better handled by calibre's conversion function.

[motherat]

Originally Posted by jhowell

The internal content of a KF8/azw3 file is very similar to EPUB so it is a fairly straightforward task to create an EPUB from it.

The content inside of the older MOBI format is based on a subset of HTML3. This needs extensive changes to turn it into a proper EPUB, which is something that would out of place in this plugin. It is better handled by calibre's conversion function.

I see -- thanks for explaining!

[PoP]

Any idea?

Unpacking this (de-DRMed) mobi7 dictionary:


I get this error:

calibre, version 6.4.0
ERREUR : Exception non gérée: <b>TypeError</b>:a bytes-like object is required, not 'str'

calibre 6.4  embedded-python: True
Windows-10-10.0.19044-SP0 Windows ('64bit', 'WindowsPE')
('Windows', '10', '10.0.19044')
Python 3.10.1
Windows: ('10', '10.0.19044', 'SP0', 'Multiprocessor Free')
Interface language: fr
Successfully initialized third party plugins: Gather KFX-ZIP (from KFX Input) (1, 49, 0) && DeDRM (10, 0, 1) && AZW6 Image Merge (1, 0, 0) && Package KFX (from KFX Input) (1, 49, 0) && APNX Generator (1, 1, 0) && Count Pages (1, 11, 3) && EpubCheck (0, 2, 4) && Find Duplicates (1, 9, 7) && Generate Cover (2, 2, 0) && Hyphenate This! (0, 1, 3) && KFX metadata reader (from KFX Input) (1, 49, 0) && KFX Input (1, 49, 0) && Set KFX metadata (from KFX Output) (1, 64, 0) && KFX Output (1, 64, 0) && KindleUnpack - The Plugin (0, 83, 1) && KindleXRay (0, 0, 4) && Obok DeDRM (7, 1, 0) && Open With (1, 7, 2) && Quality Check (1, 12, 0) && Standard Ebooks (1, 0, 0) && TTS to MP3 (0, 6, 0) && WordDumb (3, 25, 0) && X-Ray Creator (3, 2, 1)
Traceback (most recent call last):
  File "calibre_plugins.kindleunpack_plugin.action", line 271, in unpack_ebook
  File "calibre_plugins.kindleunpack_plugin.mobi_stuff", line 124, in unpackMOBI
  File "calibre_plugins.kindleunpack_plugin.kindleunpackcore.kindleunpack", line 934, in unpackBook
  File "calibre_plugins.kindleunpack_plugin.kindleunpackcore.kindleunpack", line 855, in process_all_mobi_headers
  File "calibre_plugins.kindleunpack_plugin.kindleunpackcore.kindleunpack", line 632, in processMobi7
  File "calibre_plugins.kindleunpack_plugin.kindleunpackcore.mobi_dict", line 205, in getPositionMap
  File "calibre_plugins.kindleunpack_plugin.kindleunpackcore.mobi_dict", line 288, in getInflectionGroups
  File "calibre_plugins.kindleunpack_plugin.kindleunpackcore.mobi_dict", line 377, in applyInflectionRule
AttributeError: 'array.array' object has no attribute 'tostring'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "calibre_plugins.kindleunpack_plugin.action", line 273, in unpack_ebook
  File "calibre_plugins.kindleunpack_plugin.utilities", line 105, in showErrorDlg
  File "traceback.py", line 179, in print_exc
  File "traceback.py", line 121, in print_exception
TypeError: a bytes-like object is required, not 'str'

At this point the images have been extracted, a toc.ncx has been created, but no html content.

[jhowell]

Originally Posted by PoP

AttributeError: 'array.array' object has no attribute 'tostring'

That error is most likely the result of a change to the array module in Python:

New in version 3.2: tostring() is renamed to tobytes() for clarity.

Added: I changed the name of that function in KindleUnpack and was able to unpack that dictionary without any further errors. I have attached a copy of the modified plugin in case anyone else needs to use it.

[PoP]

Thank you! Perfect

[DiapDealer]

Originally Posted by jhowell

That error is most likely the result of a change to the array module in Python:




Added: I changed the name of that function in KindleUnpack and was able to unpack that dictionary without any further errors. I have attached a copy of the modified plugin in case anyone else needs to use it.

Hmmm... Python 3.2 was long gone before calibre was altered to work with Python 3, so this has probably been broken in the plugin all along. I'll try to remember to release a version that will work with both old and new versions of calibre.

Feel free to PM me if something like this comes up in the future and I fail to respond.

[tomsem]

I tried converting a book with '&' in the title. It generates 3 FATAL errors with epub check. These need to be replaced by '&amp;'.

After I made these edits, there were still a couple of errors (AZW3 contains a fixed-layout format) but these don't seem to create issues, at least for Apple Books:

ERROR(HTM-046): Signs and Symbols_ An Illustrated Guide to - Dk.epub/OEBPS/Text/cover_page.xhtml(8,8): Fixed layout document has no "viewport" meta element.
ERROR(OPF-014): Signs and Symbols_ An Illustrated Guide to - Dk.epub/OEBPS/Text/cover_page.xhtml(-1,-1): The property "svg" should be declared in the OPF file.

By the way, thanks for this plugin.

Calibre is unable to convert this format to epub. Comic book/manga (made from series of image files) convert okay, it is those that start life as PDF that appear to be problematic.

Kovid does not claim calibre supports conversion to FXL epub (I think he has said 'does NOT support'), but it would be nice to have. I would do a PR if I had more confidence that I could figure it out, and didn't have more immediate demands on my time.

[pdurrant]

Kindle Unpack isn't/wasn't intended to produce a perfect epub. It's to extract the contents of a MOBI/AZW3/AZW4 file from the proprietary Amazon binary format into common format files.

That it can usually produce a usable ePub is a bonus.