Tools Gathering requirements for Kindle Scribe Notebook Manager

[tomsem]

I am starting to work on a script to leverage @jhowell's stellar work on KFX Input plugin, which is now able to convert Scribe's nbk files into FXL ePub files that contain SVGs for each page (one for template, if any, another for the annotations).

These are some use cases / requirements I've collected so far:

Why KSNM?

• Amazon will store the latest version of a Notebook, but this will provide a way to review and even restore previous snapshots
• Preserves stroke detail, and has separate layers for template and user pen strokes. Amazon's current export flattens both to raster image for each page
• open source toolchain that can be modified as desired
• private, secure: does not use any internet services

Use cases

• enable backup and restore of specific Notebook versions
• script the conversion to ePub
• enable export of SVG's for re-use (place into word processing documents, graphics documents, etc.)
• store ePub (with a .zip containing the nbk) in a dedicated notebook catalog in calibre with cover and metadata, including the GUID identifying the notebook.
• enable export of pen annotations made to books and personal documents to ePub, but without any versioning. Optionally store in Notebooks catalog.

Known limitations

• It's not possible to determine the Notebook Title or Folder name. This information is not available in the nbk file or in user storage.
• Pen annotations will be able to reference Title/Author of the corresponding book or document. And while it is technically possible to attach a location (it's found in the annotations data file in the associated .sdr folder), KFX Input is (by design) only concerned with basic conversion tasks. (At some point, I'll figure out how to add raw location data as an additional step, just to provide some additional context.)

Assuming I am able to create a minimal viable product, the obvious next step would be to create a calibre plugin. But first things first.

I welcome feedback and suggestions. I hope to have some code to share in a week or two.

[thereisthat]

Hi, I am interested on this and can help out if help is needed. My main use case would be offline view of the notebooks on my PC in high resolution.

[mergen3107]

QUOTE:
“It's not possible to determine the Notebook Title or Folder name. This information is not available in the nbk file or in user storage.”

How about /var/local/ databases? I have a feeling they might be in cc.db with proper notebook identifiers.

[Fitz Frobozz]

Great project idea. Is this still in the works? How can other folks be of help?

[Retired_Trout]

I'm also interested in this project.

[tomsem]

Thanks for resurrecting this thread.

In addition to being busy and distracted with a lot of other things, I had been stuck just getting python code to copy things off of Scribe, given that it changed to use MTP protocol.

There is a PyMTP package, but it has not been updated for Python 3, it didn't work out of the box and I just left it at that.

But it turns out that updating turned out to be trivial: convert with 2to3, and change 2 other lines of code, and the example code (also 2to3'ed) all works!

It's just a single python file with only standard library calls so as long as it is able to find the libmtp C library on the system, I think it can just be dropped in and used without pip installing anything.

I am planning to check it against latest libmtp C library to see if there are any filetypes or error codes to add (probably not), add type hinting to the code, convert byte strings to str, and it should be good to go for general use.

[Fitz Frobozz]

Hello! Just a shout to mention that I'm still following this thread with interest!