[Conversion Plugin] KFX Input

[Blaineoreski]

@jhowell

Just a note of gratitude for everything you've done, especially in this thread and all your kind help across the site. Thank you, jhowell

Sincerely,

Ski

[jhowell]

Originally Posted by Blaineoreski

@jhowell

Just a note of gratitude for everything you've done, especially in this thread and all your kind help across the site. Thank you, jhowell

Sincerely,

Ski

I glad that you find my efforts here to be useful.

[Topia]

I found a book that I can't convert with this plugin.

Exception content:

Exception: rb element not a child of ruby element: b'<ruby><span><rb-bad>珠</rb-bad><img src="image_rsrcCED.jpg" alt="" style="-kfx-render: inline; -kfx-style-name: s5FU; box-sizing: content-box; height: 1em; max-height: 100%; vertical-align: baseline; width: 1em"/></span></ruby>'

I analyzed this problem by myself and concluded this check is not required (at least in my case).

This book uses the image for non-standard characters, so

<ruby><rb>珠</rb><rt>ruby1</rt><rb><img ... /></rb><rt>ruby2</rt></ruby>

is desired.
I slightly modified the code to the below and confirmed it works in my case:

                        while rb_elem.getparent() is not event_elem:
                            if len(rb_elem.getparent()) != 1:
                                # rb_elem.tag = "rb-bad"
                                # raise Exception("rb element not a child of ruby element: %s" % etree.tostring(event_elem))
                                break  # CHANGED to break

                            rb_elem = rb_elem.getparent()

                    rb_elem = self.replace_element_with_container(rb_elem, "rb")
                    rb_parent = rb_elem.getparent()  # ADD

                    ruby_id = ruby_id_entry.pop("$758")
                    ruby_content = self.get_ruby_content(ruby_name, ruby_id)

                    if ruby_content.pop("$159") == "$269":
                        rt_elem = etree.Element("rt")

                        rb_parent.insert(rb_parent.index(rb_elem) + 1, rt_elem)  # MODIFIED to rb_parent

[jhowell]

Originally Posted by Topia

I found a book that I can't convert with this plugin.

Thanks for the bug report and solution.

I would like to investigate this further to see what other cases I might have missed. Would you mind sharing the ASIN of a book that exhibits this problem?

[tomsem]

With the latest Scribe update (6.16.6), Scribe now lets you set time/date automatically, just like every other internet connected device since forever.

I have in past experimented with updating/replacing notebooks, and before overwriting I had to change the modify date of the nbk file by +12h or so before overwriting the target nbk, or the change would be ignored and the previous version synced back down from cloud. This is no longer necessary, I can just copy to computer and back to the target folder, and it 'takes'.

I also thought that I had converted Print Replica document pen annotations into notebooks by overwriting a notebook nbk. But when I try it, all I see is the letter 'P' in upper left corner of the (and the notebook no longer behaves correctly).

I am probably misremembering this. Probably we've even discussed this previously.

So let me try a different question: is it feasible to convert an 'annotation' nbk into a 'notebook' nbk?

Use case is that one is using personal document as a notebook (some template or they want to be able to zoom/pan, or use collections instead of folders) and they want to send it for handwriting conversion.

As things stand you have to copy and paste using lasso tool one page at a time.

[jhowell]

Originally Posted by tomsem

With the latest Scribe update (6.16.6), Scribe now lets you set time/date automatically, just like every other internet connected device since forever.

I have in past experimented with updating/replacing notebooks, and before overwriting I had to change the modify date of the nbk file by +12h or so before overwriting the target nbk, or the change would be ignored and the previous version synced back down from cloud. This is no longer necessary, I can just copy to computer and back to the target folder, and it 'takes'.

I am surprised that replacing a notebook works at all. I wonder whether it was the time setting change or something else in the new firmware that causes the new behavior.

Originally Posted by tomsem

I also thought that I had converted Print Replica document pen annotations into notebooks by overwriting a notebook nbk. But when I try it, all I see is the letter 'P' in upper left corner of the (and the notebook no longer behaves correctly).

I am probably misremembering this. Probably we've even discussed this previously.

We discussed this back in May of last year. At that time you reported the same failure but with a different letter appearing. My response is here.

Originally Posted by tomsem

So let me try a different question: is it feasible to convert an 'annotation' nbk into a 'notebook' nbk?

It may be feasible in principle but would require further research into what is going wrong and likely an ability to modify notebooks using outside software to correct the problem. That isn't something I intend to invest my time doing.

[tomsem]

Originally Posted by jhowell

I am surprised that replacing a notebook works at all. I wonder whether it was the time setting change or something else in the new firmware that causes the new behavior.



We discussed this back in May of last year. At that time you reported the same failure but with a different letter appearing. My response is here.



It may be feasible in principle but would require further research into what is going wrong and likely an ability to modify notebooks using outside software to correct the problem. That isn't something I intend to invest my time doing.

Thanks for referencing previous posts and confirming my suspicions about misremembering things.

I will be happy to investigate and try to figure out what is involved.

Previous to the update, Kindle ran on GMT time. A file copied from computer (I'm in Pacific Time) would somehow appear 'in the past' with respect to the date Kindle last had for the file (in some 'sync' database, I infer). Scribe would infer that the cloud version was newer, and ignore the changed notebook and replace it with the earlier version on sync.

Changing the modify date time to some value greater than the current GMT time (9 ish hours for me) flagged the notebook as 'changed' and would then accept and sync it.

With the update, replacing the file generates a modify date that is greater than the last sync time, so it gets accepted as changed.

[jhowell]

Version 2.8.0 - 25 Jan 2024

Fix conversion of Print Replica books when using calibre versions older than 6.0. (Fixes "AttributeError: 'NoneType' object has no attribute 'PdfReader'")

Fix conversion of books with ruby elements containing images. (Fixes "Exception: rb element not a child of ruby element")

[vikas911]

jhowell Thank you for this tool it worked great! Question for you, does this also work with the current kindle app on Mac? i.e. version 7.x? I couldn't get it to work using the current 10.0.9 DeDRM and current version of your tool.

It works well with the Kindle Classic App however.

Thank you.

[jhowell]

Originally Posted by vikas911

jhowell Thank you for this tool it worked great! Question for you, does this also work with the current kindle app on Mac? i.e. version 7.x? I couldn't get it to work using the current 10.0.9 DeDRM and current version of your tool.

As far as I know the DeDRM plugin has not been updated to support the new Kindle for Mac app. You can raise an issue at noDRM's GitHub but I don't know if that will lead to a solution.

[jhowell]

Version 2.8.1 - 29 Jan 2024

Fix plugin not working using calibre versions older than 6.0.

[Topia]

Originally Posted by jhowell

I would like to investigate this further to see what other cases I might have missed. Would you mind sharing the ASIN of a book that exhibits this problem?

I haven't noticed it, sorry. Yes, I can share the ASIN, B0CLXPG5FN. but it's a Japanese Kindle book. maybe it's a bit challenging to find the affected position.

edit: I'll try the latest version a bit later.

[jhowell]

Originally Posted by Topia

I haven't noticed it, sorry. Yes, I can share the ASIN, B0CLXPG5FN. but it's a Japanese Kindle book. maybe it's a bit challenging to find the affected position.

I would test using that book myself but it is not available from Amazon US.

Originally Posted by Topia

edit: I'll try the latest version a bit later.

Hopefully the latest plugin release solves the problem. Please let me know if not.

[Cyn]

Hi

I am trying to perform a conversion, but I am encountering the following error: ERROR: KFX conversion failed: error('invalid group reference 1 at position 3') calibre, version 7.5.1. Does anyone know how to fix it?

[jhowell]

Originally Posted by Cyn

I am trying to perform a conversion, but I am encountering the following error: ERROR: KFX conversion failed: error('invalid group reference 1 at position 3') calibre, version 7.5.1. Does anyone know how to fix it?

That is a new error. I am going to need more information in order to diagnose the problem.

See the last paragraph in the "Dealing with conversion errors" topic in the first post of this thread for information on how to obtain a conversion log. If you could post that log here it would be very helpful.