[GUI Plugin] Modify ePub

[jhowell]

Originally Posted by jberanek

Is this a known issue, do any of the vast array of options in the plugin possibly fix this?

Known issue due to a change by Amazon. Will be fixed (as much as possible) in the next calibre release.

https://github.com/kovidgoyal/calibr...d2fa13fc1ae29e

[jberanek]

Originally Posted by jhowell

Known issue due to a change by Amazon. Will be fixed (as much as possible) in the next calibre release.

https://github.com/kovidgoyal/calibr...d2fa13fc1ae29e

Thanks for pointing me in the right direction. What an astonishing move by Amazon - it does look at face value like a "nerfing" of book sideloading.

(Just noticed that the "Send to Kindle" app doesn't even support Epub uploads)

[JSWolf]

Originally Posted by jberanek

Thanks for pointing me in the right direction. What an astonishing move by Amazon - it does look at face value like a "nerfing" of book sideloading.

(Just noticed that the "Send to Kindle" app doesn't even support Epub uploads)

Do you have the latest Send to Kindle?

[jberanek]

Originally Posted by JSWolf

Do you have the latest Send to Kindle?

Yes, I'm sure I used to use it to send books now and then, so perhaps they removed the ability to send Epubs recently?

[jhowell]

Originally Posted by jberanek

(Just noticed that the "Send to Kindle" app doesn't even support Epub uploads)

The Send-to-Kindle apps for PC and Mac have not yet been updated to handle EPUB format. Should be coming "soon".

[Pjama]

I am using the latest version of Calibre, and just installed the modify epub plug in, and am still getting the garbled signs instead of accented letters on some of my French books. Am I doing something wrong, or do I just need to wait for the next version of Calibre?

UPDATED TO ADD: It works perfectly! I was the problem, I didn't see the control panel for the modify ePub plug-in, and when I selected change to UTF and saved the book, it then sent to the kindle perfectly. (Albeit, missing the author info, which I have previously added successfully in such cases by including the author in my email subject line - I'll try resending).

[kiwidude]

Release Notes:
https://github.com/kiwidude68/calibr...fy_epub-v1.8.0

[killo3967]

Hi.
First, thanks for such a good plugin.

My problem is:

1.-I select "Add/Replace metadata Jacket", this function correctly and i could see the new jacket in my book.

2.- I check the book with the "Checkepub" in Sigil and appears
this error:

Col: 87: ERROR(RSC-005): Error al analizar el archivo: el atributo "data-calibre-jacket-searchable-tags" no se permite aquí; se esperaba atributo "align", "char", "charoff", "class", "dir", "id", "lang", "style", "title", "valign" o "xml:lang"

[kiwidude]

@killo3937 - As I only speak english I have no idea what Sigil is trying to warn you about. I will say the base functionality of this plugin has not been changed in a very long time. There is nothing I can do with that without (a) having a copy of your epub that has the jacket stilll in it and (b) knowing exactly which version of Sigil you tested it with.

So long as the epub displays in calibre viewer/still converts ok to other formats if that is what you use in your device then I simply would not worry about what that Sigil check is telling you. Things like EpubCheck can generate a lot of "noise" about things which simply do not actually matter - just chill and go read the book

[JSWolf]

Originally Posted by kiwidude

@killo3937 - As I only speak english I have no idea what Sigil is trying to warn you about. I will say the base functionality of this plugin has not been changed in a very long time. There is nothing I can do with that without (a) having a copy of your epub that has the jacket stilll in it and (b) knowing exactly which version of Sigil you tested it with.

So long as the epub displays in calibre viewer/still converts ok to other formats if that is what you use in your device then I simply would not worry about what that Sigil check is telling you. Things like EpubCheck can generate a lot of "noise" about things which simply do not actually matter - just chill and go read the book

Here is the translation to English.

Col: 87: ERROR(RSC-005): Error parsing file: attribute "data-calibre-jacket-searchable-tags" is not allowed here; expect attribute "align", "char", "charoff", "class", "dir", "id", "lang", "style", "title", "valign" or "xml:lang"

[JSWolf]

This is the line that has the error.

<tr data-calibre-jacket-searchable-tags="1" style="color:white; display:none">

Remove data-calibre-jacket-searchable-tags="1" and the error goes away.

This error is only an error with ePub 2. It verifies with ePub 3. But (IMHO) that bit of code is not needed and if it goes away, the jacket is good on both ePub 2 & 3.

One other issue. If you do not select to put the jacket at the back, it's put at the front where the cover should be. So that chages the cover since the cover should be first.

[Doitsu]

Originally Posted by killo3967

Col: 87: ERROR(RSC-005): Error al analizar el archivo: el atributo "data-calibre-jacket-searchable-tags" no se permite aquí; se esperaba atributo "align", "char", "charoff", "class", "dir", "id", "lang", "style", "title", "valign" o "xml:lang"

data-* attributes can only be be used in epub3 books. Most likely you've added a jacket to an epub2 book.
As kiwidude has already pointed out, as an end user, you don't have to worry about the error message.

@kiwidude the attribute comes from from calibre/resources/jacket/template.xhtml. Theoretically, Kovid should have created two templates: one for epub2 books and one for epub3 books.
However, since Kovid has made it abundantly clear that he's not inclined to change the Calibre code just to accommodate EPUBCheck, it's highly unlikely that this problem will get fixed.

[kiwidude]

@JSWolf, @Doitsu - thanks very much for the translation/confirmation on this not being something to be concerned about as far as this plugin is concerned.

Originally Posted by JSWolf

One other issue. If you do not select to put the jacket at the back, it's put at the front where the cover should be. So that chages the cover since the cover should be first.

There is some very crude logic I see in the plugin whereby it does try to insert what "it" considered to be the cover. It is looking for an opf spine item with id='titlepage' which if found will insert after that.

Is there a more reliable/alternative/additional way of identifying the cover from the spine? If there is a quick win on this I am happy to make a change, whereas if this requires heading back into the weeds on epub structures and files then it is not something I have time to tackle at the moment.

[JSWolf]

Originally Posted by kiwidude

@JSWolf, @Doitsu - thanks very much for the translation/confirmation on this not being something to be concerned about as far as this plugin is concerned.


There is some very crude logic I see in the plugin whereby it does try to insert what "it" considered to be the cover. It is looking for an opf spine item with id='titlepage' which if found will insert after that.

Is there a more reliable/alternative/additional way of identifying the cover from the spine? If there is a quick win on this I am happy to make a change, whereas if this requires heading back into the weeds on epub structures and files then it is not something I have time to tackle at the moment.

You could keep the logic and if it fail to find id="titlepage", place the jacket as the third file. In most cases, you have the cover followed by the title page. In some cases, it may not be optimal, but it's better then were it's landing now.

[kiwidude]

Release Notes:
https://github.com/kiwidude68/calibr...fy_epub-v1.8.1

@JSWolf - defaulting to 3 pages in isn't something I could agree with - found too many book examples where that makes no sense. The problem is there is no guarantee about what convention publishers choose for naming their covers/title pages for the plugin to identify it. Some give it the same name as the book, rather than something nice like any of "cover", "cvi" , "tp", "titlepage", which I am now detecting.

I have however made a change so it defaults to being the second page in and then keep incrementing that number until it finds a page it "guesses" is not a cover or a titlepage, based on those names above. There are books out there though (if they havent had a calibre conversion) where this still won't be "ideal" - e.g. a book that has no cover page at the start but instead 3 pages of quotes about how great the book is before a title page would now have this jacket inserted between quote pages 1 and 2.

Personally I think the jackets got hit too hard with the ugly stick and I never use them so it is not something I am going to do extensive testing on. But if someone can please try the release at the link above and give me feedback, otherwise I will just push it out to everyone in a few days...

EDIT: I have now officially released this version.