[GUI Plugin] EpubMerge

[JimmXinu]

The error suggests that there is broken XML in the content.opf of that particular epub--probably the first one in the merge.

You might be able to clear the problem by opening the epub in Calibre's Edit book, make some minor change (add a space to a file, then delete) and save it again. Or Edit book may point out the problem.

You could also try an epub validator.

Without seeing the epub myself, that's about all I can suggest.

[dmx87212]

I appreciate the recommendation! I ran it through an epub validator and it showed that several of the books had errors. Since I don't have much understanding of editing ebooks, would you be able to break down an example of one of the issues that the epub validator stated? If not I understand as this isn't the right location.

Below is an example of two of the errors displayed as well as the relevant code. One of the books only had 2 errors in this case.

ERROR(RSC-005): ./books/Bo/databases/eBookIt/temp_uploads/1646250078.epub/content.opf(46,140): Error while parsing file: value of attribute "id" is invalid; must be an XML name without colons

code:

<item id="a26-chapter-19--order-(redux).html" href="OEBPS/text/a26-chapter-19--order-(redux).html" media-type="application/xhtml+xml"/>

and then

ERROR(RSC-005): ./books/Bo/databases/eBookIt/temp_uploads/1646250501.epub/content.opf(151,58): Error while parsing file: value of attribute "idref" is invalid; must be an XML name without colons


code:

<itemref idref="a26-chapter-19--order-(redux).html"/>

[JimmXinu]

Originally Posted by dmx87212

ERROR(RSC-005): ./books/Bo/databases/eBookIt/temp_uploads/1646250078.epub/content.opf(46,140): Error while parsing file: value of attribute "id" is invalid; must be an XML name without colons

code:

<item id="a26-chapter-19--order-(redux).html" href="OEBPS/text/a26-chapter-19--order-(redux).html" media-type="application/xhtml+xml"/>

It doesn't like the ()s in <item id="a26-chapter-19--order-(redux).html"...> or <itemref idref="a26-chapter-19--order-(redux).html"/>. I suspected it was that or the '.', so I tested; Calibre's Edit book Check feature gives a bit more detail:

Warning [2 / 2]
content.opf (line: 43)
The id file0001( is not a valid id. IDs must start with a letter ([A-Za-z]) and may be followed by any number of letters, digits ([0-9]), hyphens ("-"), underscores ("_"), colons (":"), and periods ("."). This is to ensure maximum compatibility with a wide range of devices.

Note that <item id="a"...> and <itemref idref="a"/> need to match up--it's itemref because it's referring to that <item>

[JimmXinu]

New Test Version

- Add check for multiple toc.ncx files, use last to match Calibre viewer.

Somebody PM'd me about a problem that turned out to be an input file with two application/x-dtbncx+xml (toc.ncx) files. Which epubmerge doesn't like.

There is no good reason to have two, but while the standard refers to "The NCX file (Navigation Control file for XML), traditionally named toc.ncx" (singular), I don't find anything formally forbidding it and epubcheck (in Calibre's Edit book at least) doesn't complain about it.

Calibre's ebook viewer appears to use the last listed toc.ncx file. Removing the last one causes it to simply use the first. Oddly, doing Calibre convert epub->epub still keeps both. I would have expected it to remove one.

If anybody knows a better way to handle this issue, I'd be curious to hear it.

[JSWolf]

Originally Posted by JimmXinu

New Test Version

- Add check for multiple toc.ncx files, use last to match Calibre viewer.

Somebody PM'd me about a problem that turned out to be an input file with two application/x-dtbncx+xml (toc.ncx) files. Which epubmerge doesn't like.

There is no good reason to have two, but while the standard refers to "The NCX file (Navigation Control file for XML), traditionally named toc.ncx" (singular), I don't find anything formally forbidding it and epubcheck (in Calibre's Edit book at least) doesn't complain about it.

Calibre's ebook viewer appears to use the last listed toc.ncx file. Removing the last one causes it to simply use the first. Oddly, doing Calibre convert epub->epub still keeps both. I would have expected it to remove one.

If anybody knows a better way to handle this issue, I'd be curious to hear it.

It's a bug in epubchceck and a problem with that eBook. No ePub reading software that I know of handles two NCX files.

[JimmXinu]

New Version Released

Version 2.13.0 - 04 Apr 2022
- Add check for multiple toc.ncx files, use last to match Calibre viewer.
- Update Translations

[criscros]

I use FanFicFare for several webnovels and it works great. Thank you for that.
But I have one webnovel that I have to do manually. When I use epubmerge, I end up with 3 epubs, the original, the section to be added and a new epub.
Is there a way to update like FanFicFare does, so only the original is updated and I don't lose my reading positions?

[JimmXinu]

Not with EpubMerge. It's pretty tightly focused on merging, not updating/editing.

I do have a few epubs I manually update, and I use Calibre Edit book feature to do it. It does involve several steps, adding a file, putting in the content, updating the TOC, etc.

[dmx87212]

I have been using EpubMerge for years with Calibre in order to merge large book series and than easily navigate them using the generated Table of Contents on an eReader. The eReader that I have been using is Marvin 3.

Lately, I have been having an issue where when I merge multiple books, I am only receiving the table of contents of one book, even though I have the settings enabled to "Insert Table of Contents entry for each title" and "copy Table of Content entries from each title".

To give a relative time frame; I had created a merge of 10 books in Nov 2021 and it displayed each of the titles+table of contents when opened in Marvin 3. Now, I am merging an additional book to it and suddenly it doesn't work.

[JimmXinu]

I don't see anything wrong with my tests. Are you sure you're seeing nested TOC entries?

I need to see the books you're working with to know more, really.

[JimmXinu]

New Test Version Posted

2022-07-19
- Use cal6 icon theme system to allow plugin icon customization.

EpubMerge has two custom icons, but it uses several of the 'standard' Calibre icons. Now it should use installed icon themes and icon themes can include a custom icon for EpubMerge specifically (EpubMerge/images/icon.png & EpubMerge/images/unmerge.png). Only for Calibre 6+.

[thiago.eec]

Originally Posted by JimmXinu

New Test Version Posted

2022-07-19
- Use cal6 icon theme system to allow plugin icon customization.

EpubMerge has two custom icons, but it uses several of the 'standard' Calibre icons. Now it should use installed icon themes and icon themes can include a custom icon for EpubMerge specifically (EpubMerge/images/icon.png & EpubMerge/images/unmerge.png). Only for Calibre 6+.

I use the Highlights-dark icon theme. It has icons for both EpubMerge and EpubSplit, but they are not being used here with the test version.

I tried extracting the theme folder for EpubMerge on ../resources/images/EpubMerge/ and it also didn't work here.

I'm using calibre 6.1 on Windows 10 x64.

EDIT:
Just noticed that the plugin is expecting the icons to be on ../resources/images/EpubMerge/images/ (and it works if I do this). Well, that's not what all the themes I've seen are doing. They put all the icons on ../resources/images/EpubMerge/.

[JimmXinu]

Theme contains images/EpubMerge/icon.png , plugins needs it to be EpubMerge/images/icon.png, etc.

All of my kiwidude descended plugins use images like that and I'd rather not change it for backward compatibility reasons.

[thiago.eec]

Originally Posted by JimmXinu

Theme contains images/EpubMerge/icon.png , plugins needs it to be EpubMerge/images/icon.png, etc.

All of my kiwidude descended plugins use images like that and I'd rather not change it for backward compatibility reasons.

That's odd. The latest released versions of EpubMerge (2.13.0) and EpubSplit (3.3.0) works fine with the theme (under calibre 5). To be honest, all other plugins I have installed use ../resources/images/Plugin_Name/*.png.

[theducks]

There is another reason that Jimm's idea is best:
The icons IN the plugin are default and the fallback.
whereas the Resources are User supplied