[Plugin] ePub3-itizer - epub3 output plugin for Sigil

[roger64]

Hi

I tried the 0.56 version on Linux which works remarkably smoothly and quickly with a commercial ePub2 which had even some epubcheck mistakes. Thank you.

A question

This ebook has about 200/250 endnotes. I had hoped that an

epub:type="rearnote"

attribute (or other of the same kind like "note", "footnote", etc.) could have been added to each note.

This kind of attribute is requested with Koreader (and maybe other rendering engines) to display inline notes (notes de bas de page) which provide a real comfort at least for some books*. Note that a Koreader option allows the user to select either a pop-up or inline display as long as this kind of attribute is present. Koreader can be used with many different ebook-readers (Kobo, Kindle, Android, pocketbook, etc.)

If the endnotes definitions are to be found in one file, this seems to be doable. I could do it myself with a regex but maybe not every reader.

* for example when the notes add complementary information which goes beyond plain bibliographical references and when pretty small anchors can easily be missed by the reader.

[KevinH]

Since aside comes with epub3, if only anchors were there, how could it recognize that it was meant as a footnote and not just another anchor a tag.

Adding semantic tags like that could only be added if it was clear it was a link to a footnote and a return link from a footnote.

[roger64]

Originally Posted by KevinH

Since aside comes with epub3, if only anchors were there, how could it recognize that it was meant as a footnote and not just another anchor a tag.

Adding semantic tags like that could only be added if it was clear it was a link to a footnote and a return link from a footnote.

Indeed.

As it happens, this particular commercial book includes also index entries using a a tag and href link. So this semantic question has to be answered by the reader (if the editor did nothing about it).

But, maybe, if you find something like this in the opf of the proposed ePub2 :

<guide>
    <reference type="cover" title="Cover" href="text/titlepage.xhtml"/>
    <reference type="other.footnotes" title="Notes de bas-de-page" href="text/PL20.xhtml"/>
  </guide>

you could process this one and only PL20 chapter?

[KevinH]

The use of the guide to properly mark semantic types as footnotes is rare to say the least. It was not even one if the original semantic types for epub2.

So this is really something for a different plugin, not ePub3-itizer to deal with.

[roger64]

Originally Posted by KevinH

Adding semantic tags like that could only be added if it was clear it was a link to a footnote and a return link from a footnote.

Making use of aside and epub:type to enable pop-up or inline footnotes is part of the advantages of the ePub3 format and probably one of the reasons why users would like to upgrade.

That's why, even if it is a rare occurrence, if the acknowledged semantic tag "footnote" is present in the opf, I fail to see why it would be best to call for a distinct plugin for upgrading also the footnotes to the ePub3 standard. It would be a puzzling, artificial and easily avoidable complexity.

As you said above, an if... then... condition "could only be added" to the plugin code to deal with the footnotes when a semantic tag is present.

[KevinH]

It should be part of a more generic footnote plugin where the user can optionally identify a particular class if used to help detect foot vs end notes and and source vs return.

So no this is not something for a generic epub3-itizer to do. It is just the first step to making a good epub3, and not the final step. Much in the way of proper semantics should be set by the dev and something like AccessAide should be run to update alts, missing titles, add aria roles, etc.

So no, this is not going to become part of epub3-itizer when it really belongs in its own plugin that deals with identifying and updating footnotes.

[roger64]

Thank you for taking the time to reply me.

You are right, the plugin already exists with Sigil... Its name is FNLinker. It creates new ePub3 footnotes under two conditions:
- an easy one, the presence of the semantic tag (see above)
- a tricky one (depends on the book): there must be an equal number of class anchors and class definitions.

If there is a discrepancy between these two classes, it may require a careful (long) analysis to correct the mistake(s) before being able to process the links in some few seconds.

[AlanHK]

If I process an epub with filenames *.html this produces a valid epub3.
However, if I do an epubcheck, it gives a warning like:

WARNING(HTM-014a): XHTML Content Document file name 'OEBPS/Text/r1.html' should have the extension '.xhtml'.

I can fix that easily, but can the plugin do this automatically?

I realise there is the possible case of there being both a.html and a.xhtml, though no one in their right mind would do that, so in that case you could either add a suffix or abort renaming with an error message.

[Doitsu]

Originally Posted by AlanHK

If I process an epub with filenames *.html this produces a valid epub3.
However, if I do an epubcheck, it gives a warning like:

You'll have to rename the files before running the plugin:

Originally Posted by KevinH

Before Conversion
1. Make sure your epub2 validates under FlightCrew and/or EpubCheck
2. If your epub uses ".htm" or ".html" file extensions, you should first change them to use ".xhtml" file extensions as follows [...]

[KevinH]

Yes, Doitsu is 100% correct. It is just easier and faster to do the file renames and link updating inside Sigil and it can do it all at once easily.

[DNSB]

Originally Posted by AlanHK

I realise there is the possible case of there being both a.html and a.xhtml, though no one in their right mind would do that, so in that case you could either add a suffix or abort renaming with an error message.

I've run into two epubs where there were duplicate filenames with different extensions (i.e. part00019.htm and part00019.html). Both of them were omnibus editions. As you said, not sure why anyone in their right mind would do that but it has happened.

[KevinH]

I hope the were not both in the same folder at least!

[AlanHK]

Originally Posted by Doitsu

You'll have to rename the files before running the plugin:

What do you mean "have to"? If I'm doing it manually, can do it just as easily after conversion. Though it's probably preferable, even though not necessary.

Anyway, the point is: I have a epub2, using .html filenames.
ePubCheck: "No problems found!"

Run the conversion:
'WARNING(HTM-014a): XHTML Content Document file name 'OEBPS/Text/halftitle.html' should have the extension '.xhtml'."
-- for every file.

Would be nice if epub3-itizer fixed that.

[AlanHK]

Originally Posted by DNSB

I've run into two epubs where there were duplicate filenames with different extensions (i.e. part00019.htm and part00019.html). Both of them were omnibus editions. As you said, not sure why anyone in their right mind would do that but it has happened.

Actually, if you look at books, where the original structure was each volume in its own folder in the zip, after Sigil opened it and flattened the file structure it renamed any collisions automatically, (e.g. several cover.jpg files each gets a number added). So not a new problem.

[KevinH]

No. Naming your files properly is not the job of ePub3-itizer as the plugin instructions and this thread clearly states.

It will not be changed, especially when doing it inside Sigil would be much faster than doing it via python in a plugin. No reason at all to duplicate that code.