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

[KevinH]

Yes recursive named entity expansion has been used to attack html sites. Since epubs use html5 for epub3 and since the same problems exist with webkits used in ebook viewers, it was probably a good idea. That said, I have no idea if there are malicious epubs in the wild but given there are malicious Word docs and PDFs in the wild, it is just a matter of time.

[DiapDealer]

Originally Posted by AlanHK

As more entities are numeric, it becomes a stew of digits and impossible to make sense of by eye.

In my opinion, a slew of named entities make html equally impossible to make sense of. so I don't bother trying to preserve them at all. With the exception of the non-breaking space entity, I see very little compelling reason to use entities (except the standard required ones for (&, <, >) at all--named or numeric. The actual unicode characters work just fine and read much better in code, in my opinion. Why use a 4 or 5 character code to represent a single character?

[AlanHK]

Originally Posted by DiapDealer

The actual unicode characters work just fine and read much better in code, in my opinion. Why use a 4 or 5 character code to represent a single character?

The half dozen space characters all look the same when expressed as characters.

But I wasn't talking about that, but the requirement in epub3/HTML5 to replace named entities, &emsp; in particular, with a numeric ones.
All must become cryptic numbers instead of mnemonics if you want it to pass muster.

[AlanHK]

Different issue:
I've been making ePubs for iBooks.

While ePub2 worked, the TOC had glitches and omissions. Seems it really wanted ePub3. So I now use the plugin to make those.

Also, these ePubs are required to use embedded fonts.
While they worked in other ePub viewers, they did not appear in iBooks.
Eventually found that iBooks required some special Apple code to allow them.

As far as I can see the rules are:
for ePub2: include the file com.apple.ibooks.display-options.xml

Spoiler:

<?xml version="1.0" encoding="UTF-8"?>
<display_options>
<platform name="*">
<option name="specified-fonts">true</option>
</platform>
</display_options>

For ePub3 instead you should have this line in the opf:
<meta property="ibooks:specified-fonts">true</meta>
which requires a namespace added to the package:
<package version="3.0" xmlns="http://www.idpf.org/2007/opf" unique-identifier="bookid" prefix="ibooks: http://vocabulary.itunes.apple.com/rdf/ibooks/vocabulary-extensions-1.0/">

Since I have been creating the book in ePub2 and just converting to ePub3 at the end, I tried to insert these lines in the ePub2. They threw an error with ePubcheck, but I thought I could ignore that.
So then I converted it to ePub3, expecting the code to just be copied, and there be valid. Instead the namespace was omitted and the property had no value:

<package version="3.0" unique-identifier="BookId" xmlns="http://www.idpf.org/2007/opf" prefix="rendition: http://www.idpf.org/vocab/rendition/#">
    <meta property="ibooks:specified-fonts" />

I understand this behaviour is justified, since the OPF was apparently invalid as an epub2.
Also I know I could add the lines after converting to ePub3. But I would prefer to have them in my original ePub2 so I don't have to remember to add the code every time after exporting to ePub3 before submitting it.
Is there a way to have such code "pass through" to the ePub3 without further intervention?

I found what seems to be a workaround: the ePub2 method of including the XML seems to work with ePub3, despite it not being officially sanctioned. So I have been doing that since the XML does survive the ePub3-itizer; but I am uneasy that iBooks could one day be stricter and not recognise it.

[KevinH]

I really have no plans to support epub3 prefix and new style meta tags in an epub2 opf as Sigil itself does not support them (and the opf must be machine readable and grok-able at all times inside Sigil to allow things like automatic splitting, file renaming, etc).

The simplest solution is to open the epub3 after conversion in Sigil and manually add them using a clip or search and replace, or whatever is easier.

Or ask someone to make an epub3 compatible iBooks plugin adapter and run it in Sigil after conversion.

Or hard code your own copy of the epub3itizer to handle your special case for ibooks.

[odamizu]

Thank you to everyone who has worked on Sigil. The more time I spend with it the more impressed and grateful I am for this wonderful program.

I have been slowly figuring out ePub3 (yes, slow learner here ) and finally feel comfortable adopting it with all my ebooks going forward.

What I'm trying to decide now is: Should I use this nifty ePub3-itizer plug-in to go back through my entire ebook library and convert all the ePub2s into ePub3? What are the advantages to doing this? Any disadvantages (other than the considerable time it will take)? Do you foresee a time when ePub2 will no longer be readable?

Muchas gracias for any thoughts

[KevinH]

Given that most/all epub3 readers can read epub2, I do not think you need worry about epub2s becoming unreadable for a very long time.

For most typical, mainly-text based e-books, the only real advantages of epub3 are:

1. enhanced semantics/nav/audio to allow those with disabilities to use automated reading software to more easily read and understand the e-book or listen to any to synchronized audio tracks. But this requires work on the part of authors to annotate the text with semantics tags throughout of add support for audio text synching.

2. better or more complete CSS 3 support

3. audio and video support is claimed but spotty in actual readers especially for video formats

4. support for javascript to possibly "enhance" an e-book (but again spotty support in readers)

5. support for mathml but again this can be spotty.

6. more standardized fixed layout formats but possibly incompatible with Kindle formats

etc.

So a simple conversion from epub2 to epub3 without further work to improve the e-book really doesn't buy you much if anything. It does gain you access to the latest tools such as the latest epubcheck, and possibly the latest kindlegen versions but ...

That said, creating new epubs in epub3 and keeping the toc, should allow for backwards compatibility with epub2 readers while allowing the use of more advanced features if you so desire.

This is all my two cents fwiw.

[odamizu]

Originally Posted by KevinH

Given that most/all epub3 readers can read epub2, I do not think you need worry about epub2s becoming unreadable for a very long time.

For most typical, mainly-text based e-books, the only real advantages of epub3 are ...

So a simple conversion from epub2 to epub3 without further work to improve the e-book really doesn't buy you much if anything. It does gain you access to the latest tools such as the latest epubcheck, and possibly the latest kindlegen versions but ...

Okay, thank you! I will embrace ePub3 going forward and not fret about my existing ePub2s

[Julie34]

Thank you for your advice Kevin. However, I’ll tamper it a bit: most/all epub3 readers will be able to read epub2 as long as there is a sufficient number of commercial epub2 created.
Reading machines and some (the commercials one) readers (applications) follow the market. And the market prospers with increasing the wish/need to consume new toys. I observe the beginning of interactive ebooks (for children mostly). The new toy (for ebooks), will it be epub3? or something else, more proprietary? But you’re right, there is time, there is investments already done (just listening now at the radio how beautiful are those children reading “real” books and not only electronic devices) …

[KevinH]

Agreed but given 3rd party conversion tools exist and the lack of reader support for many of these features (for example: no javascript, no good video support, no real semantics, and no mathml in Kindle format books which are made from epubs) I do not see a huge push to remove support for epub2 epubs, accept for small niche product markets such as children's enhanced books.

I would eventually convert my epub2 ebooks to epub3 slowly (and maybe to epub3.1 eventually - the latest spec that no-one supports) but there really us no rush imho.

Originally Posted by Julie34

Thank you for your advice Kevin. However, I’ll tamper it a bit: most/all epub3 readers will be able to read epub2 as long as there is a sufficient number of commercial epub2 created.
Reading machines and some (the commercials one) readers (applications) follow the market. And the market prospers with increasing the wish/need to consume new toys. I observe the beginning of interactive ebooks (for children mostly). The new toy (for ebooks), will it be epub3? or something else, more proprietary? But you’re right, there is time, there is investments already done (just listening now at the radio how beautiful are those children reading “real” books and not only electronic devices) …

[sghughes]

I'm new to Sigil. Am unable to install the ePub3-itizer on my MacBook pro (high sierra 10.13.1). The files are grayed out when using the plug-in manager. Any suggestions?

[Doitsu]

Originally Posted by sghughes

I'm new to Sigil. Am unable to install the ePub3-itizer on my MacBook pro (high sierra 10.13.1). The files are grayed out when using the plug-in manager. Any suggestions?

Please read the Mac Users with Plugin Issues sticky.

[sghughes]

Originally Posted by Doitsu

Please read the Mac Users with Plugin Issues sticky.

Thanks - I had already installed the latest version of Active TCL but this had no effect on the problem.

[DiapDealer]

You (or something) have unzipped the plugin. The only type of file the Plugin Manager dialog will let you select for installation is a file with the .zip extension.

Make sure your Mac (or a download manager) is not configured to automatically unzip downloaded archives. The downloaded zip file must not be renamed, either.

[odamizu]

Originally Posted by Julie34

... most/all epub3 readers will be able to read epub2 as long as there is a sufficient number of commercial epub2 created ... But you’re right, there is time, there is investments already done …

Originally Posted by KevinH

... I would eventually convert my epub2 ebooks to epub3 slowly (and maybe to epub3.1 eventually - the latest spec that no-one supports) but there really us no rush imho.

Oh, wait. So I should (gradually) convert my epub2s — or at least the ones I may want to re-read in the future?