Yes/No custom column to automatically show whether a book is in EPUB 3 format?

[ShellShock]

Is this possible? The column would be automatically updated when adding a new book to Calibre, or when converting it in the Calibre book editor from EPUB 2 to EPUB 3.

I have recently discovered that Calibre stores series numbers as integers in EPUB 3 e.g.,

<opf:meta refines="#id-2" property="group-position">4</opf:meta>

but uses floats for EPUB 2, e.g.,

<meta name="calibre:series_index" content="4.0"/>

I prefer the integer format when it really is an integer (and not e.g., 4.1). Thus I am converting my library from EPUB 2 to 3, and it would be handy to easily tell which books are already EPUB 3.


Thanks

[kovidgoyal]

its a float in both. it just serialized as an integer in epub 3 when the floating value is actually integral. And no there is no automatic mechanism for this. But you can just use the upgrade book internals option in the POlish books tool to bulk update all your epub to epub 3 it will do nothing if a file is already epub 3

[thiago.eec]

I have a yes/no column for EPUB3 books. Although not automatic, it checks the whole library in just a few seconds using 'Quality Check' plugin.

1) Click the plugin and choose 'Search ePubs...'
2) Use this regex expression:

xmlns="http://www\.idpf\.org/2007/opf"(.*?)version="3\.0"

3) Mark 'Ignore case', 'Show all occurrences' and choose 'OPF manifest' as scope
4) Click OK
5) It will mark all books on your library that have a EPUB3 format
6) Select all the books and do a batch metadata change, setting the corresponding column to YES

The regex expression and the options will be remembered by the plugin for future use.

[Quoth]

You need an integer?
0 = epub2
1 = basic epub3, a reflowable ebook, no interactive and fall backs for epub2 (probably Project Gutenberg epub3 titles).
2 = basic epub3, but no epub 2 index etc, no fallbacks, specific hand crafted epub3 inc math ml. No javascript, transparency, animation, audio or video.
3 = FXL aka Fixed Layout aka Print Replica epub3 imitating a PDF.
4 = Multimedia and/or Interactive epub3. May have javascript, video, audio etc.

Obviously given epub 3 flexibility and features it's more complex. Maybe also the epub3 numbers should start at 2 and 1 should be for Kobo's kepub.

Any reflowable mobi, prc, docx, odt, lit, pdb, lrf, epub2, azw3, kfx or kepub converted automatically with no hand editing is going to be a type 1 epub3.

However these are the broad categories and it's what I would do, if I wanted it.

[ShellShock]

Originally Posted by kovidgoyal

its a float in both. it just serialized as an integer in epub 3 when the floating value is actually integral. And no there is no automatic mechanism for this. But you can just use the upgrade book internals option in the POlish books tool to bulk update all your epub to epub 3 it will do nothing if a file is already epub 3

Thanks, I will try the Polish tool - I had been using bulk convert.

[ShellShock]

Originally Posted by thiago.eec

I have a yes/no column for EPUB3 books. Although not automatic, it checks the whole library in just a few seconds using 'Quality Check' plugin.

1) Click the plugin and choose 'Search ePubs...'
2) Use this regex expression:

xmlns="http://www\.idpf\.org/2007/opf"(.*?)version="3\.0"

3) Mark 'Ignore case', 'Show all occurrences' and choose 'OPF manifest' as scope
4) Click OK
5) It will mark all books on your library that have a EPUB3 format
6) Select all the books and do a batch metadata change, setting the corresponding column to YES

The regex expression and the options will be remembered by the plugin for future use.

Great, this is really useful.

[ShellShock]

I have been using the Polish tool to upgrade epubs from 2 to 3, and there are some differences to how it formats the meta-data, compared to Edit book:

In my epub 2 meta-data the series position was defined in float format like this:

<meta name="calibre:series_index" content="2.0"/>

I used Polish book, Upgrade book internals to just upgrade from epub 2 to 3:

Polishing: EPUB
Upgrading book, if possible
Updated EPUB from version 2 to 3
Polishing took: 0.2 seconds

I unzipped the resulting epub to examine the meta-data; the series position is still in float format:

[CODE]<meta refines="#id-2" property="group-position">2.0</meta>/CODE]

I then opened the book in the Calibre editor (Edit book command), which I think silently re-arranges some of the books internals and automatically saves the result (without prompting). When I unzipped the epub again the series position is now in integral format and the tag has acquired an opf namespace:

<opf:meta refines="#id-2" property="group-position">2</opf:meta>

This explains why I was sometimes seeing float and sometimes integral format for series positions once I had downloaded epubs from Calibre to my reader app (I prefer integral format). It all depended on whether the epub was in version 3 format, and if I had opened it in the Calibre editor.

I like how the epub editor will automatically convert an effectively integral series position (2.0) to an integer (2) - after all, most series positions will be integral. When you do have a sub-position, e.g., 2.1, the editor respects that too.

Please would it be possible to make the Polish tool consistent with the epub editor in this regard?

[kovidgoyal]

Not something I care about but patches are welcome.