EPUBCheck finds errors with footnotes

[Tobias]

EPUBCheck from W3C is throwing a lot of errors on a converted .docx to epub 2 file:

Referring to a footnote:

<a role="doc-noteref"

ERROR(RSC-005): book.epub/index_split_051.html(58,494): Validierungsfehler: Attribut "role" ist an diesem Element nicht erlaubt. Erlaubt ist nur das Attribut "accesskey", "charset", "coords", "dir", "hreflang", "lang", "rel", "rev", "shape", "style", "tabindex", "target", "type" oder "xml:lang"

I am surprised that the conversion tool seems to produce invalid epub2 xhtml?

Is this something that I should care about and correct or does it not matter?

[DNSB]

Perhaps that 'role' is an ePub3 element has something to do with the error. For your own use, simply removing the role="whatever"s will work.

Alternatively, have calibre generate ePub3 files.

[kovidgoyal]

https://manual.calibre-ebook.com/faq...e-is-not-valid

[Tobias]

I think @kovidgoyal 's reply hits it well, thank you!

From the link:

The EPUB I produced with calibre is not valid?

calibre does not guarantee that an EPUB produced by it is valid. The only guarantee it makes is that if you feed it valid XHTML 1.1 + CSS 2.1 it will output a valid EPUB. calibre tries hard to ensure that EPUBs it produces actually work as intended on a wide variety of devices, a goal that is incompatible with producing valid EPUBs, and one that is far more important to the vast majority of its users. If you need a tool that always produces valid EPUBs, calibre is not for you. This means, that if you want to send a calibre produced EPUB to an online store that uses an EPUB validity checker, you have to make sure that the EPUB is valid yourself, calibre will not do it for you – in other words you must feed calibre valid XHTML + CSS as the input documents.

[Tobias]

Originally Posted by DNSB

For your own use, simply removing the role="whatever"s will work.

I just got feedback from a publisher who told me to make sure the file is 100% valid, because otherwise there can be error messages from shops during distribution.

[Tobias]

I brought this into the bug tracker. It does not sit right with me that the role attribute is not removed, but also not intentionally added. Unfortunately @kovidgoyal is very quick with shutting down conversations.

https://bugs.launchpad.net/calibre/+bug/2072645

[theducks]

Originally Posted by Tobias

I brought this into the bug tracker. It does not sit right with me that the role attribute is not removed, but also not intentionally added. Unfortunately @kovidgoyal is very quick with shutting down conversations.

https://bugs.launchpad.net/calibre/+bug/2072645

You DO NOT have to use HIS FREE SOFTWARE.

[retiredbiker]

Over the past decade or so I've prepared about a dozen books for publication working with several authors. At some point on nearly every single one I had to study up on some "error" thrown by some checking software, and go into the code to fix it. It is a moving target. A file Amazon took a month ago may not pass tomorrow. And it rarely makes any sense.

If you want to publish, just running a conversion will rarely be enough. You have to be willing to get into the code, learn the new glitch-of-the-day, and keep smiling. Or snarling. But that's the way it is.

[kovidgoyal]

Originally Posted by Tobias

I brought this into the bug tracker. It does not sit right with me that the role attribute is not removed, but also not intentionally added. Unfortunately @kovidgoyal is very quick with shutting down conversations.

Do you really think I haven't had this conversation before, half a dozen times. With DOCX input it started years ago, with the start attribute on <ol> tags, which also epubcheck stupidly complains about. Then too people were trying to get me to dumb down calibre's output to suit epubcheck, I refused for the same reasons.

The *only* valid reason for me to dumb down calibre's output is if that output causes issues with a significant fraction of real world reading software. epubcheck simply isn't a reason I will ever do that. As I said in the FAQ, if this is a problem for you, then calibre is not the software for you.

[DNSB]

Originally Posted by kovidgoyal

The *only* valid reason for me to dumb down calibre's output is if that output causes issues with a significant fraction of real world reading software. epubcheck simply isn't a reason I will ever do that. As I said in the FAQ, if this is a problem for you, then calibre is not the software for you.

Or you use calibre realizing that you may have to check and correct issues spit out by epubcheck if you are preparing an ePub for commercial use where they seem to love using epubcheck. For the end user, the renderer you use is supposed to treat any unknown/unimplemented items as nulls. For me, items such as role: in an ePub2 have never caused me any issues when reading.

For me, the time saved by using calibre more than covers the time to correct any issues that epubcheck may have with a calibre generated epub. Specifying ePub3 as the output helps so role:, epub:type, epub:prefix, etc. are not considered errors.