Q: When I error check my Pages-created epub, I get the error code "undeclared prefix: ibooks." How do I resolve this error?
-
Jan 21, 2014 5:42 AMby Marc Marshall,HelpfulI was getting this exact error in the epub checking tool I was using with a hand-coded EPUB file until I messed around with the XML declarations at the top of the OPF file, at which point they disappeared. Unfortunately, I don't know what, exactly, I changed, nor what was wrong, but I had copied the code from Apple's epub formatting PDF when I was getting the error based on some of the META tags. When I then re-copied the chunk out of one of the sample files available on the iTunes Connect site, they went away.
My assumption is that the error is due to some version of the header linking to an XML dictionary that doesn't include Apple's ibooks META items.
Since your EPUB was generated by Pages, it will almost certainly still work just fine even with those errors--iBooks presumably still knows what the item does, and will handle it correctly.
Unless the book is behaving weirdly when you test it, or the iBooks store is rejecting it due to the eror, I wouldn't worry about it.
-
Jan 21, 2014 6:23 AMby Amelia1927,That's interesting (although I've never actually tried altering an OPF file). The only reason I even know about the error code is that initially the test file wouldn't upload on one of several sites I plan to distribute on (Kobo) other than iBooks. The help people at that site directed me to this epub testing software, which is where I learned about the error.
http://www.pagina-online.de/software/epub-checker
After wasting many hours trying to understand that error code as the reason for the upload failure (and panicking because I thought maybe some proprietary coding makes it so Pages epubs will ONLY upload to iBooks), on a whim, I reduced the file name to letters with no spaces. It uploaded and when redownloaded it was still intact and still working perfectly on Overdrive, Bluefire, Kobo, and iBook reading apps. But the error code shows up in the epub test the same as before.
Your experience might suggest that the problem is actually with the epub checker's failure to recognize some aspect of Apple's META items. Just hope it doesn't cause any sort of issue further down the road to publication, as with the application of DRM by the distributing sites.
-
Jan 21, 2014 6:34 AMby Marc Marshall,HelpfulBased on that, it does sound like the sites you were having problems with had to do with filenames, not this error. I can imagine a site running a standard EPUB verifier, and if any errors at all are thrown rejecting the file, but practically speaking an unknown META tag (what's causing this error) should be ignored by the parser and therefore by definition be harmless.
I was using a free Java-based checker myself, and the fact that I managed to clear up the error indicates to me (unless I'm remembering wrong or misunderstood what fixed it in my case) that it's probably not technically the EPUB checker's failure to recognize the Apple META property; more accurately, it's that the XML dictionary the file is referencing doesn't include the Apple items in it, so the EPUB checker is correctly complaining about it.
It could also be something as simple as Pages generating a file that is marked as an EPUB v2 file but that uses EPUB v3 features (I know my book was intended to be v3, and I could have inadvertently used headers for v2).
Regardless, if the sites accept it, there's a good chance it is okay; DRM, at least, isn't going to change anything--only if an e-reader changes its EPUB parser at some point and breaks something, which seems unlikely.
However, the best way to find out if it's going to work right is to test it. Kindle is far and away the most popular reader, so try it with that and see if it comes out looking how you expect. You should be able to view it in the Kindle application, and there is also a Kindle simulatior available from Amazon that lets you see how an e-paper Kindle will render books without having to buy one. They have very limited EPUB support as far as features go, so it's definitely worth testing even if there are no errors--a lot of things iBooks will display just fine a Kindle will choke on. The older e-paper Nooks, in my experience, are even worse.
-
Jan 21, 2014 7:13 AMby Amelia1927,I wondered about the epub 2 vs epub 3 question. Not knowing whether Pages had created a version 2 or 3 epub (how can you tell?), I loaded the epub into Calibre and set up a conversion from epub to epub--I thought I might be able to convert it from one to other and retry the upload. Calibre assumed my file was an epub 3 and will (so far as I can tell) only convert it to an epub 3 document, so that was sort of pointless.
I also tried going through the steps of generating an epub file by a different process. My original is actually in Word (2008), which I have formatted using Word headers (redefined) that are recognized by Pages as toc elements. (Perhaps this is the source of the error?) It also has inline illustrations and superscripted footnotes. I saved this Word document into an html document, put it in Calibre and converted it to an epub. The results were coherent but not aesthetically acceptable. The superscripted footnotes enlarged so much that the lines they are in spaced out, the text indents vanished, etc. .... All of which fuelled my concern about what that error code means. And all of which it seems is probably irrelevant.
-
Jul 22, 2016 3:55 PMby mdriftmeyer,Inside the sample epub documents that you can extract Apple includes the xml namespace and header structure that needs to be placed at the top of every xhtml document you intended to have in your book.
Between the XML opening <? ?> there is the requisite <html ... "> before you get to the <head> tag. Inside those sample Epub documents are Apple specific vocabulary xmlns and epub:prefix for ibooks. Make sure those are included verbatim, on every xhtml document extension file.
