Landmark in ePub

[Ryan]

Q:
We’re using subscription to convert PDF files into fixed ePub format.

After conversion we’re trying to upload the ePubs to iTune Connect, but always getting the following error message:

ERROR ITMS-5107: "[our ebook title].epub: Fixed-layout books must supply a landmarks" at Book (MZItmspBookPackage)

-----------------------------------------------------------------------
A:
Most, perhaps all, of our ToEpub customers do post-processing of the generated EPUB content. While we generate fully complaint EPUB3 files (and landmarks is optional in the spec), the output is not necessarily ready for online stores. Especially with regards to meta data, such as languages, author, title, bookmarks etc. This is because most PDF's don't contain this information, many are missing title for example, and there is no such thing as landmarks in PDF. We do convert the PDF outline to EPUB toc, but our ToEpub wouldn't know what the landmarks are. For example we have no way to know if there is a bibliography, and where it begins.

Since the nav.xhtml file where the landmarks would go is a XML file, you can use any sort of XML utility to edit it, and add your landmarks, in your own XML post-processing step.

Steps:
1. If you are using PDFNet, or DocPub CLI, you can generate the EPUB unzipped/expanded format. Otherwise, if using online conversion, you would simply unzip the .epub file
2. Post-processing: make XML changes to nav.xhtml
3. run EPUB check on the unzipped epub folder, with the following commands: "epubcheck.bat -mode exp -save epub_root_folder"

The last step will validate the epub, and then package it in the specific epub manner, as epub_root_folder.epub file.

Here is the contents of the epubcheck.bat file that I personally use. Simply put it with the epubcheck files that you can get here. https://github.com/IDPF/epubcheck

@ECHO OFF
java -jar %~dp0\epubcheck-3.0.jar %*
if errorlevel 1 (
   echo Failure Reason Given is %errorlevel%
   exit /b %errorlevel%
)

We may add landmarks, and other iBook store metadata requirements in our conversion, later on, but as mentioned above, it would most likely not be publisher ready and you would still need to post-process the nav.xhtml and package.opf files with a XML utility.