On Jun 11, 2012, at 12:21 AM, Hans Hagen wrote:
The validator likes it, next is the hard part: generating a valid xhtml. Is the xhtml generator only for the epub, i.e. can it be changed or is another output file just for the epub appropriate?
the only issue is that css has no method for mapping a link which is why the exporter produces two files (xml and xhtml)
I get an error by the epub validator if I feed it the generated epub. The content file is: -START------------ <?xml version='1.0' encoding='UTF-8' standalone='yes' ?> <!-- input filename : test_00 --> <!-- processing date : Mon Jun 11 07:03:39 2012 --> <!-- context version : 2012.06.11 00:21 --> <!-- exporter version : 0.30 --> <document xmlns:m="http://www.w3.org/1998/Math/MathML" version="0.30" language="de" date="Mon Jun 11 07:03:39 2012" file="test_00" context="2012.06.11 00:21" xmlns:xhtml="http://www.w3.org/1999/xhtml"> Hello world! </document> -END------------ the error is: -START------------ Epubcheck Version 3.0b5 Validating against EPUB version 2.0 ERROR: test_00.epub/OEBPS/test_00.xhtml(9,202): elements from namespace "" are not allowed Check finished with warnings or errors -END------------ A possible validatable file would be -START------------ <?xml version="1.0" encoding="UTF-8" ?> <!-- input filename : test_00 --> <!-- processing date : Sat Jun 9 14:36:39 2012 --> <!-- context version : 2012.06.09 01:22 --> <!-- exporter version : 0.30 --> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en"> <head> <title>TITLE</title> </head> <body> <div id="test_00"> Hello world! </div> </body> </html> -END------------ It just wraps the filename without extension in title plus head and the main part in body plus a div. However, I did not try, yet, how many or which devices can read it just fine.
Calibre is already happy with it and shows it just fine. Please note that some ebook readers (e.g. apple devices, calibre) have a nasty caching mechanism. If you want to see the new changes, I found that it is best to delete the old book before replacing it.
Andy
xmlns + doctype is probably a bit over the top for the coverpage
I slavishly follow the mandatory epub specification and then add some things to make common devices work. Calibre, Adobe Digital Editions and Apple devices want a xhtml cover page in a very special way and ignore the cover image otherwise. The next thing I would add is a toc.xhtml. Here, a Kindle ignores the toc.ncx and wants a html toc as well. Unfortunately, this blows up mtx-epub, but at least that is the last issue with reading devices that I know. Andy