On 12/13/2011 04:27 PM, Jirka Kosek wrote: > On 13.12.2011 12:03, Hussein Shafie wrote: > >> After rewriting the EPUB stylesheets from scratch, we have contributed >> them to the DocBook XSL Stylesheets open source project. However, this >> open source project has preferred to keep the original stylesheets >> (which are maintained by Keith Fahlgren -- >> http://www.oreillynet.com/pub/au/2878 -- who seems deeply involved in EPUB). > > Hi, > > that's pitty, I wasn't aware of this.
Having our contribution ``rejected'' by the DocBook XSL Stylesheets open source project is absolutely not a problem for XMLmind. We are committed to always provide our users with production quality software. Unfortunately for us, from time to time, this means reinventing the wheel (because the existing wheel is not round enough). > > Anyway to make it even more complicated there is now third possibility. > New EPUB3 stylesheets has been recently contributed by Bob Stayton, see: > http://docbook.svn.sourceforge.net/viewvc/docbook/trunk/xsl/epub3/ > > They are definitively much better then previous one, but they target > EPUB3. But in my experience is that files are very compatible with > existing readers. Thank you for the information. > >> We actually have no problem with the standard DocBook XSL Stylesheets. It's >> just >> that we are not familiar with generating EPUB formats and the DocBook way to >> do >> it in the DocBook XSL version we use is a ruby script. Our application is >> made >> in Java so using this ruby script is not convenient at all for us. That's >> why we >> thought we might use XXE capabilities to perform the work. Checking the >> document >> XXE_install_dir/addon/config/docbook/xsl/CHANGES_MADE_BY_XMLMIND.txt, it >> looked >> like the XSL coming with XXE fixed some issues from the standard one. > > In Java world you can use ant build resulting EPUB file, using something > like: > > <target name="epub"> > <xslt in="${in}" out="dummy.html" > style=".../docbook-xsl/epub3/chunk.xsl" force="true"> > <factory name="com.icl.saxon.TransformerFactoryImpl"/> > <classpath> > <pathelement location="${saxon.jar}"/> > <pathelement location="${saxon.extensions.jar}"/> > <pathelement location="${xslthl.jar}"/> > <pathelement location="${xml-apis.jar}"/> > </classpath> > <xmlcatalog> > <catalogpath location="catalog.xml"/> > </xmlcatalog> > </xslt> > <zip destfile="${out}.epub" > basedir="ebook" > includes="mimetype" > compress="false" > /> > <zip destfile="${out}.epub" > basedir="ebook" > excludes="mimetype" > update="true" > compress="true" > keepcompression="true" > /> > </target> Thanks. It's clearly not the use of a Ruby script which prevented us from using the stock DocBook XSLT stylesheets which generate EPUB2. > > But of course XXE convert command can provide much more, like copying > resources, handling image conversion, ... > > Sorry if this too off-topic for this list. > It's absolutely not off topic. We'll of course get rid of our XSLT stylesheets that generate EPUB2 when EPUB3 will become a little more mainstream. For now, we prefer to keep using our own XSLT stylesheets mainly because we have successfully tested their outputs (including cover pages) against many EPUB readers and checkers: epubcheck, Apple iBooks, Adobe Digital Editions, Calibre, Bookworm, etc. http://code.google.com/p/epubcheck/ http://www.apple.com/ipad/built-in-apps/ibooks.html http://www.adobe.com/products/digitaleditions/ http://calibre-ebook.com/ http://bookworm.oreilly.com/ -- XMLmind XML Editor Support List [email protected] http://www.xmlmind.com/mailman/listinfo/xmleditor-support
