Hi,
For PDF export the first thing to check is if you can export an empty or
near empty document
If not then you probably have a setup problem (temp dir, etc..)
If yes, then it has to do with the content
Known problems are:
- PDF export does not always support custom HTML. HTML easily breaks the
export
- PDF export refuses to have to subtitles with the same name. Currently
it generates anchors with the same ID which is non valid HTML and breaks
the PDF export
Somebody is actually working to fix this issue in the anchor ID generation
Concerning non latin issues, I'm not sure I'm a good help  !
Ludovic
Ryszard Lach a écrit :
  Hi.
 We are using xwiki for almost two years (as internal knowledge base) and
 although we still have some problems - thank you very much for all you
 work, I hope we will be able to solve some paining issues ;-), I'm
 pretty determined to dig until success, I count on your help and promise
 to give sufficient feedback.
 Our main problems are:
 * broken PDF export (generated null-byte files)
 * broken search on non-latin1 characters (that I'll describe in separate
    thread)
 Hera are the details
 My platform is:
 tomcat 5.5.17 running in locale pl_PL.UTF-8
 JDK 1.5.0
 mysql 5.0 with default charset UTF-8 (and DB in UTF-8 too).
 xwiki.war 9.840, but the same bahaviour is on latest SVN trunk.
 I suppose that problem with PDF export has something to do with UTF-8.
 I'm pretty sure I have UTF-8 configured properly (xwiki.encoding in
 xwiki.cfg and filter in web.xml), non-latin1 characters are being properly
 written to and read from database, are also properly rendered in xwiki
 pages. When I click PDF export link (even viewing document containing
 only ASCII characters) I receive an empty file and some entries in
 catalina.out:
 ########
 <?xml version="1.0" encoding="UTF-8" ?>
 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
     "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
     <html 
xmlns="http://www.w3.org/1999/xhtml" lang="en"
xml:lang="en">
     <head>
     <title>
     Wiki - Sandbox - TestTopic1
     </title>
     <meta http-equiv="Content-Type" content="text/html;
charset=UTF-8"
     />
     <meta name="revisit-after" content="7 days" />
     <meta name="description" content="Intranet Wiki" />
     <meta name="keywords" content="wiki" />
     <meta name="distribution" content="GLOBAL" />
     <meta name="rating" content="General" />
     <meta name="copyright" content="you need to choose your
copyright"
     />
     <meta name="author" content="" />
     <meta http-equiv="reply-to" content="" />
     <meta name="language" content="pl" />
     <meta http-equiv="Content-Type" content="text/html;
charset=utf-8"
     />
 [...]
 line 5 column 1 - Warning: <html> has XML attribute "xml:lang"
 [Fatal Error] :78:27: Invalid byte 1 of 1-byte UTF-8 sequence.
 19:53:36,729  WARN TP-Processor4
 
https://xwiki.biuro.contium.pl:5443/xwiki/bin/pdf/Sandbox/TestTopic1
 XWikiAction:execute:185 - Uncaught exception: Error number 12003 in 12:
 XSL Transformation Failed
 Wrapped Exception: Invalid byte 1 of 1-byte UTF-8 sequence.
 com.xpn.xwiki.XWikiException: Error number 12003 in 12: XSL
 Transformation Failed
 Wrapped Exception: Invalid byte 1 of 1-byte UTF-8 sequence.
         at com.xpn.xwiki.pdf.impl.PdfExportImpl.applyXsl(PdfExportImpl.java:176)
         at
com.xpn.xwiki.pdf.impl.PdfExportImpl.convertXHtmlToXMLFO(PdfExportImpl.java:154)
         at com.xpn.xwiki.pdf.impl.PdfExportImpl.exportXHtml(PdfExportImpl.java:86)
         at com.xpn.xwiki.pdf.impl.PdfExportImpl.exportHtml(PdfExportImpl.java:115)
         at com.xpn.xwiki.pdf.impl.PdfExportImpl.export(PdfExportImpl.java:129)
         at com.xpn.xwiki.pdf.impl.PdfExportImpl.exportToPDF(PdfExportImpl.java:119)
         at com.xpn.xwiki.web.PDFAction.render(PDFAction.java:45)
         at com.xpn.xwiki.web.XWikiAction.execute(XWikiAction.java:146)
 [...]
 ########
 Do you any have any idea what can be the cause of this exception?
 I suppose that
 '    <html 
xmlns="http://www.w3.org/1999/xhtml" lang="en"
xml:lang="en">'
 has nothing to do, although I don't know why lang, content-type and
 charset are derived both from templates/httpheader.vm and
 xwiki->preferences.
 I hope I will be able to run it under eclipse soon, maybe I find the
 problem....
 Best regards,
 Richard.
 ------------------------------------------------------------------------
 --
 You receive this message as a subscriber of the xwiki-users(a)objectweb.org mailing list.
 To unsubscribe: mailto:xwiki-users-unsubscribe@objectweb.org
 For general help: mailto:sympa@objectweb.org?subject=help
 ObjectWeb mailing lists service home page: 
http://www.objectweb.org/wws