31 Aug
2006
31 Aug
'06
8:01 p.m.
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.
--
"First they ignore you. Then they laugh at you. Then they
fight you. Then you win." - Mohandas Gandhi.
31 Aug
31 Aug
9:28 p.m.
New subject: [xwiki-users] Broken PDF export
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
--
Ludovic Dubost
XPertNet: http://www.xpertnet.fr/
Blog: http://www.ludovic.org/blog/
XWiki: http://www.xwiki.com
Skype: ldubost AIM: nvludo Yahoo: ludovic
9:53 p.m.
New subject: [xwiki-users] Broken PDF export
On Thu, Aug 31, 2006 at 09:28:24PM +0200, Ludovic Dubost wrote:
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..)
I've created a document containig single letter 'a'. Xwiki behaves
in just the same way - the same error in catalina.out.
I cannot find any setting regarding temporary dir. Besides, I guess it
would influence also on file uploads and I can upload attachments
without problems.
Richard.
--
"First they ignore you. Then they laugh at you. Then they
fight you. Then you win." - Mohandas Gandhi.
10:01 p.m.
New subject: [xwiki-users] Broken PDF export
Sorry I didn't look at the exception.. Clearly it seems to be a UTF-8
problem.
Maybe Xavier or Stephane have an idea since they are running xwiki on UTF-8
Ludovic
Ryszard Lach a écrit :
On Thu, Aug 31, 2006 at 09:28:24PM +0200, Ludovic
Dubost wrote:
--
Ludovic Dubost
XPertNet: http://www.xpertnet.fr/
Blog: http://www.ludovic.org/blog/
XWiki: http://www.xwiki.com
Skype: ldubost AIM: nvludo Yahoo: ludovic
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..)
I've created a document containig single letter 'a'. Xwiki behaves
in just the same way - the same error in catalina.out.
I cannot find any setting regarding temporary dir. Besides, I guess it
would influence also on file uploads and I can upload attachments
without problems.
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
1 Sep
1 Sep
3 p.m.
New subject: [xwiki-users] Broken PDF export of UTF-8 documents
On Thu, Aug 31, 2006 at 10:01:04PM +0200, Ludovic Dubost wrote:
Sorry I didn't look at the exception.. Clearly it seems to be a UTF-8
problem.
Maybe Xavier or Stephane have an idea since they are running xwiki on UTF-8
Now I can confirm it. Problem is for sure in xsl transformation and
occurs only when using non-latin1 characters.
With brand new xwiki database I can export document containing 'a' or 'ó'
(oacute), which have the same codes in LATIN1, LATIN2 and UTF-8. When I
enter some other non-LATIN1 characters - export fails with exception
[Fatal Error] :66:27: Invalid byte 2 of 2-byte UTF-8 sequence.
11:43:33,806 WARN http-8080-Processor24
http://localhost:8080/xwiki/bin/pdf/Sandbox/TestTopic1
XWikiAction:execute:185 - Uncaught exception: Error number 12003 in 12:
XSL Transformation Failed
Wrapped Exception: Invalid byte 2 of 2-byte UTF-8 sequence.
com.xpn.xwiki.XWikiException: Error number 12003 in 12: XSL
Transformation Failed
Wrapped Exception: Invalid byte 2 of 2-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)
[...]
Problem with saving a document containing only single letter 'a' in my
previous tests has probably something to do with my database, which I
used - it has plenty documents with non-LATIN1 content and probably some
charactes not belonging to editable part of xwiki page (maybe some
preferences?) was transformed with xslt throwing an exception.
I have xwiki run on eclipse, but I'm not a programmer, so I affraid I
won't be able to find much more....
Does anybody have working pdf-export of UTF-8 documents?
Richard.
--
"First they ignore you. Then they laugh at you. Then they
fight you. Then you win." - Mohandas Gandhi.
4 Sep
4 Sep
12:46 p.m.
New subject: [xwiki-users] Broken PDF export of UTF-8 documents
Dear all,
For UTF-8 pdf generation, you have to patch the following class :
com.xpn.xwiki.pdf.impl.PdfExportImpl
(http://svn.forge.objectweb.org/cgi-bin/viewcvs.cgi/xwiki/xwiki/trunk/src/ma…)
by replacing the ligne :
tidy.setCharEncoding(Configuration.LATIN1);
with
tidy.setCharEncoding(Configuration.UTF8);
It works with French characters but not completly with Chinese or Vietnamese.
Regards
--
Xavier MOGHRABI - Consortium ObjectWeb
Jabber: moghrabi(a)jabber.objectweb.org - Phone: +33 4 76 61 52 35
Le vendredi 1 septembre 2006 15:00, Richard Lach a écrit :
On Thu, Aug 31, 2006 at 10:01:04PM +0200, Ludovic
Dubost wrote:
Sorry I didn't look at the exception..
Clearly it seems to be a UTF-8
problem.
Maybe Xavier or Stephane have an idea since they are running xwiki on
UTF-8
Now I can confirm it. Problem is for sure in xsl transformation and
occurs only when using non-latin1 characters.
With brand new xwiki database I can export document containing 'a' or
'ó'
(oacute), which have the same codes in LATIN1, LATIN2 and UTF-8. When I
enter some other non-LATIN1 characters - export fails with exception
[Fatal Error] :66:27: Invalid byte 2 of 2-byte UTF-8 sequence.
11:43:33,806 WARN http-8080-Processor24
http://localhost:8080/xwiki/bin/pdf/Sandbox/TestTopic1
XWikiAction:execute:185 - Uncaught exception: Error number 12003 in 12:
XSL Transformation Failed
Wrapped Exception: Invalid byte 2 of 2-byte UTF-8 sequence.
com.xpn.xwiki.XWikiException: Error number 12003 in 12: XSL
Transformation Failed
Wrapped Exception: Invalid byte 2 of 2-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) [...]
Problem with saving a document containing only single letter 'a' in my
previous tests has probably something to do with my database, which I
used - it has plenty documents with non-LATIN1 content and probably some
charactes not belonging to editable part of xwiki page (maybe some
preferences?) was transformed with xslt throwing an exception.
I have xwiki run on eclipse, but I'm not a programmer, so I affraid I
won't be able to find much more....
Does anybody have working pdf-export of UTF-8 documents?
Richard.
5 Sep
5 Sep
8:18 a.m.
New subject: [xwiki-users] Broken PDF export of UTF-8 documents
On Mon, Sep 04, 2006 at 12:46:56PM +0200, Xavier MOGHRABI wrote:
Dear all,
For UTF-8 pdf generation, you have to patch the following class :
com.xpn.xwiki.pdf.impl.PdfExportImpl
(http://svn.forge.objectweb.org/cgi-bin/viewcvs.cgi/xwiki/xwiki/trunk/src/ma…)
by replacing the ligne :
tidy.setCharEncoding(Configuration.LATIN1);
with
tidy.setCharEncoding(Configuration.UTF8);
It works with French characters but not completly with Chinese or Vietnamese.
Thank's a a lot, that's it!
Documents are generated now, at least, and are returned with
appplication/x-pdf content type. Polish characters are not working yet,
but I've heard I should configure polish fonts for FOP.
Best regards,
R.
--
"First they ignore you. Then they laugh at you. Then they
fight you. Then you win." - Mohandas Gandhi.
6691
days inactive
6696
days old
6 comments
4 participants
participants (4)
-
Ludovic Dubost -
Richard Lach -
Ryszard Lach
-
Xavier MOGHRABI