You are viewing a plain text version of this content. The canonical link for it is here.
Posted to c-users@xalan.apache.org by Holger Flörke <fl...@doctronic.de> on 2004/01/09 15:51:00 UTC
A Simple Stylesheet failed on CVS-Xalan
Hi there,
I have checked out the tagged version "Xalan-C_1_7_0" (it is not
official I know) from the apache CVS yesterday and build it against
Xerces-C 2.4. The simple stylesheet
<?xml version="1.0" encoding="iso-8859-1"?>
<xsl:stylesheet
version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:template match="/">
<table border="" cols=""/>
</xsl:template>
</xsl:stylesheet>
fails, because a bug of Xerces-C 2.4 within IGXMLScanner.cpp (fixed in
revision 1.44) complaining about the attribute "cols" double defined. You
can execute Xerces-C 2.4 DOMPrint.exe on this document and Xerces will also
complain.
My question: Am I right with the simple stylesheet fault? (Maybe this is a
bug on my system) If the assumption holds true, a Xalan-C 1.7 build against
Xerces-C 2.4 would be useless. In this case I would prefer a build against
Xerces-C 2.3 (or somebody know similar bugs of Xerces-C 2.3?).
HolgeR
--
holger floerke d o c t r o n i c
email floerke@doctronic.de information publishing + retrieval
phone +49 2222 9292 90 http://www.doctronic.de
Re: A Simple Stylesheet failed on CVS-Xalan
Posted by Berin Lautenbach <be...@wingsofhermes.org>.
david_n_bertoni@us.ibm.com wrote:
> I didn't realize we could actually build on Nagoya, so having something
> could be really useful. The thing about that's troublesome is it wasn't
> uncovered in Xerces-C's normal test process, so I'm not sure if even
> automating the testing process would help. Still, it's something that
> can't hurt, and might be really helpful.
Would the Xalan tests have caught it? If so, we should be able to
automatically capture.
>
> Is there hardware available within Apache to cover a reasonable subset of
> our platforms, if we were to use your system?
Nagoya gives us Solaris 2.8, with gcc 3.1 and CC 5.4.
Moof gives us Mac OS X 10.3, but I've never got myself access.
I have some systems that I am setting up to autobuild my own stuff,
which gives us Linux (gcc 2.95.4 - Debian stable and gcc 3.3.3 - Debian
Unstable), FreeBSD (4.9) and NetBSD (1.6.1).
Cheers,
Berin
Re: A Simple Stylesheet failed on CVS-Xalan
Posted by da...@us.ibm.com.
Hi Berin,
Sorry, I guess I'm also taking a long time to reply! ;-)
I didn't realize we could actually build on Nagoya, so having something
could be really useful. The thing about that's troublesome is it wasn't
uncovered in Xerces-C's normal test process, so I'm not sure if even
automating the testing process would help. Still, it's something that
can't hurt, and might be really helpful.
Is there hardware available within Apache to cover a reasonable subset of
our platforms, if we were to use your system?
Thanks!
Dave
Berin Lautenbach
<berin@wingsofhe To: xalan-c-users@xml.apache.org
rmes.org> cc: (bcc: David N Bertoni/Cambridge/IBM)
Subject: Re: A Simple Stylesheet failed on CVS-Xalan
01/19/2004 01:01
AM
Please respond
to xalan-c-users
David,
Apologies - this is *long* after your original e-mail.
I've been playing with a *very simple* system to do some autobuilding
out of CVS. Similar to gump, but much simpler. The reason for the
simpler being dependency checking is more difficult, and I want to
autobuild multiple OSes off a single NFS mount. Linux and *BSD being
the main ones. So it will start clean each time and then cause a build
to occur using a given compiler set and defined set of depending
libraries (the idea being to autotest a CVS project against multiple CVS
libraries - as is done with gump - but also autotest against defined
previous versions. E.g. Xalan CVS against Xerces 2.3.
The script involved is a Perl script using an XML configuration file.
Now the thought crosses my mind that we should be able to set something
up on Nagoya to do this for Xerces/Xalan weekly/nightly so that a bug
like this might be caught as it occurred, rather than after a release.
It's very primitive, but it might be useful?
Cheers,
Berin
david_n_bertoni@us.ibm.com wrote:
>
>
>
> Hi Holger,
>
> Yes, this is a major bug in Xerces-C, but I don't know what else we can
do.
> I will post a note on the Xerces-C list to see if they have any intention
> of doing a 2.4.1 release anytime soon.
>
> Any other suggestions are welcome, but I'm not sure releasing with Xerces
> 2.3 is a good idea.
>
> Thanks!
>
> Dave
>
>
>
>
> Holger Flörke
> <floerke@doctron To:
xalan-c-users@xml.apache.org
> ic.de> cc: (bcc: David N
Bertoni/Cambridge/IBM)
> Subject: A Simple
Stylesheet failed on CVS-Xalan
> 01/09/2004 06:51
> AM
> Please respond
> to xalan-c-users
>
>
>
>
> Hi there,
>
> I have checked out the tagged version "Xalan-C_1_7_0" (it is not
> official I know) from the apache CVS yesterday and build it against
> Xerces-C 2.4. The simple stylesheet
>
> <?xml version="1.0" encoding="iso-8859-1"?>
> <xsl:stylesheet
> version="1.0"
> xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
>
> <xsl:template match="/">
> <table border="" cols=""/>
> </xsl:template>
>
> </xsl:stylesheet>
>
> fails, because a bug of Xerces-C 2.4 within IGXMLScanner.cpp (fixed in
> revision 1.44) complaining about the attribute "cols" double defined. You
> can execute Xerces-C 2.4 DOMPrint.exe on this document and Xerces will
also
>
> complain.
>
> My question: Am I right with the simple stylesheet fault? (Maybe this is
a
> bug on my system) If the assumption holds true, a Xalan-C 1.7 build
against
>
> Xerces-C 2.4 would be useless. In this case I would prefer a build
against
> Xerces-C 2.3 (or somebody know similar bugs of Xerces-C 2.3?).
>
> HolgeR
> --
> holger floerke d o c t r o n i c
> email floerke@doctronic.de information publishing + retrieval
> phone +49 2222 9292 90 http://www.doctronic.de
>
>
>
>
>
Re: A Simple Stylesheet failed on CVS-Xalan
Posted by Berin Lautenbach <be...@wingsofhermes.org>.
David,
Apologies - this is *long* after your original e-mail.
I've been playing with a *very simple* system to do some autobuilding
out of CVS. Similar to gump, but much simpler. The reason for the
simpler being dependency checking is more difficult, and I want to
autobuild multiple OSes off a single NFS mount. Linux and *BSD being
the main ones. So it will start clean each time and then cause a build
to occur using a given compiler set and defined set of depending
libraries (the idea being to autotest a CVS project against multiple CVS
libraries - as is done with gump - but also autotest against defined
previous versions. E.g. Xalan CVS against Xerces 2.3.
The script involved is a Perl script using an XML configuration file.
Now the thought crosses my mind that we should be able to set something
up on Nagoya to do this for Xerces/Xalan weekly/nightly so that a bug
like this might be caught as it occurred, rather than after a release.
It's very primitive, but it might be useful?
Cheers,
Berin
david_n_bertoni@us.ibm.com wrote:
>
>
>
> Hi Holger,
>
> Yes, this is a major bug in Xerces-C, but I don't know what else we can do.
> I will post a note on the Xerces-C list to see if they have any intention
> of doing a 2.4.1 release anytime soon.
>
> Any other suggestions are welcome, but I'm not sure releasing with Xerces
> 2.3 is a good idea.
>
> Thanks!
>
> Dave
>
>
>
>
> Holger Flörke
> <floerke@doctron To: xalan-c-users@xml.apache.org
> ic.de> cc: (bcc: David N Bertoni/Cambridge/IBM)
> Subject: A Simple Stylesheet failed on CVS-Xalan
> 01/09/2004 06:51
> AM
> Please respond
> to xalan-c-users
>
>
>
>
> Hi there,
>
> I have checked out the tagged version "Xalan-C_1_7_0" (it is not
> official I know) from the apache CVS yesterday and build it against
> Xerces-C 2.4. The simple stylesheet
>
> <?xml version="1.0" encoding="iso-8859-1"?>
> <xsl:stylesheet
> version="1.0"
> xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
>
> <xsl:template match="/">
> <table border="" cols=""/>
> </xsl:template>
>
> </xsl:stylesheet>
>
> fails, because a bug of Xerces-C 2.4 within IGXMLScanner.cpp (fixed in
> revision 1.44) complaining about the attribute "cols" double defined. You
> can execute Xerces-C 2.4 DOMPrint.exe on this document and Xerces will also
>
> complain.
>
> My question: Am I right with the simple stylesheet fault? (Maybe this is a
> bug on my system) If the assumption holds true, a Xalan-C 1.7 build against
>
> Xerces-C 2.4 would be useless. In this case I would prefer a build against
> Xerces-C 2.3 (or somebody know similar bugs of Xerces-C 2.3?).
>
> HolgeR
> --
> holger floerke d o c t r o n i c
> email floerke@doctronic.de information publishing + retrieval
> phone +49 2222 9292 90 http://www.doctronic.de
>
>
>
>
>
Re: A Simple Stylesheet failed on CVS-Xalan
Posted by da...@us.ibm.com.
Hi Holger,
Yes, this is a major bug in Xerces-C, but I don't know what else we can do.
I will post a note on the Xerces-C list to see if they have any intention
of doing a 2.4.1 release anytime soon.
Any other suggestions are welcome, but I'm not sure releasing with Xerces
2.3 is a good idea.
Thanks!
Dave
Holger Flörke
<floerke@doctron To: xalan-c-users@xml.apache.org
ic.de> cc: (bcc: David N Bertoni/Cambridge/IBM)
Subject: A Simple Stylesheet failed on CVS-Xalan
01/09/2004 06:51
AM
Please respond
to xalan-c-users
Hi there,
I have checked out the tagged version "Xalan-C_1_7_0" (it is not
official I know) from the apache CVS yesterday and build it against
Xerces-C 2.4. The simple stylesheet
<?xml version="1.0" encoding="iso-8859-1"?>
<xsl:stylesheet
version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:template match="/">
<table border="" cols=""/>
</xsl:template>
</xsl:stylesheet>
fails, because a bug of Xerces-C 2.4 within IGXMLScanner.cpp (fixed in
revision 1.44) complaining about the attribute "cols" double defined. You
can execute Xerces-C 2.4 DOMPrint.exe on this document and Xerces will also
complain.
My question: Am I right with the simple stylesheet fault? (Maybe this is a
bug on my system) If the assumption holds true, a Xalan-C 1.7 build against
Xerces-C 2.4 would be useless. In this case I would prefer a build against
Xerces-C 2.3 (or somebody know similar bugs of Xerces-C 2.3?).
HolgeR
--
holger floerke d o c t r o n i c
email floerke@doctronic.de information publishing + retrieval
phone +49 2222 9292 90 http://www.doctronic.de