Thomas, I am having some problems with the generated xml:base, so yes,
it my case it's harmful. The issue is, I have a directory hierarchy like
this:
book/chapter/section/images
In each directory there's a docbook file to be included in the upper
level document. So, I have in section1.xml and XInclude with a
reference to images/image1.svg, now the issue is the final generated
"FO" has xml:base "section/images" only. The first two are gone. When I
generate pdf using fop I get an error about not finding the given file.
I don't know how to explain this, but this is the case with xsltproc and
with xalan-j ! this is what really confusing me.
I am including the whole SVG graphic file in my documents so there's no
need for xml:base attribute especially if it's behaving this way. I
looked around for a way to disable this attribute but couldn't find.
My way around was a little bit weired. Since I am generating the SVG
graphics from an xml files. I included the xml file and added a template
to process it in my customization layer. This way it's not relying on an
external graphics file but definitely it takes longer to compile the
document. So it's working for me, but the original issue is not
resolved, which is, the wrong xml:base attribute.
Thomas Schraitle wrote:
Hi,
On Montag, 3. März 2008, Mansour wrote:
I am looking to include a file in my docbook document, but I don't want
to generate xml:base. Is there a way to do this ??
I know "keep.relative.image.uris" can be used. But I am not using
images. I need to include a text/xml file.
I am not sure what you want to achieve, so maybe some background
information might be useful:
According to the XInclude specification[1], "each element information item
in the top-level included items which has a different base URI than its
include parent has an attribute information item added to its attributes
property. [see 4.5.5 Base URI Fixup]" (which is in fact the xml:base
attribute).
There *may* be an option in your XML parser or XInclude processor,
because "An XInclude processor may, at user option, suppress xml:base
and/or xml:lang fixup." [see 4.5 Creating the Result Infoset] So you
should check the available options.
If there is no option, I fear, the only method is to post process your
document with an XSLT stylesheet. You have to copy every element,
attribute, comment and processing instruction node except xml:base. I
don't think, this is a good idea. Normally, xml:base isn't harmful.
Hope it was a bit helpful. :)
Bye,
Tom
-------
[1] http://www.w3.org/TR/xinclude/
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]