Keith,
As I discussed at last week's CUD Transition meeting, it does
not appear that there is any better way to handle this issue, other
than stripping out the additional xml:base attributes after processing
the xinclude statements.
Therefore, I'd like to proceed with my original suggested fix:
http://cr.opensolaris.org/~dermot/webrev-6997263/
unless you have any further comments?
- Dermot
On 11/05/10 18:42, Dermot McCluskey wrote:
On 11/05/10 17:54, Keith Mitchell wrote:
Is there perhaps a way in our DTDs to universally allow an attribute
(such as xml:base)? Or to allow it via ManifestParser?
Initially, I thought not. Now, I'm finding a few web pages
that describe how to do something this, or something very similar,
but I cannot get them working as expected on Solaris (possibly
because we have a slightly older version of lxml?).
Anyway, I agree this would be a better solution and I'll put the
current fix on hold while I continue to look into this.
Thanks for bringing this up.
- Dermot
On 11/ 5/10 10:38 AM, Dermot McCluskey wrote:
Keith,
I thought about that, but the problem is that we might, potentially,
want to insert xincluded sub-documents at any point in the XML tree, ie
each of these are potentially legal (I'm simplifying the XML a
little):
<dc>
<include href="inc_target.xml">
</dc>
or
<dc>
<target>
<include href="inc_target_device.xml">
</target>
</dc>
or
<dc>
<target>
<target_device>
<include href="inc_disk.xml">
</target_device>
</target>
</dc>
And the new "base" attribute will get added at the point where
the xincluded sub-doc is inserted. Therefore, we'd have to change the
schema to add an optional "base" attribute to almost every element?
Or perhaps I'm misunderstanding how the whole XInclude mechanism
is meant to work? I'm certainly open to correction on this.
- Dermot
On 11/05/10 17:21, Keith Mitchell wrote:
Hi Dermot,
On a higher level, would it perhaps be better to allow this
attribute to pass validation? This would provide an audit trail of
sorts for xincluded elements, which could be valuable.
- Keith
On 11/ 5/10 04:42 AM, Dermot McCluskey wrote:
Hi,
Can I get a review for this fairly simple fix for ManifestParser bug:
http://monaco.sfbay/detail.jsf?cr=6997263
Webrev is at: http://cr.opensolaris.org/~dermot/webrev-6997263/
Thanks,
- Dermot
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
_______________________________________________
caiman-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
