Thanks, I found out where actually the problem is. By the way, you have any documents and samples on DOMNodeIterator. Does it make the traversal easier ? I need to traverse the entire tree; but with recursive DFS, the thing is not really obvious... I need to carry a few objects through the recursive call... Hope there is something better ...
Vu
-----Original Message-----
From: Alberto Massari [mailto:[EMAIL PROTECTED]]
Sent: Friday, February 25, 2005 3:18 PM
To: xerces-c-dev@xml.apache.org
Cc: Nguyen, Vu [CAR:SI12:EXCH]
Subject: Re: FW: [jira] Commented: (XERCESC-1355) DOMBuilder parse wrong node type
Hi Vu,
At 18.46 24/02/2005 -0500, Vu Nguyen wrote:
>Just in case you did not get this one. Here all the necessary
>attachments.
It looks like I finally got your attachments ;-)
There is a bug in your code, that will cause to expand just the first element. Inside buildCLISubTree(DOMNode *n, int depth) you first display the
attributes, then you do this:
if (n->getNodeType() == DOMNode::ELEMENT_NODE )
{
for(child= n->getFirstChild(); child!=0; child=child->getNextSibling())
{
buildCLISubTree(child, depth++);
}
} else {
buildCLISubTree(n->getNextSibling(), depth++);
}
This means that when you invoke buildCLISubTree on an element, you will
drill down to its children, but you will never explore its siblings. The correct code is:
if (n->getNodeType() == DOMNode::ELEMENT_NODE )
{
for(child= n->getFirstChild(); child!=0; child=child->getNextSibling())
{
buildCLISubTree(child, depth++);
}
}
buildCLISubTree(n->getNextSibling(), depth++);
Alberto
P.S. Given the type of navigation you are doing, have you considered using
DOMNodeIterator?
>-----Original Message-----
>From: Nguyen, Vu [CAR:SI12:EXCH]
>Sent: Thursday, February 24, 2005 9:14 AM
>To: 'Alberto Massari (JIRA)'
>Subject: RE: [jira] Commented: (XERCESC-1355) DOMBuilder parse wrong
>node
>type
>
>Hi Alberto,
> I already tried but not working. So I will attach them to this
> email.
> I'd like to summarize here for more precision about potential
> error in my finding:
> 1) DOMBuilder did not build the node type correctly. It only set
> the all nodes to ELEMENT_NODE type.
> 2) DOMBuilder did not build the node correctly. For example,
> there is xml tag <help>"BulkInputService"</help> will result in two nodes:
>
> The first of type ELEMENT_NODE whose name is help; ofcourse
> there
> is no value for it.
> The second node that DOMBuilder inserted in of type TEXT_NODE,
> whose name is '#text' and its value is "BulkInputService" and added as a
> child node...
>
> From the CLITree.cc
> just look at: parseCLIModel(xml file);
> buildCLITree() and buildCLISubtree()
>
>My hunch is something to do with the code to setup the DOMBuilder, and
>the
>actual xml, though well-formed and validated, has something to screwup the
>dom parser ....
>
>Thanks a lot for your help,
>
>I stuck with this code for about few days and sitting at very tight
>schedule. I would be appreciate your help a lot. Looking forward for your
>assistance. Don't hesitate to email me if need anything...
>
>Thanks again
>Vu Nguyen
>(613) 768-2622
>
>One more thing, the xerces-c lib is the binary distribution downloaded
>from your website for LINUX.
>
>-----Original Message-----
>From: Alberto Massari (JIRA)
>[<mailto:xerces-c-dev@xml.apache.org>mailto:xerces-c-dev@xml.apache.org]
>Sent: Thursday, February 24, 2005 3:24 AM
>To: [EMAIL PROTECTED]
>Subject: [jira] Commented: (XERCESC-1355) DOMBuilder parse wrong node type
>
> [
> <http://issues.apache.org/jira/browse/XERCESC-1355?page=comments#action_59693>http://issues.apache.org/jira/browse/XERCESC-1355?page=comments#action_59693
> ]
>
>Alberto Massari commented on XERCESC-1355:
>------------------------------------------
>
>Could you attach the XML and the code you use to initialize the parser?
>It's hard for us to guess what's wrong here...
>
>Alberto
>
> > DOMBuilder parse wrong node type
> > --------------------------------
> >
> > Key: XERCESC-1355
> > URL:
> <http://issues.apache.org/jira/browse/XERCESC-1355>http://issues.apach
> e.org/jira/browse/XERCESC-1355
>
> > Project: Xerces-C++
> > Type: Bug
> > Components: DOM
> > Versions: 2.6.0
> > Environment: Linux gcc 3.2 Redhat Linux 8.0 3.2-7
> > Reporter: Vu Nguyen
>
> >
> > Basically I wrote a simple code (similar to DOMCount) to printout
> > the DOM tree content. I found out that all of the node type
> > (node.getNodeType()) always return 3(which is ELEMENT_NODE) while
> > it should sometimes return TEXT_NODE so I can obtain the value via
> > node.getNodeValue(). As the result, I always get NULL. Not sure if
> related to this thing, before setFeature(...whitespace = true), the
> DOMTree has a lot of bogus TEXT_NODEs. ==================== OUTPUT LOG
> ========================= Building CLI tree from data model ...
>
> > Element: command-groups
> > Attributes
> > ----------
> > id=_1
> > name=bulk-input
> > Rootnode has child: 1
> > Element: help = (null) with type 1
> > Element: command = (null) with type 1
> > Attributes
> > ----------
> > id=_1
> > name=show
> > Element: command = (null) with type 1
> > Attributes
> > ----------
> > id=_2
> > name=start
> > Node type == 1
> > Element: help and value (null) at 0
> > Node type == 3
> > Element: #text and value "BulkInputService" at 0
> > we done
> > =============================================================
> > Note that: help is a text node and expected to return some value.
> > Here the sample of xml and dtd.
>
>--
>This message is automatically generated by JIRA.
>-
>If you think it was sent incorrectly contact one of the administrators:
>
><http://issues.apache.org/jira/secure/Administrators.jspa>http://issues
>.apache.org/jira/secure/Administrators.jspa
>
>-
>If you want more information on JIRA, or have a bug to report see:
>
><http://www.atlassian.com/software/jira>http://www.atlassian.com/softwa
>re/jira
>
>
>
>
>
>
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: [EMAIL PROTECTED]
>For additional commands, e-mail: [EMAIL PROTECTED]
