Re: [xml] Child Traversing loop runs twice

[Csaba Raduly <rcsaba gmail com>]

On Wed, Feb 17, 2010 at 8:12 AM, andrew james
<andrew systemssingular com> wrote:
> siddharth sehgal wrote:
> > Hi
> >
> > I am new to libxml2 and intend to use it in my application which will run
> > on embedded windows/ linux. I was trying to get hold of some common xml
> > parsing routines and there is an interesting observation which I cannot
> > resolve. although the xml I am parsing is a bit more elaborative but the
> > observation persists on simple xml like this also.
> >
> > <?xml version="1.0" ?>
> > <CAST>
> >    <a />
> >    <b />
> >    <c />
> > </CAST>
> >
> >
> > xmlDocPtr doc;
> > xmlNodePtr topcur;
> > .
> > .
> > doc = xmlParseFile(ansiSTR);
> > .
> > .
> > topcur = doc->children;//xmlDocGetRootElement(doc);//At <CAST>
> > .
> > .
> >   topcur=topcur->children;
> > while(topcur != NULL){
> >    printf("%s\n",topcur->name);
> >    topcur=topcur->next;
> > }
> >
> > Output expected is:
> > a
> > b
> > c
> >
> > Actual Output:
> > text
> > a
> > text
> > b
> > text
> > c
> > text
> >
> > The loop for child traversing runs twice. I cant figure this out . Any
> > suggestions will be deeply appreciated.
> >
> > thanks
> > Sid
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > xml mailing list, project page  http://xmlsoft.org/
> > xml gnome org
> > http://mail.gnome.org/mailman/listinfo/xml
>
>
> Though I am inexperienced with the specifics of libxml2, I have used other
> parsers for html (xml), that also read what to us is like extra text nodes.
>
> Anyone may want to correct this if I am wrong,
>
> For some reason, someone else may know the reason, some parsers read text
> nodes between specific elements (is it the new line?).
>
> for this particularity of parsers, the suggestion is that you add some
> conditions in the code to evaluate node name.

No no, there's no need for that. xmlNode contains a "type" member.
Its values are from the xmlElementType enum. Just compare with
XML_ELEMENT_NODE or XML_TEXT_NODE.
See tree.h

> xmlStrcmp is almost illogical for a human to read because it
> evaluates to 0 when strings are the same

That's because xmlStrcmp is not a test for equality, but a comparison,
so it has to cater for three possible situations:
1. string1 is lexicographically "less" than string 2 (returns a value
less than zero)
2. strin1 is the same as string2 (returns 0)
3. string1 is lexicographically "greater" than string 2 (returns a
value greater than zero)

Hope this helps,
Csaba
-- 
Life is complex, with real and imaginary parts