I am trying to parse some simple XML in XML::LibXML and I have struck a problem. The domain is WebDAV.
A PROPFIND request from a user can send some very simple XML, but in a variety of semantically identical but syntactically different forms.
For a simple example:
<?xml version="1.0" encoding="utf-8" ?>
<propfind xmlns="DAV:">
<propname/>
</propfind>
<?xml version="1.0" encoding="utf-8" ?>
<D:propfind xmlns:D="DAV:">
<D:propname/>
</D:propfind>
TO get the 'propfind' node valid XPaths are
//propname
//DAV:propname
/propfind/propname
/DAV:propfind/DAV:propname
surely? But none of them work with XPathContext except in one special case (below) where there is no need for it anyway.
Below are my test code and the results. In it I register 'D' as a prefix for 'DAV:' and so using a XPath with 'D' as prefix works where the XML uses it too. But that is not good enough for where the XML uses 'DAV:' as a default namespace 'D' as a prefix should work too. Surely?
#!/usr/bin/perl -w
use strict;
use XML::LibXML;
sub testfn {
my $content = shift or die;
my $xpath = shift or die;
$|++;
my $parser = XML::LibXML->new();
my $dom = $parser->parse_string($content);
my @propfind = ();
@propfind = $dom->findnodes($xpath);
print ref($dom)."::findnodes('$xpath') (NOT XPathContext) ".
scalar(@propfind)." nodes\n";
my $xc = XML::LibXML::XPathContext->new($dom);
$xc->registerNs('D', 'DAV:');
@propfind = $xc->findnodes($xpath);
print ref($xc)."::findnodes('$xpath') ".scalar(@propfind)." nodes\
+n";
}
# Example from RFC4918
my $txt1 = '<?xml version="1.0" encoding="utf-8" ?>
<D:propfind xmlns:D="DAV:">
<D:propname/>
</D:propfind>
';
my $txt2 = '<?xml version="1.0" encoding="utf-8" ?>
<propfind xmlns="DAV:">
<propname/>
</propfind>
';
my $xpath1 = '/DAV:propfind/DAV:propname';
my $xpath2 = '/propfind/propname';
my $xpath3 = '/D:propfind/D:propname';
print "\$txt1\n".$txt1."\n";
print "\$txt1 \$xpath1 \n";
eval {
&testfn($txt1, $xpath1);
};
if($@){
print "Failed \$xpath $xpath1\n";
}
print "\$txt1 \$xpath2 \n";
eval{
&testfn($txt1, $xpath2);
};
if($@){
print "Failed \$xpath $xpath2\n";
}
print "\$txt1 \$xpath3 \n";
eval{
&testfn($txt1, $xpath3);
};
if($@){
print "Failed \$xpath $xpath3\n";
}
print "\$txt2\n".$txt2."\n";
print "\$txt2 \$xpath1 \n";
eval{
&testfn($txt2, $xpath1);
};
if($@){
print "Failed \$xpath $xpath1\n";
}
print "\$txt2 \$xpath2 \n";
eval {
&testfn($txt2, $xpath2);
};
if($@){
print "Failed \$xpath $xpath2\n";
}
print "\$txt2 \$xpath3 \n";
eval {
&testfn($txt2, $xpath3);
};
if($@){
print "Failed \$xpath $xpath3\n";
}
The results:
$txt1
<?xml version="1.0" encoding="utf-8" ?>
<D:propfind xmlns:D="DAV:">
<D:propname/>
</D:propfind>
$txt1 $xpath1
Failed $xpath /DAV:propfind/DAV:propname
$txt1 $xpath2
XML::LibXML::Document::findnodes('/propfind/propname') (NOT XPathConte
+xt) 0 nodes
XML::LibXML::XPathContext::findnodes('/propfind/propname') 0 nodes
$txt1 $xpath3
XML::LibXML::Document::findnodes('/D:propfind/D:propname') (NOT XPathC
+ontext) 1 nodes
XML::LibXML::XPathContext::findnodes('/D:propfind/D:propname') 1 nodes
$txt2
<?xml version="1.0" encoding="utf-8" ?>
<propfind xmlns="DAV:">
<propname/>
</propfind>
$txt2 $xpath1
Failed $xpath /DAV:propfind/DAV:propname
$txt2 $xpath2
XML::LibXML::Document::findnodes('/propfind/propname') (NOT XPathConte
+xt) 0 nodes
XML::LibXML::XPathContext::findnodes('/propfind/propname') 0 nodes
$txt2 $xpath3
Failed $xpath /D:propfind/D:propname
