Why is XML::LibXML->toString() showing expected XML but findnodes() finding more XML?
I have this subroutine that is passed in a chunk of xml and tries to locate some elements (using XML::LibXML and XPath):
sub new_from_xml {
my $class = shift;
my ( $xml ) = @_;
my $self = {};
foreach (qw[ width height ]) {
$self->{$_} = $xml->findnodes("//$_")->[0]->textContent;
}
$self->{type} = $xml->findnodes("//type")->[0]->textContent;
$self->{url} = URI->new( $xml->findnodes("//url")->[0]->textContent );
return $class->new( $self );
}
It's get called from here:
sub new_from_xml {
my $class = shift;
my ( $xml ) = @_;
my $self = {};
foreach (qw[id caption orientation]) {
$self->{$_} = $xml->findnodes("//$_")->[0]->textContent;
}
$self->{alt} = $xml->findnodes('//htmlAlt')->[0]->textContent;
foreach my $instance ( $xml->findnodes("//instance") ) {
my $photo =
WWW::NewsReach::Photo::Instance->new_from_xml( $instance );
push @{$self->{instances}}, $photo;
}
return $class->new( $self );
}
What I expect is two <instance> ... </instance>
blocks that findnodes() returns, then when I loop over I pass the first instance on the first call and the second instance on the second call.
This is what I see in the debugger (I'm in the first subroutine above, WWW::NewsReach::Photo::Inst开发者_如何学运维ance->new_from_xml).
DB<13> x $xml->toString
0 '<instance><width>100</width><height>66</height><type>Small</type><url>http://pictures.newsreach.co.uk/liveimages/Decor-tips-for-guaranteed-unsecured-loans-users.jpg</url></instance>'
Okay good that's what I expected.
DB<14> x $xml->findnodes("//type")->[0]->textContent
0 'Medium'
Wait, what? That wasn't present in the XML shown from toString. Where did this come from?
DB<15> x $xml->findnodes("//type")
0 XML::LibXML::Element=SCALAR(0x101d03780)
-> 4334788352
1 XML::LibXML::Element=SCALAR(0x101cdd5c0)
-> 4334949168
Hmm, so there are two <type> ... </type>
elements.
DB<16> x $xml->toString;
0 '<instance><width>100</width><height>66</height><type>Small</type><url>http://pictures.newsreach.co.uk/liveimages/Decor-tips-for-guaranteed-unsecured-loans-users.jpg</url></instance>'
Umm, there's definitely only meant to be one <type> ... </type>
element. What is going on here?
Why is toString showing one <instance> ... </instance>
element but clearly the actual XML contains two <instance> ... </instance>
elements? Any help would be greatly appreciated.
//foo
will start searching from the document root. You want .//foo
精彩评论