andrewheiss has asked for the wisdom of the Perl Monks concerning the following question:

I'm using XPath to parse an XML file of blog posts and comments. Here's an example of the file structure:

<entry>
  <title>Title of the post</title>
  <content>Content of the post</content>
  <category term="blog"/>
  <category term="label"/>
  <category term="another label"/>
  <category term="blah"/>
  ...
</entry>

[download]

The first category element indicates the type of entry--whether or not the entry is a blog post or a comment. Any subsequent category elements indicate the blog post tags.

I can easily get the first instance of category like so:

foreach my $entry (reverse($xc->findnodes('//post:entry'))) {
    my $type =  $xc->findvalue('./post:category[1]/@term', $entry);
    ...
}

[download]

And I can get the subsequent values by referencing post:category[2]/@term, or any other number.

I can even get all subsequent values using this xpath:

my $category = $xc->findvalue('./post:category[position()>1]/@term', $entry);

Unfortunately, though, when I use position()>1 it sticks all the values in one variable, like "labelanother labelblah". What's the best way to keep the values separate? when retrieving them with xpath?

Replies are listed 'Best First'.
Re: Loop through multiple elements with same name in XPath
by ikegami (Patriarch) on May 30, 2009 at 16:58 UTC
    You need findnodes if you want to return multiple results 
    for my $attr_node (
    reverse
    $xc->findnodes('//post:entry/post:category/@term')
) {
    my $attr_val = $attr_node->getValue();
    ...
}

    [download]

    Or maybe you want 

    for my $entry_node ( reverse $xc->findnodes('//post:entry') ) {
    for my $attr_node ( $xc->findnodes('post:category/@term', $entry_n
+ode) ) {
        my $attr_val = $attr_node->getValue();
        ...
    }
    ...
}

    [download]
[reply]
[d/l]
[select]
Re: Loop through multiple elements with same name in XPath
by mirod (Canon) on May 30, 2009 at 09:01 UTC

    Use findnodes and loop through the nodes. Something like (untested):

    my $category = join " ", map { $_->getNodeText } $xc->findnodes('./post:category[position()>1]/@term', $entry);

    updated:really use findnodes (thanks ikegami)

[reply]
[d/l]
[select]
Re: Loop through multiple elements with same name in XPath
by Anonymous Monk on May 30, 2009 at 09:11 UTC
[reply]
Re: Loop through multiple elements with same name in XPath
by Jenda (Abbot) on May 30, 2009 at 20:51 UTC

    I know I'm getting tiresome ... 

    my $parser = XML::Rules->new(
  rules => {
    _default => 'content',
    category => 'content array',
    entry => 'array no content',
    post => 'pass no content',
  }
);
my $data = $parser->parse($the_xml);
foreach my $entry (@$data) {
  print "Title: $entry->{title}\nContent: $entry->{content}\n";
  if ($entry->{category)[0] eq 'blog') {
    print "It's a blog post\n\n"; 
  }
  ...

    [download]
    or 
    my $parser = XML::Rules->new(
  rules => {
    _default => 'content',
    category => 'content array',
    entry => sub {
      my ($tag,$attr) = @_;
      delete $attr->{_content};
      $attr->{type} = shift(@{$attr->{category}});
      return '@entry' => $attr;
    },
    post => 'pass no content',
  }
);
my $data = $parser->parse($the_xml);
foreach my $entry (@$data) {
  print "Title: $entry->{title}\nContent: $entry->{content}\n";
  if ($entry->{type) eq 'blog') {
    print "It's a blog post with labels: ".join(', ', @{$entry->{categ
+ory}})."\n\n"; 
  }
  ...

    [download]
    or if you like you can handle each the <entry> within the sub specified in the rules and forget the value then and save memory.

    Jenda
    Enoch was right!
    Enjoy the last years of Rome.

[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom