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

Dear Monks,

I am doing a big project using mainly XML::Twig. Today during testing i come across a strange problem. I tried in several ways but i could not able to find the why the problem comes. So i need some help here from you guys.

<?xml version="1.0"?>
<article>
<label>here label for testing</label>
<fig id="fig001" position="float"><label>Fig. 1.</label><caption><p>fi
+rst</p></caption></fig>
<fig id="fig002" position="float"><label>Fig. 2.</label><caption><p>se
+cond</p></caption></fig>
<fig id="fig003" position="float"><label>Fig. 3.</label><caption><p>th
+ird</p></caption></fig>
<fig id="fig004" position="float"><label>Fig. 4.</label><caption><p>fo
+urth</p></caption></fig>
<fig id="fig005" position="float"><label>Fig. 5.</label><caption><p>fi
+fth</p></caption></fig>
<fig id="fig006" position="float"><label>Fig. 6.</label><caption><p>si
+xth</p></caption></fig>
</article>

[download]

The above is the sample xml file. There i need to cut the 'label' element and place it as a first_child for the 'caption'. So i tried as shown below. The below code is sample code from my original code.

use strict;
use XML::Twig;

my $twig = new XML::Twig(                    
                                        
);

$twig->parsefile('1.xml');

my $count = $twig->get_xpath('//fig//label'); 

for my $c (0..($count-1))
{

    my $t = $twig->get_xpath('//fig//label', $c);            #//fig/la
+bel -here problem comes            
                                                    
    my $pl = $twig->get_xpath('//fig//caption', $c);
    
    #my $cc = $t->print;# print for testing
    
    #my $aa = $twig->print; #print for testing
    
    #print "$cc\n\n$c\n**$aa\n";                                      
+                  
                                
    my $cut = $t->cut ; 
                            
    $cut->paste('first_child', $pl) ;                                 
+               

}

$twig->print;

[download]

I got the exact output as shown below by running the above code.

Correct Output: XPath Expression - "//fig//label"
---------------
<?xml version="1.0"?>
<article>
<label>here label for testing</label>
<fig id="fig001" position="float"><caption><label>Fig. 1.</label><p>fi
+rst</p></caption></fig>
<fig id="fig002" position="float"><caption><label>Fig. 2.</label><p>se
+cond</p></caption></fig>
<fig id="fig003" position="float"><caption><label>Fig. 3.</label><p>th
+ird</p></caption></fig>
<fig id="fig004" position="float"><caption><label>Fig. 4.</label><p>fo
+urth</p></caption></fig>
<fig id="fig005" position="float"><caption><label>Fig. 5.</label><p>fi
+fth</p></caption></fig>
<fig id="fig006" position="float"><caption><label>Fig. 6.</label><p>si
+xth</p></caption></fig>
</article>

[download]

My question comes here, I am getting the exact output when i give the XPath Expression as '//fig//label'. But if i change the XPath Expression as '//fig/label', i am getting error. To identify the problem i checked by printing in the loop every time and the final output i got is as shown below. The XPath Expression is correct, because it is very first child, but i am getting error and not able to get exact output. Where i am going wrong?

Wrong output:  XPath Expression - "//fig/label"
-------------
<article>
<label>here label for testing</label>
<fig id="fig001" position="float"><caption><label>Fig. 1.</label><p>fi
+rst</p></caption></fig>
<fig id="fig002" position="float"><label>Fig. 2.</label><caption><labe
+l>Fig. 3.</label><p>second</p></caption></fig>
<fig id="fig003" position="float"><caption><p>third</p></caption></fig
+>
<fig id="fig004" position="float"><label>Fig. 4.</label><caption><p>fo
+urth</p></caption></fig>
<fig id="fig005" position="float"><label>Fig. 5.</label><caption><p>fi
+fth</p></caption></fig>
<fig id="fig006" position="float"><label>Fig. 6.</label><caption><p>si
+xth</p></caption></fig>
</article>

[download]

Regards,
Prasad

Replies are listed 'Best First'.
Re: XML::Twig 'cut' and 'paste' Question
by Tanktalus (Canon) on Jul 28, 2006 at 19:19 UTC

    I'm wondering if you're complicating things a bit. Rather than count, and then loop, try simply looping. I have a suspicion that your cuts and pastes are changing the number of matches that //fig/label sees - once you've changed the first fig/label, it's no longer a fig/label. Try this: 

    #!/usr/bin/perl

use strict;
use warnings;
use XML::Twig;

my $twig = new XML::Twig(
                         pretty_print => 'record_c'
                        );

$twig->parsefile('1.xml');

for my $fig (
             $twig->get_xpath('//fig')
            )
{
    my $label   = $fig->get_xpath('./label', 0);
    my $caption = $fig->get_xpath('./caption', 0);

    next unless $label and $caption;

    $label->cut();
    $label->paste('first_child', $caption);

}

$twig->print;

    [download]
    That seems to do the trick. Also note the better variable names ;-)

[reply]
[d/l]
Re: XML::Twig 'cut' and 'paste' Question
by Ieronim (Friar) on Jul 28, 2006 at 19:30 UTC
    You modify the document with every cut'n'paste—the <label> are direct children of the <fig> element the first time, but then the first label becomes a grandchild of <fig>! Of course, you get the error Can't call method "cut" on an undefined value when you try to get the $c'th element from the list containing only one label.

    This error appeared because you chose an extremely ineffective way—you call get_xpath on every element in the list instead of fetching all elements once.

    The corected variant of your code:

    Read more... (2 kB)

    The more efficient and clean way resulting in the same output:

    #!/usr/bin/perl

use warnings;
use strict;
use XML::Twig;

my $twig = new XML::Twig(                   
pretty_print => 'indented',                                     
);

$twig->parse(<<'XML');
<?xml version="1.0"?>
<article>
<label>here label for testing</label>
<fig id="fig001" position="float"><label>Fig. 1.</label><caption><p>fi
+rst</p></caption></fig>
<fig id="fig002" position="float"><label>Fig. 2.</label><caption><p>se
+cond</p></caption></fig>
<fig id="fig003" position="float"><label>Fig. 3.</label><caption><p>th
+ird</p></caption></fig>
<fig id="fig004" position="float"><label>Fig. 4.</label><caption><p>fo
+urth</p></caption></fig>
<fig id="fig005" position="float"><label>Fig. 5.</label><caption><p>fi
+fth</p></caption></fig>
<fig id="fig006" position="float"><label>Fig. 6.</label><caption><p>si
+xth</p></caption></fig>
</article>
XML

my @labels = $twig->get_xpath('/article/fig/label'); 
my @captions = $twig->get_xpath('/article/fig/caption');
for my $i (0..$#labels) {
    $labels[$i]->move('first_child', $captions[$i]);
}

$twig->print;

    [download]
    But i myself would prefer to let twig_handlers do the job: 
    #!/usr/bin/perl

use warnings;
use strict;
use XML::Twig;

my $twig = new XML::Twig(
    twig_handlers => {
        '/article/fig' => sub {
            my $label = $_->first_child('label') or return;
            my $caption = $_->first_child('caption') or return;
            $label->move('first_child', $caption);
            $_->flush;
        }
},
    pretty_print => 'indented',                                     
);

$twig->parse(<<'XML');
<?xml version="1.0"?>
<article>
<label>here label for testing</label>
<fig id="fig001" position="float"><label>Fig. 1.</label><caption><p>fi
+rst</p></caption></fig>
<fig id="fig002" position="float"><label>Fig. 2.</label><caption><p>se
+cond</p></caption></fig>
<fig id="fig003" position="float"><label>Fig. 3.</label><caption><p>th
+ird</p></caption></fig>
<fig id="fig004" position="float"><label>Fig. 4.</label><caption><p>fo
+urth</p></caption></fig>
<fig id="fig005" position="float"><label>Fig. 5.</label><caption><p>fi
+fth</p></caption></fig>
<fig id="fig006" position="float"><label>Fig. 6.</label><caption><p>si
+xth</p></caption></fig>
</article>
XML

$twig->print;

    [download]
         s;;Just-me-not-h-Ni-m-P-Ni-lm-I-ar-O-Ni;;tr?IerONim-?HAcker ?d;print
[reply]
[d/l]
[select]
Re: XML::Twig 'cut' and 'paste' Question
by GrandFather (Saint) on Jul 28, 2006 at 19:43 UTC

    In XPATH syntax / specifies an absolute path where as // specefies a relative path. /article/fig/label would work.

    See http://www.w3.org/TR/xpath for the full XPATH spec.

    DWIM is Perl's answer to Gödel
[reply]
[d/l]
      You are not fully right. Directly from the document you cite:
      //para selects all the para descendants of the document root and thus selects all para elements in the same document as the context node
      So //fig/label selects all 'label' childern of all 'fig' elements in the document.
           s;;Just-me-not-h-Ni-m-P-Ni-lm-I-ar-O-Ni;;tr?IerONim-?HAcker ?d;print
[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom