Perl - Modify the nested XML tags

[Murari]

I have a complex xml which I need to parse in PERL to output to different format input xml:

    <?xml version="1.0" encoding="UTF-8"?>
     <response>
     <result name="response" numFound="58582" start="0"> 
     <doc>
<str name="body">Have a great time at this park!</str><int name="is_pa
+rk_id">317851</int><str name="ss_image_thumb_small"/>
    <str name="title">Playground</str>    
     </doc>
<doc>
<str name="body">Have a great time at this park!</str><int name="is_pa
+rk_id">317851</int><str name="ss_image_thumb_small"/>
    <str name="title">Playground</str>    
     </doc>
     </result>
     </response>

and the required output format is:

     <?xml version="1.0" encoding="iso-8859-1"?>
     <Feed>
     <Products>
     <Product>
<Description>Have a great time at this park!</Description><ExternalId>
+PF317851</ExternalId><ImageUrl/>
     <Name>Playground</Name>
</Product>
<Product>
<Description>Have a great time at this park!</Description><ExternalId>
+PF317851</ExternalId><ImageUrl/>
     <Name>Playground</Name>
</Product>
     </Products>
     </Feed>

<str name="body"> should be treated as description, <int name="is_park_id"> should be treated as <ExternalId> and <str name="title"> should be treated as <Name>. Could you help in formatting the XML to new format?

[tobyink]

use strict;
use warnings;
use XML::LibXML 2;
use XML::LibXML::PrettyPrint;

my $xml = XML::LibXML::->load_xml(string => <<'INPUT');
<?xml version="1.0" encoding="UTF-8"?>
<response>
    <result name="response" numFound="58582" start="0">
        <doc>
            <str name="body">Have a great time at this park!</str>
            <int name="is_park_id">317851</int>
            <str name="ss_image_thumb_small"/>
            <str name="title">Playground</str>
        </doc>
    </result>
</response>
INPUT

my %changes = (
    '//response'                          => sub { $_->setNodeName('Fe
+ed') },
    '//result'                            => sub { $_->setNodeName('Pr
+oducts'), %$_ = (); },
    '//doc'                               => sub { $_->setNodeName('Pr
+oduct'), %$_ = (); },
    '//str[@name="body"]'                 => sub { $_->setNodeName('De
+scription'), %$_ = (); },
    '//int[@name="is_park_id"]'           => sub { $_->setNodeName('Ex
+ternalId'), %$_ = (); },
    '//str[@name="ss_image_thumb_small"]' => sub { $_->setNodeName('Im
+ageUrl'), %$_ = (); },
    '//str[@name="title"]'                => sub { $_->setNodeName('Na
+me'), %$_ = (); },
);

while (my ($xpath, $processor) = each %changes)
{
    $xml->findnodes($xpath)->foreach($processor);
}

print(
    XML::LibXML::PrettyPrint::
        -> new(element => {
            compact => [qw/ Description ExternalId ImageUrl Name /],
        })
        -> pretty_print($xml)
        -> toString
);

[Murari]

I have the these libraries installed in my machine, couldn't use XML::LibXML. XML::NamespaceSupport, XML::Parser, XML::RegExp, XML::SAX, XML::Simple My requirement is multi-register document, could you help in the solution for the input nested XML?

[Anonymous Monk]

use XML::Twig, the code won't look much different, XML::Twig :) it comes with many examples/tutorials

[choroba]

There is how I would do it using XML::XSH2, a wrapper around XML::LibXML:

open 1.xml ;

rename Feed        response ;               cd Feed ;
rename Products    result ;                 cd Products ;
rename Product     doc ;                    cd Product ;
rename Description str[@name="body"] ;
rename ExternalId  int[@name="is_park_id"] ;
rename ImageUrl    str[@name="ss_image_thumb_small"] ;
rename Name        str[@name="title"] ;

insert text "PF" prepend ExternalId ;

delete //*/@* ;

save :b ;

[sundialsvc4]

It may also be possible to express this using XSLT technology, entirely apart from Perl.   For an idea of what sort of things this technology can do, without programming, consider the interactive Periodic Table of the Elements example on the Internet.   (With your free hand, hold your lower jaw in place before you click on that link ...)   There is no programming being used to build the entire HTML of that display, and only a slight amount of JavaScript being used to respond to clicks and mouse-overs.   Your browser is the one doing all the work, on the fly, to produce what you see.   If your problem can be described as “fetch elements from an XML data source and make something else out of it,” this might be the cat’s meow.   A number of Perl libraries provide an industry standard XSLT capability, if it turns out that you need it.

[tobyink]

"without programming"

Without any programming whatsoever... apart from over 4000 lines of Javascript, and an unspecified amount (it doesn't all seem to be publicly available, so I can't count the lines) of XSLT, which is itself a Turing-complete declarative programming language.