Having trouble with Perl script that converts XML to hash - perl

Question

I have a Perl script to convert the XML file below into a hash:
<university>
<name>svu</name>
<location>ravru</location>
<branch>
<electronics>
<student name="xxx" number="12">
<semester number="1"subjects="7" rank="2"/>
</student>
<student name="xxx" number="15">
<semester number="1" subjects="7" rank="10"/>
<semester number="2" subjects="4" rank="1"/>
</student>
<student name="xxx" number="16">
<semester number="1"subjects="7" rank="2"/>
<semester number="2"subjects="4" rank="2"/>
</student>
</electronics>
</branch>
</university>.
.
.
.
.
.
<data>
<student name="msr" number="1" branch="computers" />
<student name="ksr" number="2" branch="electronics" />
<student name="lsr" number="3" branch="EEE" />
<student name="csr" number="4" branch="IT" />
<student name="msr" number="5" branch="MEC" />
<student name="ssr" number="6" branch="computers" />
<student name="msr" number="1" branch="CIV" />
.............................
..............................
.....................
</data>
How can I create a hash table for the data elements, with the name and number as the key and branch is the value in that hash. I need this because some students have the same name and some students have same number.
By using this hash key I have to search in the university node for student if found and print the branch name of each student.
I written some script in XML::Simple but am not able to create a hash.
#!/usr/bin/perl
use warnings;
use strict;
use Data::Dumper;
use XML::Simple;
my $xml = new XML::Simple;
my $data = $xml->XMLin("data.xml", forcearray => [ 'student' , 'semister' ],
KeyAttr => { student => "+Name" } );
print Dumper($data);
by using data dumper I am printing hole xml information. but I need to print only Data Node elements only please help me how to do this.

Answer 1

I would probably write my own XML::Parser handler to combine attributes into key values (if that's something supported by XML::Simple I couldn't find it in the docs). This example should get you started:
#!/usr/bin/perl
use strict;
use warnings;
use XML::Parser;
use Data::Dumper;
my %hash;
sub tag_start { my ($expat, $tagname) = (shift, shift);
# attributes are now in #_
my %a = grep { $_=$_=>shift } #_; # attribute hash for this tag
my $context = join('/',$expat->context()) || '';
if ($context eq 'xml/data') {
if ($tagname eq 'student') {
push #{($hash{"$a{name}:$a{number}"}||=[])}, $a{branch};
}
} elsif ($context eq ...) {
...
}
}
my $p = new XML::Parser(Handlers => { Start=>\&tag_start });
$p->parsefile('file.xml');
print Dumper \%hash;
Note that to get this to work I had to clean up your XML a bit by enclosing it in an <xml> tag and adding some missing spaces:
<xml>
<university>
<name>svu</name>
<location>ravru</location>
<branch>
<electronics>
<student name="xxx" number="12">
<semester number="1" subjects="7" rank="2"/>
</student>
<student name="xxx" number="15">
<semester number="1" subjects="7" rank="10"/>
<semester number="2" subjects="4" rank="1"/>
</student>
<student name="xxx" number="16">
<semester number="1" subjects="7" rank="2"/>
<semester number="2" subjects="4" rank="2"/>
</student>
</electronics>
</branch>
</university>
<data>
<student name="msr" number="1" branch="computers" />
<student name="ksr" number="2" branch="electronics" />
<student name="lsr" number="3" branch="EEE" />
<student name="csr" number="4" branch="IT" />
<student name="msr" number="5" branch="MEC" />
<student name="ssr" number="6" branch="computers" />
<student name="msr" number="1" branch="CIV" />
</data>
</xml>
Result:
$VAR1 = {
'ksr:2' => [
'electronics'
],
'msr:1' => [
'computers',
'CIV'
],
'csr:4' => [
'IT'
],
'ssr:6' => [
'computers'
],
'msr:5' => [
'MEC'
],
'lsr:3' => [
'EEE'
]
};

Answer 2

There is no need to use XML::Simple and XML::Fast together. Both perform essentially the same thing.
Invoking multiple XML parsers for the same functionality invites trouble in the form of undesired behavior, code that should work but doesn't and debugging that will leave you holding your hands in your head because identically-named methods are treading on one another's toes.
I'd stick with XML::Fast for this case:
use strict;
use warnings;
use XML::Fast;
my $data = xml2hash 'data.xml', array => [ 'student', 'semester' ];
Even if the structure is not exactly the desired one, $data can easily be post-processed and seasoned to taste (it is a data structure after all).