asked on

XML parsing with Perl

hello, I begin to study XML parsing and I'm involved in 2 little problems.
my xml (a collection of bibliographic records) has the following structure:

<?xml version="1.0" encoding="UTF-8"?>

<collection xmlns="http://www.loc.gov/MARC21/slim">


<record>
<leader>-----nas-a2200000z--4500</leader>
<controlfield tag="008">140922uuuuuuuuuxx-uu-|------u|----|eng-d</controlfield>
<datafield tag="010" ind1="" ind2="">
<subfield code="a">01015589</subfield>
</datafield>
<datafield tag="245" ind1="" ind2="0">
<subfield code="a">Publishers weekly</subfield>
</datafield>
<datafield tag="260" ind1="" ind2="">
<subfield code="a">New York, NY</subfield>
<subfield code="b">Reed Business Information</subfield>
</datafield>
<datafield tag="022" ind1="" ind2="">
<subfield code="a">0000-0019</subfield>
</datafield>
<datafield tag="776" ind1="" ind2="">
<subfield code="x">2150-4008</subfield>
</datafield>
<datafield tag="090" ind1="" ind2="">
<subfield code="a">954921332001</subfield>
</datafield>
<datafield tag="866" ind1="" ind2="">
<subfield code="a">Available from 1997. </subfield>
<subfield code="s">1000000000001224</subfield>
<subfield code="t">1000000000000630</subfield>
<subfield code="x">EBSCOhost Business Source Complete:Full Text</subfield>
<subfield code="z">1000000000125212</subfield>
</datafield>
</record>

....more records...
</collection>

and I'd like to make 2 manipulations:

1) adding a single constant line (with constant content) in a precise position in <datafield tag="866" ind1="" ind2="">".
in one word, I.E., the
<datafield tag="866" ind1="" ind2="">
<subfield code="a">Available from 1997. </subfield>
<subfield code="s">1000000000001224</subfield>
<subfield code="t">1000000000000630</subfield>
<subfield code="x">EBSCOhost Business Source Complete:Full Text</subfield>
<subfield code="z">1000000000125212</subfield>
</datafield>

should be transformed to:

<datafield tag="866" ind1="" ind2="">
<subfield code="a">Available from 1997. </subfield>
****add the following line with "code" attribute in alphabetical order, after "a" and before "s"****
<subfield code="i">DEFAULT</subfield>
<subfield code="s">1000000000001224</subfield>
<subfield code="t">1000000000000630</subfield>
<subfield code="x">EBSCOhost Business Source Complete:Full Text</subfield>
<subfield code="z">10000000value 00125212</subfield>
</datafield>

2) find ALL and ONLY the records titles (it's the content of /collection/record/datafield tag='245'/subfield code='a') that have:

a) value of "/collection/record/datafield tag='866']/subfield code='x'" equal to "Elsevier SD Freedom Collection:Full Text"
b) the "/collection/record/datafield @tag='866'/subfield @code='a'" is totally absent, OR -if present- is empty.I.E.:

<datafield tag="866" ind1="" ind2="">
<subfield code="s">1000000000000992</subfield>
<subfield code="t">1000000000000473</subfield>
<subfield code="x">Elsevier SD Freedom Collection:Full Text</subfield>
<subfield code="z">1000000000043233</subfield>
</datafield>

OR

<datafield tag="866" ind1="" ind2="">
<subfield code="a"></subfield>
<subfield code="s">1000000000000992</subfield>
<subfield code="t">1000000000000473</subfield>
<subfield code="x">Elsevier SD Freedom Collection:Full Text</subfield>
<subfield code="z">1000000000043233</subfield>
</datafield>

thanks a lot for your reply,

fabianope

wilcoxon

A few questions...

Why does it matter if the codes are in alphabetic order? XML doesn't care.

Are you positive that the field could not also be <subfield code="a"/> which is also valid empty XML? It won't matter to a parser but, if someone tried to do it with regex or something, it would.

Personally, I'd look at either XML::Simple (which will definitely not preserve order) or XML::SAX. Both will write new files rather then editing the existing one in-place.

fabiano petrone

ASKER

Hi again, wilcoxon
Thanks a lot also for the interest on this issue.
here are the answers:
1) the xml is also for my friends librarians, and I'd like to give them a "ordered" code (for example, the various tags attribute corresponds to precise cataloging items, etc.).
2) you're right about this issue...the only thing I can say and that...I've never found <subfield code="a"/> in the file to process until now, but you're right...it can happens in the future so the point 2 should be re-formulated as follows:
*************************************************************************************************
2) find ALL and ONLY the records titles (it's the content of /collection/record/datafield tag='245'/subfield code='a') that have:

a) value of "/collection/record/datafield tag='866']/subfield code='x'" equal to "Elsevier SD Freedom Collection:Full Text"
b) the "/collection/record/datafield @tag='866'/subfield @code='a'" is totally absent, OR -if present- is empty.

c) a <subfield code="a"/>

thanks a lot,
fabiano

wilcoxon

Are ind1 and ind2 part of the "key" for datafield or is tag unique on its own?

Maintaining ordering is a lot more difficult so I'm going to ignore that requirement since it does not actually affect the XML. Are the codes already in alphabetic order? If not, it's even harder.

This should be the simplest code to accomplish the rest of it. If this doesn't exactly do what you want (such as changing elements to attributes or vice-versa), it will require playing around with options to XMLin and XMLout.

use strict;
use warnings;
use XML::Simple;
my $file = shift; # get file name from command line
my %opt = (KeepRoot => 1, ForceArray => [qw(datafield subfield record)],
           KeyAttr => { datafield => 'tag', subfield => 'code' },
          );
my $xml = XMLin($file, %opt) or die "could not parse $file: $!";
foreach my $rec (@{$xml->{collection}{record}}) {
    my $data = $rec->{datafield}{866};
    if (not $data->{ind1} and not $data->{ind2}) {
        $data->{subfield}{i} = 'DEFAULT';
    }
    next if ($data->{subfield}{a} or $data->{subfield}{x} ne 'Elsevier SD Freedom Collection:Full Text');
    print 'Title: ', $rec->{datafield}{245}{subfield}{a}, "\n";
}
print XMLout($xml, %opt);

Open in new window

fabiano petrone

ASKER

Hi, there
just launched:
acnp2.pl e-collection.20140922145816.xml-marc >acnp2.txt

(acnp2.pl your script, e-collection.20140922145816.xml-marc the input file, acnp2.txt the output file)

I obtain the error:
Unrecognised option: ForceArray at C:\Perl64\eg\xml\acnp2.pl line 17.

at line 17 I've the following code:
print XMLout($xml, %opt);

surely my fault... where :=))

Thanks a lot,

Fabiano

ASKER CERTIFIED SOLUTION

wilcoxon

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

fabiano petrone

ASKER

Hi there
tested & all OK!! :=))
Thanks a lot,
Fabiano