XML::LibXML Line Ending (whitespace) Problem.

Question

HI, I am parsing an XML file using LibXML in Perl. The problem that I have is the ending characters (whitespace) is treated as a text node. For instance, given an input like the following

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE books [
    <!ELEMENT title  (#PCDATA)>
    <!ELEMENT author (#PCDATA)>
    <!ELEMENT year   (#PCDATA)>
    <!ELEMENT price  (#PCDATA)>
    <!ELEMENT book   (title, author, year, price)>
    <!ELEMENT books  (book*)>
]>
<books>
<book>
<title>Everyday Italian</title>
<author>Giada De Laurentiis</author>
<year>2005</year>
<price>30.00</price>
</book>
</books>

The parser thinks that the number of child of node "books" is 3, they are:

• text node (containing the char between <books> and <book>)
• element node of <book>
• text node (containing the char between </book> and </books>)

Question is how do I tell LibXML to ignore whitespaces? I tried with no_blanks (that is $parser = XML::LibXML->new(no_blanks => 1) when construction the parser) but it seems that it has no effect.

Thanks in advance

Answer 1

XML::LibXML::Parser has $parser->keep_blanks(0); . It's supposed to do the opposite of no_blanks - see if that works

Answer 2

Strictly-speaking, XML::LibXML is doing the correct thing... there are three child-nodes of the <books> element. The question is, how are you parsing the content, and why is this a problem?

Assuming you've parsed your content and assigned the result to $document, you now have an instance of the XML::LibXML::Document class. Using this, you can get the <books> element by using documentElement():

$books = $document->documentElement();

This returns an instance of XML::LibXML::Element. From this, you can get just the <book> child-elements using getChildrenByTagName():

@book_elements = $books->getChildrenByTagName('book');

Does this help?