Tryag File Manager

//proc/self/root/usr/share/doc/perl-XML-DOM-1.44/FAQ.xml

This document contains answers to common questions.
I need to add a lot more stuff and make sure it is valid XML.
(The format may change, but since it's XML, it's easy to transform...)

</description>

<title>libxml-enno FAQ</title>

<section>
  <title>Known Build/Make Problems</title>

'unrecognized pod directive in ...: head3' warnings in 'make install'

</question>
   <answer>

These are caused by pod2man. (A bug in pod2man, IMHO.) 
It doesn't seem to recognize '=head3' in pod files. Ignore the warnings.

</answer>
  </topic>

t/out/dc_attr3.err missing from distribution

</question>
   <answer>

Create the file. It should be empty (i.e. size 0.)
WinZIP sometimes doesn't add or extract the file when it has size 0.
I've only seen this on Windows.

</answer>
  </topic>

<section>Getting Started</section>

<topic>
   <short>Element node has too many children.</short>
   <question>
<![CDATA[

I have a simple XML file:
--------------------------------------------------
<?xml version='1.0'?>

<istCH>
  <Image>NAME</Image>
</istCH>
--------------------------------------------------

and the perl code:
--------------------------------------------------

my $parser = new XML::DOM::Parser;
my $doc = $parser->parsefile("test.xml");
my $root = $doc->getDocumentElement();

my $i = 0;
for my $kids ($root->getChildNodes()) {
  print STDERR "  Child $i is $kids\n";
  print STDERR "             name  ", $kids->getNodeName(), "\n";
  print STDERR "             type  ", $kids->getNodeType(), "\n";
  print STDERR "             value ", $kids->getNodeValue(), "\n";
  $i++;
}
--------------------------------------------------

And I found that my root node has 3 children, where I thought it should have one:

1) a text node having as value a cariage return and two spaces

2) an element node named Image with no value (in fact the value is in a
text child of that element)

3) an other text child having as value a carriage return

I thought the XML root would have only one child (Image) with the value 'NAME'.

]]>
   </question>
   <answer>

That's what an XML processor is supposed to do: all characters, including
whitespace (outside of markup) are reported to the application. Your DOM
script should then decide what to do with unnecessary whitespaces.

You can use a PerlSAX filter (like XML::Handler::DetectWS) to filter out the
whitespace at parse time, before it reaches the DOM document.
   </answer>
  </topic>
  
  <topic>
   <question>

How do I move (not copy) parts of one XML::DOM::Document (or Element) to another?

</question>
   <answer>

You can use cloneNode() to copy a subtree or use removeNode()
to cut the subtree out of document A.
Then use setOwnerDocument($docB) on the subtree and use
insertNode() or appendNode() to add it to document B.
(Note that setOwnerNode is not part of the DOM Level 1 specification, so
your code won't be portable to other DOM implementations.)

One problem: if you have attributes in the subtree with
defaulted values (i.e. they were not specified in document A,
but XML::Parser (expat) generated them because an ATTLIST
declaration specified a default value for that attribute),
the attributes will still point to the default values in 
document A.

I haven't found a good solution for this problem yet,
because users may want different things in different situations.
Any thoughts are welcome.

</answer>
  </topic>