bobf has asked for the wisdom of the Perl Monks concerning the following question:

I am trying to define handlers for XML::Twig that will allow me to parse out different sections of an XMI document (created by exporting a UML model from Enterprise Architect) and process them separately, but I can't seem to grok the syntax. A very simplified version of the document looks like this: 

<XMI xmlns:UML="omg.org/UML1.3" xmi.version="1.1" timestamp="2006-08-0
+4 16:29:41">
  <XMI.header>
    <XMI.documentation>
      <XMI.exporter>Enterprise Architect</XMI.exporter>
      <XMI.exporterVersion>4.1RR</XMI.exporterVersion>
    </XMI.documentation>
  </XMI.header>
  <XMI.content>
    <UML:Model name="EA Model" xmi.id="MX_EAID_...">
      <UML:Namespace.ownedElement xmi.id="MX_EAID_...">

<!-- note the UML:Class tag on the next line -->
        <UML:Class name="EARootClass" xmi.id="EAID_..." />
        <UML:Package name="Logical View" xmi.id="EAPK...">
          <UML:Namespace.ownedElement xmi.id="EAPK...">
            <UML:Package name="Logical Model" xmi.id="EAPK...">
              <UML:Namespace.ownedElement xmi.id="EAPK...">

<!-- many UML:Package and UML:Namespace tags removed for brevity -->

<!-- I want to pull out the following UML:Class block -->
                            <UML:Class name="DataType" xmi.id="EAID...
+">
                              <UML:Classifier.feature xmi.id="EAID..."
+>

<!-- I want to pull out the following UML:Attribute blocks -->
                                <UML:Attribute name="dataTypeId">
                                </UML:Attribute>
                                <UML:Attribute name="name">
                                </UML:Attribute>
                              </UML:Classifier.feature>
                            </UML:Class>

<!-- I want to pull out the following UML:Association blocks -->
                            <UML:Association xmi.id="EAID...">
                            </UML:Association>
                            <UML:Association xmi.id="EAID...">
                            </UML:Association>

<!-- many UML::Class and UML::Association blocks removed (I want these
+, too) -->
<!-- miscellaneous blocks removed -->

              </UML:Namespace.ownedElement>
            </UML:Package>
          </UML:Namespace.ownedElement>
        </UML:Package>
        <UML:DataType xmi.id="eaxmiid3" />
        <UML:DataType xmi.id="eaxmiid1" />
      </UML:Namespace.ownedElement>
    </UML:Model>

<!-- I want to process each of the UML::TaggedValue elements -->
    <UML:TaggedValue tag="complexity" />
    <UML:TaggedValue tag="ea_stype" />

<!-- many UML::TaggedValue tags removed -->

  </XMI.content>
  <XMI.extensions xmi.extender="Enterprise Architect 2.5" />
</XMI>

[download]

I tried a few things, but the closest I came to getting what I want is the following: 

use strict;
use warnings;

use Data::Dumper;
use XML::Twig;

my $twig = XML::Twig->new( twig_roots => { 'UML:Class' => \&uml_class 
+} );
$twig->parsefile( 'testfile.xmi' );

sub uml_class 
{
    my ( $twig, $section ) = @_;

    my $elt = $twig->first_elt;
    my $struct = $elt->simplify( forcearray => 1 );

    print Dumper( $struct );
    # parse the block and extract the data elements

    $twig->purge;
}

[download]

I can't seem to figure out how to grab a UML:Class block and send it to a parser - it always contains additional root elements and/or is missing child tags. I think I can figure out how to write separate parsers for the UML::Class, UML::Attribute, UML::Association, and UML::TaggedValue tags, but I need to isolate them first.

I must be making this too difficult. Could someone clue me in, please?

Thanks in advance!

Replies are listed 'Best First'.
Re: Pulling out sections of an XMI file with XML::Twig
by GrandFather (Saint) on Sep 27, 2006 at 21:35 UTC

    Is not $section what you want as a starting point?

    sub uml_class
{
    my ( $twig, $section ) = @_;

    $section->print ();
}

    [download]

    Prints:

    <UML:Class name="EARootClass" xmi.id="EAID_..."/><UML:Class name="Data
+Type" xmi.id="EAID..."><UML:Classifier.feature xmi.id="EAID...">

<!-- I want to pull out the following UML:Attribute blocks -->
                                <UML:Attribute name="dataTypeId"></UML
+:Attribute><UML:Attribute name="name"></UML:Attribute></UML:Classifie
+r.feature></UML:Class>

    [download]

    Update:

    Note the "Processing just parts of an XML document" section in the XML::Twig documentation that describes twig_roots and in particular the line my( $t, $elt)= @_; in the sample code in that section.

    Update:

    and to reparse the sub-element:

    sub uml_class
{
    my ( $twig, $section ) = @_;
    my $xml = $section->sprint ();
    my $subTwig = XML::Twig->new ( twig_roots => { 'uml:attribute' => 
+\&uml_attr});
    
    $subTwig->parse ($xml);
}

sub uml_attr {
    my ($twig, $elt) = @_;
    
    $elt->print ();
    print "\n";
}

    [download]

    Prints:

    <uml:attribute name="datatypeid"></uml:attribute>
<uml:attribute name="name"></uml:attribute>

    [download]
    DWIM is Perl's answer to Gödel
[reply]
[d/l]
[select]

      *click* ...a lightbulb turns on...

      Thank you for the example code - that made all the difference! I didn't realize that $section was actually an XML::Twig::Elt object. Using your example as a starting point, I was able to get what I needed.

      At first I thought I needed to parse the XML chunks into a data structure (such as that returned by XML::Simple), which I accomplished using 

      my $struct = $section->simplify( forcearray => 1 );

      [download]
      but I soon realized that was overkill for what I really needed - the value of the attributes for the class (etc) tags. The $elt->att( $attribute ) method did the trick. An example of the class handler from my working code is below. 
      sub uml_class
{
    my ( $twig, $section ) = @_;

    print "data for class:\n";
    print "   name = ", $section->att( 'name' ), "\n";
    print "   xmi.id = ", $section->att( 'xmi.id' ), "\n";

    my $subTwig = XML::Twig->new( twig_roots => {
                                      'UML:Attribute' => \&uml_attr } 
+);
    # $subTwig->parse( $xml ); # original code (typo)
    $subTwig->parse( $section->sprint() );
}

      [download]

      Thanks again! I knew I was making it too hard. :-)

      Update: corrected typo in the example code

[reply]
[d/l]
[select]

        I must admit that I don't quite understand what you are doing, or even trying to do, but it looks like you are parsing things several time ( the call to parse in uml_class, but I fail to see what's exactly in $xml). It should not be necessary, you can set handlers at different levels of the tree (not twig_roots, but regular twig_handlers).

        If you could repost a complete example I might be able to help.

[reply]

Back to Seekers of Perl Wisdom