GUS XML and XML handling in the GUS object layer
------------------------------------------------

The GUS distribution includes a general purpose plugin that accepts data in XML format 
and loads it in GUS, handling a limited set of key constraint cases. 
This XML is called "simple GUS XML":

        my $purpose = <<PURPOSE;
        Insert data in from a simple GUS XML format into GUS.
        PURPOSE
        my $purposeBrief = <<PURPOSE_BRIEF;
        Insert data in from a simple GUS XML format into GUS
        PURPOSE_BRIEF 

       - excerpt from LoadGusXML.pm, GUS distribution


In GUS 3.0 was /GUS/Common/plugin/perl/UpdateGusFromXML.pm which was replaced 
in GUS 3.5 by /GUS/Supported/plugin/perl/LoadGusXml.pm

The plugin is devised to load data from "simple GUS XML format" into GUS. 
'GUS XML', coined by the staff in CBIL, is extended in the context of XMLGUS, 
to define a format that manages fully the input data key constraints, allows
for specialized processing, and other features (see references below). 

The XML used by CBIL as input is in the form:

<Schema::table>
      <column>value</column>
      ...
</Schema::table>

This is very similar to the GUS XML recognized by XMLGUS except instead of underscore 
CBIL uses '::' to delimit schema and table names.

In addition, an extension to the CBIL format is introduced to control commits to the database. 
'//' at the beginning of a line denotes objects that need to be submitted 
to the database since the last commit.

The CBIL plugin returns the number of processed, inserted and updated objects.

The only reference found for usage of LoadGusXml is for entering User, Group, and Project 
entries in the database in the User Guide

To add a project this is the sample XML used:

<Core::ProjectInfo>
     <name>Sample project</name>
     <description>A sample project</description>
</Core::ProjectInfo>
//

The plugin is invoked with:

ga GUS::Supported::Plugin::LoadGusXml --filename sampleproject.xml --commit


If elements referring to tables are nested in the XML, LoadGusXml will create the appropriate 
foreign keys between tables. Candidate keys for reserved names, such as DNA, are not resolved
by simple GUS XML.

Usage notes about LoadGUSXml.pm (restrictions with simple GUS XML):
  1. The elements referring to columns must be in the order they appear in the tables.
  2. If a primary key column is specified it must be in lower case, the other column names 
    are not case sensitive
 3. A column must be all on one line with no leading/trailing whitespace


Details about (not simple) GUS XML can be found at
http://flora.ittc.ku.edu/xmlgus2.0/help.txt and in Clark, et al., 
Applied Bioinformatics 2005, 4(1):13-24.