[Pymerase-cvs-commit] Docs/input/modules parseXMI.tex,NONE,1.1

[<ki...@us...>]

Update of /cvsroot/pymerase/Docs/input/modules
In directory sc8-pr-cvs1:/tmp/cvs-serv18028

Added Files:
	parseXMI.tex 
Log Message:
first version of parseXMI docs.


--- NEW FILE: parseXMI.tex ---
% Pymerase Docs - parseXMI
% Copyright (c) California Institute of Technology
%
% Authors: Brandon King
% $Revision: 1.1 $
% Modified $Date: 2003/05/07 23:27:05 $

\section{\cb parseXMI}
\subsection{\cb Description}
Provides support for using UML to provide information to
Pymerase. Currently three UML programs generate XMI that is parseable
by Pymerase -- Poseidon, ArgoUML, and SMW UML. 

\subsection{\cb Usage}
Use one of the four methods listed in 'Pymerase Docs - Running
Pymerase' and input a .xmi and/or .zargo file.

\subsection{\cb Caveats}
\subsubsection{\cb XMI Standard}
Unfortunately, UML and XMI are under active developement, so new
versions come out fairly often and they aren't always backwards
compatible. This makes it difficult for Pymerase to support all
versions of XMI. When you are using a UML program that works,
upgrading it may or may not change the compatibilty of Pymerase with
that UML program.

\subsubsection{\cb Object Model vs Relational Model}
There are significant differences between an Object Model (UML) and a
relational model (relational database). Because of this, UML doesn't
always provide all the information you need to when providing
information to Pymerase. In these cases it can be better to use the
table.dtd XML format which is read in by the parseGenexSchemaXML input
module. You may, however, use UML and then output table.dtd XML files
using the Pymerase CreateTableXML output module, add more information,
and feed the XML files back into Pymerase.