CHANGES

Changes wrt previous versions:
------------------------------

dd. 7 april 2008
Corrected support for writing the data - the placeholder problem
has been solved.
This is now version 1.00

dd. 27 december 2007
Added support for writing the data to an XML file in
xmlreader.f90 (the program now generates a routine write_xml_* as well
that writes a file with the same structure as the routine read_xl_*
reads). There is one problem left: if the definition contains
"placeholders", the code is not written in the correct order.

dd. 7 december 2007
Added new predefined types, so that a specification like:

<?xml version="1.0"?>
<definition>
    <options rootname="words"/>
    <variable name="word" type="word" dimension="1"/>
</definition>

works as well (for all primitive types)

Added the start of a program to convert XSD-files.

dd. 25 june 2007
Adjusted the xmlreader program to properly accept smaller arrays
for fixed shape variables.

dd. 19 june 2007
Corrected a mistake with character variables with fixed shape.
Allow larger arrays than according to shape

dd. 17 june 2007
Updated the documentation, prepared release of version 0.97

dd. 12 june 2007
Added support for arrays of fixed "shape" (see
the menuitems example)

Fixed a bug with handling the default values for variables
(the code ended up in the wrong place)

dd 10. june 2007
Added routines to write basic data to an XML file.
(Start of implementation of generating export routines
in the xmlreader program)

Corrected bug in xmlreader program (had to do with
whitespace that interfered with the processing in the
generated reader code)

NOTE: Still need to check that this solves the
problem entirely!

Applied patch by Davide Cesari: array of strings and
initialising pointer components.

dd 26. february 2007
Fixed a bug in the xmlreader program wrt reading
attributes as data (self-contained elements were
sometimes skipped - i.e. <element data="xxx"/>
was not treated properly in all cases)

dd 1. july 2006
Corrected the initialisation of the array "entities"
- the strings must all be the same length (not
all compilers will complain about it though).

And some corrections by Jacques Germishuizen for the
examples.

dd. 26 june 2006
Fixed a few bugs pointed out by
Jacques Germishuizen (rootname, double precision
variables)

dd. 16 june 2006
Added a small configuration file to the
xmlreader program - some global settings and
a way to work around a limitation in some
compilers.
Also added a facility whereby attribute
values are treated as data for a subelement:
In the menu example (tst_menu.f90, menu.xml,
menuitems.xml) the following fragments are
equivalent:

<item type="item" name="Open ..."/>

and

<item>
   <type>item</type>
   <name>Open ...</name>
</item>

dd. 9 june 2006
Corrected a severe misunderstanding about XML -
the requirement that there is a root element
This has now been built into the xmlreader
program.
Some refinements to xmlparse.f90 (extra check
on <?xml ...?> elements) and added proper
comments.

dd. 5 june 2006
Added two examples
Added subroutine xml_process
Updated the documentation
Corrected bugs in xmlreader.f90

dd. 30 may 2006
Added TODO file to the project
Added support for "strict" parsing of the
XML file via the generated reader routine.

dd. 28 may 2006
Solved a problem with place holder tags
(anything after the closing tag was
ignored)

dd. 17 may 2006
Corrected a few glitches in xmlreader.f90 -
The code generated for place holder tags
was incorrect.

Note: There is still a bug which appears
in example tst_grid. With the tag "general"
in place, the gridarray tags are not seen:

<?xml version="1.0">
<comment>
Read a couple of integers and a derived type
</comment>
<grid><X>10</X><Y>20</Y>
</grid>
<general>
   <A>1</A>
   <B>2</B>
   <C>2</C>
</general>
<grid_array><X>11</X><Y>22</Y></grid_array>
<grid_array><X>12</X><Y>23</Y></grid_array>
<grid_array><Y>23</Y></grid_array>

Maybe an ordering/loop problem?

dd. 7 february 2006
Corrected a stupid mistake in xmlreader.f90 -
the spurious error message is gone

dd. 6 february 2006
Added placeholders to the xmlreader utility. More
thorough tests required: the generated source code is
composed of several pieces, the ordering can be
wrong when the XML file is ordered differently.

Note:
The generated code is not tested yet. It looks okay,
but that is all.

dd. 24 january 2006
Extended the xmlreader utility with new features.
(mostly support for default values)


dd. 19 october 2005
Added support for "entities" (like an ampersand) in
character data.

Expanded the capabilities of the xmlreader program
with arrays of primitives and derived types.

Updated the project documents at http://xml-fortran.sf.net


dd. 3 october 2005
Integrated a patch by cinonet with the file xmlparse.f90.

Updated the documentation with information on the xmlreader
program.

New version of the test program (more flexible) for the
parser. Not finished yet.

New version of the xmlreader program - added support for
derived types.


dd. 29 september 2005
Added two articles on the use of XML in Fortran that appeared in
the Fortran Forum.

Updated the project page - new information on available libraries among
other things.

Added new code:
- tree_struct.f90 and tree_find.f90:
  Module for manipulating general tree structures

- read_xml_prims.f90 and test_read_prims.f90
  Code to read (integer) data returned from xml_get.
  Will be part of a new approach to dealing with XML files