[sbml-discuss] Release of libSBML-5.19.0

[Sarah Keating]

---------------------------------------
libSBML-5.19.0
----------------------------------------

This is our first release on GitHub.
https://github.com/sbmlteam/libsbml/releases/tag/v5.19.0


The usual collection of source and binary files can still be found on
SourceForge.
https://sourceforge.net/projects/sbml/files/libsbml/5.19.0/

Python installation is available via PyPI.

We also now have a package on CondaForge.
Many thanks to Moritz Beber for his persistence and help in making this
happen.

------------------------------------------------------------------------------------------

The libSBML source code has moved to GitHub.
(https://github.com/sbmlteam/libsbml).

The stable branch reflects the latest released code. The development
branch is the main branch and we encourage users to create and
submit pull requests against this branch. We are no longer
maintaining a separate repository for experimental code for as yet
unfinished packages.

Please use the GitHub issue tracker to file any reports or submit a
fix and do a Pull Request.

We will gradually move away from SourceForge, but for this release
code will be available from both SourceForge and GitHub.

-----------------------------------------------------------------------------------------------
* Addition of support for the SBML Distributions package

- Now that the first official version of the specification for the
SBML Level 3 Distributions package is available, the libSBML API for
'distrib' is now included in the stable release of libSBML. This
means that all prebuilt binaries for the stable release will
include the 'distrib' package code. The src archive containing
libSBML core code will continue to be available, with additional
archives of the individual package code.

NOTE: The libSBML GNU make-based build system has not been updated
for packages. Thus, to build from src, it is necessary to
use the CMake build system.

* New features

- Functions get/set/isSetMath have been added to the SBase class.
These do not allow interacting with a non-existant <math> element
but may facilitate using more generic functions.

* Bug fixes

- There was a memory leak introduced in the SWIG wrapper. Thanks to
Fabian Fröhlich for spotting and fixing this.

- The SBMLNamespaces class was being leaked in a number of places,
these have now been plugged.

- The MathML to infix parser was also leaking memory. Thanks to
Liam Keegan for spotting and fixing these.

- The validation code detecting assignment cycles was incorrectly
including local parameters within its search. These have now
been removed from the relevant code.

- Conversion code was failing to detect when a local parameter used
the 'id' of a speciesReference which is allowed in SBML Level 3.
These can be converted to SBML Level 2 and the code has now been
corrected so this conversion will take place.

- The L3 infix parser has been improved to recognize the use of a
name without parentheses as a misused function name or a csymbol,
if appropriate.

- There was an anomaly when using the addCVTerm() and
appendAnnotation() functions which meant that if used in a certain
order information could be lost. This has been corrected. Thanks
to Colin Halupczok for the detailed report.

- Converting a <localParameter> to a global <parameter> object was
failing if the metaid attribute was set. This has been changed to
produce the correct behavior.

* Configuration/build system changes

- Import of the python library has been sorted in line with recent Linux
changes to the location of the shared file.

- Code has been fixed to be MingGW compliant.

- The build configuration was updated to work with the latest version
of R.

* Miscellaneous

- A number of memory leaks and possible access violations have been
solved.


STABLE PACKAGES


- 'distrib' package-specific updates:

- A new converter, AnnotationToDistribConverter, has been added
to allow conversion from the existing annotation format used
with <functionDefinitions> to the new 'distrib' format. This
complements the existing DistribToAnnotationConverter that converts
'distrib' to annotations.


* Bug fixes

- 'comp' package-specific bug fixes:

- Comp flattening dropped conversion factors for particular
instances of assignment rules. This has been corrected.
Thanks to Matthias König for reporting this.

- 'distrib' package-specific bug fixes:

- There was unfortunately an infinite loop in the code using
the <uncertainty> element. This has been removed.

- 'render' package-specific bug fixes:

- Some render classes were not handled correctly by language
bindings. This has been fixed.

EXPERIMENTAL PACKAGES

* New features

- 'spatial' package-specific updates:

- Compression in the 'spatial' package is now based on strings.
The code has been changed to reflect this.

- Support for int and uint as DataKinds has been added to the
spatial code in line with changes in the specification.

- A range of validation tests have been added to the spatial
code. As this is an experimental passage package still in
development, we cannot guarantee that all validation has been
implemented.

* Bug fixes

- 'spatial' package-specific bug fixes:

- There were some memory leaks in the spatial code.
Thanks to Liam Keegan for spotting these and congratulations
for being our first GitHub user to address problems using a
Pull Request.


=====================================================

Please report any problems you encounter using libSBML either to the
mailing list libsbm...@googlegroups.com or using the issue tracker for
libSBML at

https://github.com/sbmlteam/libsbml/issues

Thanks,

the LibSBMLTeam