[Cdk-devel] RFC: Subclassing CDKException?

[E.L. W. <eg...@us...>]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


Hi all,

some long time ago 'we', i.e. the project, decided that we should only use=
=20
CDKException, and not subclasses of this exception...

However, I've experienced this as a burden, more than that it reduced=20
complexity in the system. It would be easier to trace a problem if we would=
=20
use subclassed CDKExceptions, because the Exception class name is prominent=
=20
in the stacktrace, and handling exception in source code is much more=20
specific...

Therefore, I think it's time to discuss the issue again, and would like to=
=20
propose this:

Let's have a core CDKException which will be thrown in any situation where=
=20
something goes wrong in the CDK algorithms, or when some weird condition is=
=20
encountered. But, in cases where default Java exception can be used, they=20
must. For example, when a wrong argument is given to a method, then the=20
IllegalArgumentException must be used.

And this class may be subclassed for a specific (java) package.... for=20
example, cdk.qsar.DescriptorException and cdk.io.CDKIOException...

At this moment, many subclasses of CDKException are used, so the standing=20
'rule' is not enforced.

Egon

=2D --=20
eg...@us...
=2D ---------------------------------------
CDK: http://cdk.sf.net/
JChemPaint: http://jchempaint.sf.net/
Jmol: http://www.jmol.org/
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (SunOS)

iD8DBQFBpFRWd9R8I9Yza6YRAkD0AJ9sCOEY9VuhpzJLUKgTMzKXzfWKeACeLyv2
p9Cys4Y/tnUAw3YJzXshfDI=3D
=3Dwp4y
=2D----END PGP SIGNATURE-----