Hi Gilles,
I made the change because it seemed logical (to me) and sufficiently trivial
(internal class with 2 usages). I don't really care too much either way so I'll
go ahead and revert it. Commons Math has MathInternalError so there is
precedent for this naming convention.
In regard to it not being part of the public API, this is indicated by its
presence in the "o.a.c.geometry.core.internal" package.
-Matt
________________________________
From: Gilles Sadowski <gillese...@gmail.com>
Sent: Sunday, July 5, 2020 2:12 PM
To: Commons Developers List <dev@commons.apache.org>
Subject: Re: [commons-geometry] 11/11: change GeometryInternalError to
GeometryInternalException to better reflect its actual type
Hi.
Le dim. 5 juil. 2020 à 13:26, Matt Juntunen
<matt.juntu...@hotmail.com> a écrit :
>
> What are your objections to the name? The new name is more consistent with
> JDK conventions.
It is a programming error (as opposed to a usage error, which
subclasses of "RuntimeException" represent). As such it is
more self-documenting (IMHO) to keep the previous name.
JDK's use of the English word "Error" should not prevent all
application to use it. There is no implied relationship to the
JDK's "Error" hierarchy.
>
> Also, this class is not part of the public API.
An alternative to make it that clear is perhaps to have an
"InternalUtils" class (untested):
public class InternalUtils {
private static String INTERNAL_ERROR_MESSAGE = "...";
public static void throwInternalError() {
throw new GeometryInternalError();
}
private static class GeometryInternalError extends IllegalStateException {
// ...
}
}
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
