[
https://issues.apache.org/jira/browse/MATH-195?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12576784#action_12576784
]
Phil Steitz commented on MATH-195:
----------------------------------
Could be though we had it wrong in [math] to begin with. IIRC, early on there
were discussions about cluttering client code with catches for checked
exceptions due to bad arguments that amounted to violations of the API contract
- i.e., violations of stated preconditions or "nonsensical" arguments. For
example, trying to get the (not generalized) inverse of a matrix that is not
square. We agreed that it was better to make these RTEs and in many cases what
is thrown is IllegalArgumentException.
We knew from the beginning that the exceptions hierarchy in the initial
releases of [math] was weak and that we would have to refactor / rethink it if
the little library was to grow. Well, now it has grown, largely thanks to the
Luc's Mantissa contribution and I think it is fair for us to rethink. I think
we should do that now and I would be OK with any of
a) leaving "legacy" unchecked exceptions as is for backward compatabiltiy but
adopting the Mantissa guidlines above moving forward
b) merging the two sets of guidelines by defining a category of "truly illegal
argument" cases where we throw sublcasses of IAE, but keeping the Mantissa
guidelines for non-obvious violations of preconditions. Perhaps an example of
that would be NotPositiveDefiniteMatrixException.
c) make all exceptions associated with arguments subclass IAE
d) deprecate "legacy" methods throwing IAE and replace with methods advertising
checked exceptions (this one might be painful)
The least traumatic is b) and I think that may be possible to precise, so that
has my vote. Could be a) and b) amount to the same thing.
> Inconsistencies in the Exception hierarchy
> ------------------------------------------
>
> Key: MATH-195
> URL: https://issues.apache.org/jira/browse/MATH-195
> Project: Commons Math
> Issue Type: Improvement
> Affects Versions: 1.1, 1.2, 2.0, Nightly Builds
> Reporter: Thomas Chust
> Assignee: Brent Worden
> Priority: Minor
> Fix For: 2.0
>
> Original Estimate: 0.17h
> Remaining Estimate: 0.17h
>
> I think there are some inconsistencies in the hierarchy of exceptions used by
> the Commons Math library which could easily be fixed:
> - org.apache.commons.math.linear.MatrixIndexException should really be a
> subclass of
> java.lang.IndexOutOfBoundsException.
> - Either org.apache.commons.math.linear.InvalidMatrixException should be
> checked or
> org.apache.commons.math.random.NotPositiveDefiniteMatrixException and
> org.apache.commons.math.geometry.NotARotationMatrixException should be
> unchecked, but in any case the latter two
> should be subclasses of the former. An unchecked InvalidMatrixException
> should probably be a subclass of
> java.lang.ArithmeticException while a checked InvalidMatrixException
> should be a subclass of
> org.apache.commons.math.MathException.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
