Re: CLP(FD) team for the ASP solver competition



On Thu, 26 Feb 2009 00:55:38 GMT, ulrich@xxxxxxxxxxxxxxxxxxxxxxxxxx
(Ulrich Neumerkel) wrote:

A.L. <alewando@xxxxxxxxxx> writes:
On Wed, 25 Feb 2009 22:26:40 GMT, ulrich@xxxxxxxxxxxxxxxxxxxxxxxxxx
In create_shipment_/4 you have your own convention for terms to throw.
Consider throwing the standard terms (in those situations applicable).
The reason is that the standard terms are best understood - explained
in 13211-1 7.12.2 Error classification.

No. We have our own standards for exception format, and this format is
not restricted to Prolog, but spans over all other languages we are
using. Since Prolog is called from Java and Java is parsing and
processing Prolog exceptions, format of the exceptions must fit the
common standard.

From the outside for me, it is certainly impossible to judge that.

Some errors are very specific to Prolog - there it is particularly
interesting to consider the ISO model. ISO error terms look like this:

error(Error_term, Imp_def)

Where Error_term is defined by ISO, and Imp_def can be used by you.
So there is a prevision to put several conventions together.

I can do whatever I want on Prolog side, but if exception is
probagated outside of Prolog, then my freedom is restricted.
Therefore, my life is much easier if I restrict the format from very
beginning and translate whatever exception/error Prolog generates to
standard fromat as early as possible

A.L.
.



Relevant Pages

  • Re: CLP(FD) team for the ASP solver competition
    ... Consider throwing the standard terms. ... not restricted to Prolog, but spans over all other languages we are ... interesting to consider the ISO model. ...
    (comp.lang.prolog)
  • Re: WANTED: Volunteer to Scan Old Programs
    ... BASIC to the American National Standard for Minimal BASIC, ... Implementation-defined features ... Error and exception reporting ... of interpretation, and Section 5 for information peculiar to ...
    (comp.os.cpm)
  • Re: c++ documentation
    ... > The standard is terse and somewhat difficult to read. ... >> and seem less like a black art to the novice. ... Yes, you're right, calling algorithms "number crunching" totally misses ... > than abort when an exception is thrown. ...
    (comp.os.linux.development.apps)
  • Re: A minimal module system for Prolog
    ... The only Prolog implementation that I'm aware whose developers had the ... Solving these "issues" results necessarily in a new module system. ... What good is a standard without actual implementations and a user ... facto standard for programming in the large using Prolog. ...
    (comp.lang.prolog)
  • Re: Allocators and exceptions
    ... Nothing in the Ada standard is about "goodness". ... nothing anywhere in the standard that resources like memory ever get ... Never, ever, write code that intentionally raises an exception during ...
    (comp.lang.ada)